Sean Pitman yet again (with apologies)

[sweetnes...@yahoo.com]

Sorry for the delay, much more important things got in the way. And
apologies for starting a new thread; apologies to everyone else, too,
for starting so many threads involving rebutalls of this half-assed
theory. But let us continue.

I will begin by exploring yet another catastrophic error in your
calculation:

>You have shown that short amino acid sequences can
>come together to form a new unified function that
>is indeed unique and of greater complexity. The
>only problem you have here is that the minimum amino
>acid number required for your most complex example
>is only 3 or 4 hundred amino acids. My argument is
>that evolution becomes more and more difficult the
>greater the minimum amino acid requirement until it
>becomes impossible this side of zillions of years
>when the minimum requirement reaches a few thousand
>amino acids in fairly specified order.

I have given many reasons why this is wrong already, but let us go on
to some new ones, until the state of affairs becomes painfully
obvious.

Let's start with a simple comparison. Let's take an imaginary protein,
call it A, size 300. It acquires a point mutation, becoming B, still
size 300. According to you, there is no problem in this happening,
since the number of amino acids is low.

Now, the same protein A acquires a large insert, and its size
increases to, say, 2000. This is now protein C. According to you,
there is no way whatsoever that this new protein could have a
beneficial effect on the cell.

I am mystified. Why?

It is because the strawman you are presenting has large proteins being
built up by completely random association of amino acids. This, as I
have pointed out again and again, does not happen at all.

Let's say (erring on your side of the argument) that one mutation in
ten billion produces a significantly beneficial new function. So, for
every ten billion point mutations in small proteins, there will be one
that will produce a new small protein, with a new, beneficial
function. Also, for every ten billion large insertions, there will be
one that will produce a new LARGE protein, with a new, beneficial
function.

So even ignoring myriad other factors that go against your theory (I
will add some more in this post too), just sticking to the most
general, basic facts of genetic mutation, your entire spiel on how it
is "more difficult" to produce a large protein then a small one fails
completely.

Let's carry this further. Because of domains, their recombination, and
some thermodynamic factors (which I will cover below, in answer to
another particular whopper of yours), larger proteins are far more
versatile then the small ones - it is *easier* to adapt a large
proteins for a completely novel function then a small one; you simply
have more space to work with. This can be seen by wide varieties of
independently produced structural proteins, membrane proteins,
signaling complexes, etc; all of which depend on certain protein
architecture rather then your "amino acids in specific order". The
truly difficult feats are coming up with small but very
sequence-specific proteins, such as, say, proteins that facilitate
glucose transport across membranes.

>Where are your examples of evolution requiring such a
>level of minimum amino acid specificity? I see a lot
>of hot air coming from you, but no such example.

There is no such thing as minimum amino acid specificity. You invented
it. I cannot show you something that exists only in your imagination.

To point out just one piece of evidence you completely ignored, I have
first asked you to look up some sequence alignments, and then (after
you ignored my suggestion) I have provided you with a link to one.
There you have several proteins that perform the same function, but
agree with each other, sequence-wise, only ~20%. There are few
exceptions to this rule, and no exceptions that I am aware of among
large proteins (the larger the protein, less % agreement).

If a protein needs to have a "minimum" number of amino acids in a
"specific sequence", how come that so many different-sized proteins
(sometimes tenfold size difference for same function, same efficiency)
with wildly diverging sequences can produce the same effect? If you
want a correct definition of what is required, you have to say that it
is necessary to have certain small stretches of specific amino acids
occuring somewhere on the protein. Which screws up your calculation,
and allows for directed evolution to be efficient (see references
below).

>This is good thinking since it is actually an attempt
>to controvert my hypothesis whereas you haven't even
>tried to present the de novo evolution of anything with
>a minimum function requiring over a few hundred amino
>acids working together at the same time.

I did, among other things, a search on your previous posts. As far as
you are concerned, if something is observed in lab, it is not
sufficiently complex; if it hasn't been observed in the lab, it is
obviously too complex to arise spontaneously. The approach does not
suprise me, but let's see if something can be done about your
credibility.

If you paid any attention to what I said, you would have found that
many of the proteins I spoke about (including those in pterine
pathway) require many large proteins to function together. It is just
that the one which provided a clue was 300 residues long. The next one
is 800. You could have also noted that novel proteins of sizes
measured in thousands of residues arise in labs all the time, only
slightly less frequently then the smaller ones; usually, they evolve
down in size, since large protein sizes are inefficient (contrary to
your claim that they are *necessary*; again, see references below).

>What you have shown are simpler systems combining to
>produce a novel function that requires, at minimum,
>only 3 or 4 hundred amino acids. You have also provided
>examples of a simple function requiring less than a few
>hundred amino acids at minimum evolving a new type of
>function within the same level of complexity, but not anything
>much greater. Please, I think you can at least try and
>do better than this.

Sure. As illustrated above, there is no correlation between size and
effects of a mutation. Since you repeat the same argument many times
in your posts, I will repeat mine a few times, in different forms, to
attempt to minimize the amount of smoke you will try to cover them
with.

Your model is this: there is a function X, that requires at minimum a
10000-residue protein. The evolution works like this: it creates a
random 10000 residue protein. Then it modifies one amino acid at a
time, until it produces the protein that performs function X. You say
that this is impossible, since it would take impossibly long to search
through all the combinations of amino acids.

As someone has nicely put it before, well, duh.

The evolution actually works a bit differently. I will illustrate this
in two steps. For the first step, let's say that you are right, and
that there is an absolute minimum for a size of a protein before it
can perform function X; so 10000 residues or nothing. In this case, it
is hit and miss. The population will relatively frequently produce
10000-residue proteins; occasionaly, some of them will have a
beneficial function, and will be conserved. If one that is close
enough to peform function X arises by accident, then the cell will be
able to perform function X. If not, cell will not be able to perform
function X, and will have to live without it.

There are some proteins that fit this description. For examples, you
can compare the cell structures of eukariotes with those of
prokaryotes, or animal cells with plant cells, or diatoms with
amoebae... Occasionaly, a group will stumble upon something great, and
will develop it further. Other groups have to do without it. See
"nested hierarchy" for examples too.

One thing that is absolutely certain even here is that you *will*
eventually have *many* very large proteins performing many different
functions within the cell. You might not have one that performs
functon X. Luck of the draw.

Second step of my illustration is a bit futher away from your strawman
and closer to evolutionary genetics. In real world, there is no such
thing as minimum residue requirement; small polypeptides can perform
analogous functions quite well (small polupeptides that perform the
galactosidase function - which should be impossible if your theory is
correct - have been shown to perform quite efficient catalytic
function). Those small polypeptides get domains added to them, which
then provide specificity, signaling, and all other functions a large
protein has. This is, again, *observed to happen* in lab, especially
during directed evolution experiments (which have the benefit of
speeding up things to an observable rate).

>You haven't explained how my ideas are wrong in this regard.

I have, but my words fell on deaf ears. I did it again just now. I can
hardly wait to see the semantic acrobatics you are going to employ to
get around everything written above.

>but not how new types of function within higher levels
>of complexity can be evolved without searching randomly
>through a very large non-beneficial sequence space.

So, let us repeat again. Since evolution is not teleological (it
doesn't go after specific functions), and since it does not work by
producing large proteins and then mutating one amino acid at a time
until it hits the right sequence, *and* since proteins (especially
large proteins) *do not* requre large stretches of specific amino
acids to perform specific functions, there is never such a thing as a
"search through non-beneficial sequence space".

>However, with each amino acid increase in the minimum
>random walk required to achieve the success of any new
>type of beneficial function within that level of
>functional complexity, the time required grows exponentially.

If you are going to a specific place. Let us go through yet another
example.

Imagine the cell as a man who walks around in a room. Each minute, he
steps in a random direction. Room is, say, one mile by one mile in
radius. In the room are spread, say, ten thousand functions that 50
residue proteins can possibly have (imagine them as small dark spots
where the function is optimal, surrounded with vast grey areas where
function somewhat works, but not very well).

It is obvious that the cell will frequently step on various grey
areas, and produce various 50 residue proteins, right?

Now, let's increase the size of the protein to 5000 residues. Now the
room is the size of a continent. The man continues to walk randomly.
Your argument is: there is no way that, randomly walking, the man
could cross from this particular spot in LA to that particular spot in
NY without taking trillions of years. And you are right.

But that has no effect. You see, the functionalities of proteins are
as dense on a continental scale as on the room scale. If 50 residue
proteins can have ten thousand different functions, 5000 residue
proteins can have trillions of different functions. These are spread
around, covering the whole continent.

So the men walk, multiply, occasionaly fall down cliffs and into
ditches...one of them ends up in Las Vegas, one of them ends up in
Miami, one of them ends up in New York. Then I come around and call
one "plant", the other "bacteria", the third one "animal"...

Then you come along, and say "there is no way this man could have ever
reached New York by walking randomly!" Of course not. But he wasn't
going there in the first place. He was waking randomly. He just
happened to be there when you looked.

And, the buildup of complexity is a factor also. Take five people.
Land them on a new, Earth-like planet. Tell them to build a computer.
They will fail: they do not have the tools, the expertise or the
infrastructure. It takes thousands of years to build the
infrastructure necessary for someone to produce a computer; but this
does not mean that production of a computer is a momentuous task when
infrastructure is present. It is difficult, but not impossibly hard.

>Ok - take, for example, a particular function that
>requires, at minimum 5,000aa at minimum to be realized.
>Say this sequence happens to get duplicated so that it
>can undergo various mutations without risking significant
>loss to the original beneficial function (which has been
>optimized for its host by now - as far as *level* of
>function is concerned). The sequence space at this level
>of complexity is more than 10e6500 sequences.

Let's point out another error in this. Genetic code is degenerate;
which means that the mutation that produces methionine at a certain
place is far less likely (what, 64 times less likely?) to come up then
one that produces glycine. Since mutations happen at genetic level,
your numbers are utterly meaningless.

>The question is, out of all of these universes of possibilities,
>how many are or would be beneficial to the given organism in
>question? If the organism could use a million different types
>of functions to some benefit at this level of complexity and if
>each of these million different types of functions had at least
>10e1000 different sequences with at least some selectably
>advantageous level of function, then there would be 10e1006 total
>beneficial sequences in sequence space with a different type of
>function.

Completely meaningless calculation. You (and I) have absolutely no
means to ascertain how many of all possible mutations might be
beneficial to the organism. Your calculation is further meaningless
since the protein of that size will fold up into many separate
domains, and it will be its final configuration that will determine
its use; the number of different combinations of domains and motifs is
far lower then the number of different amino acid configurations (read
the discussion of domains and subdomains below). There is no reason
whatsoever to assume that their density is lower then the density of
beneficial functions of the smaller proteins.

The problem here is cost (is the function worth the cost of
production), and optimization takes care of that. The protein either
gets kicked out, or gets funneled to another function that "requires"
that size, where the cost is worth it.

There is more on this just a little bit below.

>Of course, you will say that evolution doesn't work
>like this. Evolution works by getting various fully
>formed and functional sequences to get pasted together
>to gain a new collective function of higher complexity.
>What many don't realize is that the same statistical
>problems are involved here.

I have to ask yet again: why are scientific journals so obstinate in
refusing to print your obviously correct calculation?

(Clue: it's not correct.)

And no, evolution does not work in that way. I have written quite a
bit on this above.

>You certainly haven't been able to provide me yet with
>any such real time demonstration. I'm still waiting . . .

Sigh. Yes, I have. I keep hoping that you might take a peek into the
Journal of Molecular Evolution, or any other journal dedicated to this
line of research (which might get you to ask yourself just how idiotic
all those people are, writing all those papers and publishing all
those journals, when everything they do has been proven incorrect by
your simple calculation) but wishing does not make it so. But never
mind.

First, Frances Arnold kicks ass:

Peters MW, Meinhold P, Glieder A, Arnold FH. "Regio- and
enantioselective alkane hydroxylation with engineered cytochromes P450
BM-3." J Am Chem Soc. 2003 Nov 5;125(44):13442-50.
Collins CH, Yokobayashi Y, Umeno D, Arnold FH. "Engineering proteins
that bind, move, make and break DNA." Curr Opin Biotechnol. 2003
Aug;14(4):371-8.
Wintrode PL, Zhang D, Vaidehi N, Arnold FH, Goddard WA 3rd. "Protein
dynamics in a family of laboratory evolved thermophilic enzymes." J
Mol Biol. 2003 Mar 28;327(3):745-57.
Sun L, Bulter T, Alcalde M, Petrounia IP, Arnold FH. "Modification of
galactose oxidase to introduce glucose 6-oxidase activity."
Chembiochem. 2002 Aug 2;3(8):781-3.
Schwaneberg U, Otey C, Cirino PC, Farinas E, Arnold FH.
"Cost-effective whole-cell assay for laboratory evolution of
hydroxylases in Escherichia coli." J Biomol Screen. 2001
Apr;6(2):111-7.

You may note that her work (check out the other articles she put out
in last few years also, this is just a selection) includes complex
evolutionary events (not just point mutations). The second article on
the list involves practicaly completely random evolution of enzymes
that perform VERY specific and VERY complex functions.

Then, your favorite lactase:

Stefan, A.; Radeghieri, A.; Gonzalez Vara y Rodriguez, A.;
Hochkoeppler, A. Directed evolution of b-galactosidase from
Escherichia coli by mutator strains defective in the 3' 5' exonuclease
activity of DNA polymerase III. FEBS Letters (2001)
Alexeeva, Marina; Carr, Reuben; Turner, Nicholas J. Directed
evolution of enzymes: new biocatalysts for asymmetric synthesis.
Organic & Biomolecular Chemistry (2003)
#awa, Yoshihiro. Directed evolution of a new enzyme L-aspartate
dehydrogenase from an amino acid dehydrogenase by DNA shuffling.
Hoffmeister, Yang, Liu, Thorson; work on novel anomeric sugar kinases.
Related work of Chi-Huey et al.
Xia, Gang; Chen, Liangjing; Sera, Takashi; Fa, Ming; Schultz, Peter
G.; Romesberg, Floyd E. Directed evolution of novel polymerase
activities: mutation of a DNA polymerase into an efficient RNA
polymerase. Proceedings of the National Academy of Sciences of the
United States of America (2002)
Camps M, Naukkarinen J, Johnson BP, Loeb LA. "Targeted gene evolution
in Escherichia coli using a highly error-prone DNA polymerase I." Proc
Natl Acad Sci U S A. 2003 Aug 19;100(17):9727-32.
Zhang, Ji-Hu; Dawes, Glenn; Stemmer, Willem P. C. Directed evolution
of a fucosidase from a galactosidase by DNA shuffling and screening.
Proceedings of the National Academy of Sciences of the United States
of America (1997)
Yang, H.; Carr, P. D.; McLoughlin, S. Yu; Liu, J. W.; Horne, I.; Qiu,
X.; Jeffries, C. M. J.; Russell, R. J.; Oakeshott, J. G.; Ollis, D. L.
Evolution of an organophosphate-degrading enzyme: a comparison of
natural and directed evolution. Protein Engineering (2003)
Wiseman, A. Novel cytochromes P450 applications arising from the
directed-evolution of recombinant micro-organisms. Letters in Applied
Microbiology (2003)

A bit of more theoretical stuff, but just so you see some glimpse of
how are people using things that "cannot possibly happen" (and an
analysis as to reasons why the processes are slow and picky):

#oltermann, Andre; Kettling, Ulrich; Haupts, Ulrich; Tebbe, Jan;
Scholz, Peter; Pilling, Jens; Werner, Susanne; Rarbach, Markus.
Process for generating sequence-specific proteases by specificity
screening-based directed evolution using target substrate peptides and
use thereof. Eur. Pat. Appl. (2003)
Cohen, Noa; Abramov, Simona; Dror, Yael; Freeman, Amihay. In vitro
enzyme evolution: the screening challenge of isolating the one in a
million. Trends in Biotechnology (2001)
Farinas ET, Bulter T, Arnold FH. "Directed enzyme evolution." Curr
Opin Biotechnol. 2001 Dec;12(6):545-51.
Wong TS, Schwaneberg U. "Protein engineering in bioelectrocatalysis."
Curr Opin Biotechnol. 2003 Dec;14(6):590-6.

We spoke of domains? I'll give you some more reading material below,
but for now:

Voigt CA, Martinez C, Wang ZG, Mayo SL, Arnold FH. "Protein building
blocks preserved by recombination." Nat Struct Biol. 2002
Jul;9(7):553-8.
#washita, Shintaro; Osada, Naoki; Itoh, Tomohito; Sezaki, Mariko;
Oshima, Kenshiro; Hashimoto, Etsuko; Kitagawa-Arita, Yuko; Takahashi,
Ichiro; Masui, Tohru; Hashimoto, Katsuyuki; Makalowski, Wojciech. A
transposable element-mediated gene divergence that directly produces a
novel type bovine bcnt protein including the endonuclease domain of
RTE-1. Molecular Biology and Evolution (2003)

Regarding your insistence that there are "minimum amino acid
requirements" for a certain function, and your flat-out statement that
~400 residues are required for lactase? See here:

Matsumura, Ichiro; Ellington, Andrew D. In vitro evolution of
beta-glucuronidase into a beta-galactosidase proceeds through
non-specific intermediates. JMB
#ochavi, E.; Bar-Nun, A.; Fleminger, G. Substrate-directed formation
of small biocatalysts under prebiotic conditions. Journal of
Molecular Evolution 1997

Regarding your "antibiotic resistance does not arise spontaneously":

Blazquez J. "Hypermutation as a factor contributing to the acquisition
of antimicrobial resistance." Clin Infect Dis. 2003 Nov
1;37(9):1201-9.

Aires de Sousa M, de Lencastre H. "Evolution of sporadic isolates of
methicillin-resistant Staphylococcus aureus (MRSA) in hospitals and
their similarities to isolates of community-acquired MRSA." J Clin
Microbiol. 2003 Aug;41(8):3806-15.

Enough reading material for you? There is so much more; literally
thousandfold more.

>A bit of clarification: I am well aware that many
>multi-domain proteins use all or many of their domains
>at the same time for a collective function. However,
>when you are taking about a particular function, multiple
>domains may not be required to achieve this type of
>function. The question is, what is the minimum amino
>acid part requirement to achieve a particular type of
>function to a level where it becomes selectably advantageous?

So let me see here. For you to be able to drive a car, it is not
important how many cylinders you engine has, but how many molecules of
iron is the engine made of?

Before I go on to domains, subdomains, and motifs, let me cover a tiny
chemical fact. No matter how large the protein is, it's actual active
site is going to be rather small. There will be one or two amino
acids, and maybe an ion or two involved. So this entire story of
"large proteins" ignores basic chemistry: if protein needs to be very
large, it going to be such only because it needs either to cover a lot
of space (which means that the largest part of it is functionally
irrelevant), or because it needs to do more then one thing (in which
case different domains wil perform different functions).

What is the minimum amino acid requirement? Look here: if you have
more then about three hundred amino acids, you will almost certainly
have two or more domains. Reason is thermodynamical stability. When
you have 2000-residue protein, you will have quite a few domains.

Specific functions require certain domains with certain charges. Let's
take a large channel protein, that transports something across a
membrane; you will find that such channels (when they are one
molecule, rather then a noncovalent complex of many smaller proteins).
What do you need? You need a beta-barell of certain size, with two or
three specific charges distributed on the inside; and with mostly
hydrophobic residues on the outside.

You are familiar that, for a random sequence of residues, there is a
decent chance (~20%) it will fold into a beta-sheet. If you make a
beta sheet with six or more antiparallel strands, it will, due to its
curvature, likely form a beta barell. Therefore, formation of a
protein that has a large beta-barell is not a problem. Hydrophobic
residues are also not a problem: if you generate the code randomly,
most of the side chains will be nonpolar.

The big problem are the charges on the inside: how do you get a
channel that transports one particular amino acid, rather then
anything that comes along? Here, you would say "it has to be
designed".

But it doesn't. Our protocell, where this happens, has a primary goal
of equalizing the osmotic pressure (to a certain point). As shown
above, proteins that can function as channels not only can, but *will*
arise quite frequently by spontaneous mutations. And most of them will
kill a cell, by disrupting the concentration of salts inside of it.
But, as structural studies have shown, there are a lot of them that
will expell only certain sized ions, only under certain concentrations
(i.e. it will transport Ca2+, but only if the concentration
differential is higher then a certain number); this has to do with
overall internal charge, and the size of the opening.

We are talking quite large numbers here. One fifth of all random
proteins produced will form a beta-sheet. The data gets fuzzy beyond
this point, but at least half of those that produce a beta-sheet will
have that sheet rolled into a barell. At least half of those will have
hydrophobic side chains vastly predominant on their outside. At least
one tenth will have the channel of the size that allows ions through,
but nothing else. At least half of those will have predominantly
negative charge within the tunnel, with one or two positive bumps. At
least one in a thousand will allow only Ca2+ ions to pass, due to the
position of those bumps. At least one in a hundred of those will do so
at a concentration differential which is beneficial to the cell (ie
when the cell is about to burst).

So, the chances that a randomly produced protein will form a useful
Ca2+ ion channel are about one in forty million. Those are not bad
odds, considering how many cells are producing how many billions of
new proteins on a yearly basis. Even if you complicate some of the
factors in the above calculation (and there is enough experimental
evidence that you can't complicate them too much; see the huge body of
work on artificial membrane ion channels), you can raise the number by
three or four orders of magnitude. It will still be within reach of
few millenia of bacterial evolution, easily.

And once you have a channel, everything else is specificity. You have
the founding block of other channels to come.

But I took that example into digression. The point was the chance of a
particular structure arising. One of many things you constantly ignore
in your calculations is that there is a very limited number of actual
structures that separate domains fold into. Those things are
proscribed thermodynamically and sterically, and will be the same in
all proteins, regardless of the size. This is why you have domains,
subdomains, and motifs: things that get repeated over and over again.

To perform a function you don't need specific sequences of amino
acids. You need specific charges kept in specific spatial order, by
whichever combination of these larger structural motifs will do the
job.

The repetitiveness of these motifs is crucial. Generate a million
random large (say ~1000 residue) proteins, and you will have thousands
of them that contain structures similar enough to those of the large
proteins already in existence to mimic them to some degree. At least a
few hundred will be quite efficient at what they do.

Side note: it occurs to me that you might be saying that large
proteins, for some reason, are less likely to be beneficial. A large
protein might have five or six domains. Vast majority of random
mutations are neutral. So, even if development was completely random
(mix several thousand base pairs and run them through a ribosome),
chances are that you would get one or two domains that would do
something; chances for those domains to do something beneficial are
the same as the chances for small proteins.

>(i.e., bacterial motility),

Yes, motility. In looking through the references, I found out that it
has been established for quite some time that only three of the
proteins you listed are actually involved in torque generation (FliG,
MotA and MotB). Significant progress has been made in understanding
the exact manner of motility, and it seems that you will soon lose yet
another "they don't fully understand how it happened, so it must have
been designed" argument. (See the work of Hill and especially Blair on
the subject)

>Interesting. Give me the details of this demonstration
>and/or the reference for the publication of this demonstration.

Unpublished yet. There should be enough reading material to make up
for it above.

>Does it involve the deregulation of the kinase function, as
>is the case with the chimeric bcr-able function mentioned above?

It has nothing to do with kinase function (you insistence on this as
an important point is just another measure of how poorly you
understand the subject you are attacking, but this post has gotten far
too long for me to start yet another line of determined ignorance
correction).

>As I have said over and over again in this forum, I do not
>think we know enough about the functional genetic differences
>between humans and apes to rule out the possibility of common
>origin with the use of genetics alone.

We know that there are significant differences in many complex protein
structures between humans and chimps.

>The definition of "species" is rather subjective. I do
>not believe that all of what are now referred to as "separate
>species" where designed separately - although I do believe
>that they were all here relatively recently.

I went to your site, read your texts on geology...and I won't say
anything further; if I do, I will have to get involved in a separate
discussion on geology, which is not my subject.

However: they were all here relatively recently? Do explain in a bit
more detail. I would like you to tell me what do you actually think
happened, and how is it that we are all getting it all wrong (even
when such clear, simple calculations as yours disprove all of our
work).

>When such sequences are beneficial (which not all of
>the possibilities are), they are useful only because
>the system of the organism "recognizes" them as useful
>in a particular environment.

Which is why you have divergence, yes.

>Then, adding letters to that word, evolve it were each
>addition makes beneficial sense in that situation.

Can you understand this sentence?
Can you understand this sentencence?
Can you understand this sentencenoe?
Can you understand this sentence noe?
Can you understand this sentence now?

Pronounce the last sentence before you answer. And keep in mind that
words are far less flexible (orders and orders of magnitude) then
proteins. The second and third mutations are neutral, since the
information is still kept. The fourth is already beneficial, since
(when pronounced "noe" sounds similar enough to "now" to occasionaly
be understood).

>But they are very similar in that they
>are not completely flexible in their
>functional sequencing. There is a
>minimum limit to the size and order required
>by a protein-based function - beyond which
>all beneficial function suddenly ceases.

Incorrect. See the above examples, or do a single fricking protein
alignment and see how far same functionalities from different species
actually match. (Clue: not very far at all.)

>A many types of protein functions do not necessarily
>require multiple domains in order to be realized.
>You yourself admitted as much in your first post when
>you claimed that the lactase function is actually much
>simpler than its usual 1000 acids would make it appear.

Sigh. What I'm telling you is that every large protein has multiple
domains. Not that it requires multiple domains, but that it *has*
mutiple domains. Once you get above 200-300 residues in length, the
protein will fold in such a way that multiple, separable domains
*will* form. And even the minor structures will take recognizable,
repeatable shapes (take a look at the frequency of appearance of beta
bends, or omega bends).

Lactase has three domains. Simpler proteins can perform the function,
but they lose the specificity and efficiency.

>If you can present the evolution of such a collective
>function that requires, at minimum, all of the various
>amino acids in all of the various domains of a particular
>protein working at the same time, then you will have
>something.

? Can you present a single large protein that satisfies the above?

Goodness gracious, how convincingly you present yourself as a person
that understands biochemistry; I almost forget how many basic things
you don't understand.

Take your lactase. There are maybe six or seven residues that are
involved (working together), and maybe twenty or thirty more that are
kinda-sorta aiding the situation. The rest is not "working together".
It is same for any protein you can name.

But wait a minute: you *said* that lactase is a protein in which
several hundred residues work together, right? Well, if that is
"working together", every protein ever evolved has all of its residues
working together. So I have given you dozens of examples.

>Yes - I agree and know full well about such collective
>functions. I think you have misinterpreted what I was
>trying to say.

By your response (you are quite obviously flapping your hands and
trying to cover up your ignorance, not realizing that you are showing
off even more of it in the process), I wouldn't say so.

Do you, or do you not, realize that all large proteins consist of
separate domains? That within those domains exist secondary structures
and motifs that repeat again and again? Do you understand how limited
this is on a large scale (as specific as it can get on a small scale)?

Perhaps additional reading *is* in order. Since you are fond of
calculations and solution spaces, take a look at this:

Deeds, Eric J.; Dokholyan, Nikolay V.; Shakhnovich, Eugene I. Protein
evolution within a structural space. Biophysical Journal (2003)

or this:

Bairoch, Amos; Murzin, Alexey G. Sequences and topology. Predicting
evolution. Current Opinion in Structural Biology (1997)

...to see what a more reasonable calculation might look like.

Some more information on domains and folding:

Heringa J; Taylor W R Three-dimensional domain duplication, swapping
and stealing. Current Opinion in Structural Biology (1997)

Then, just a few teeny-tiny examples of structural evolution work:

Robinson, Douglas M.; Jones, David T.; Kishino, Hirohisa; Goldman,
Nick; Thorne, Jeffrey L. Protein evolution with dependence among
codons due to tertiary structure. Molecular Biology and Evolution
(2003)
Schulz, Angela; Schoeneberg, Torsten. The structural evolution of a
P2Y-like G-protein-coupled receptor. Journal of Biological Chemistry
(2003)
Unligil, Ulug Mete. Protein structure and evolution: a study of the
x-ray crystal structures of rabbit n-acetylglucosaminyltransferase i
and haemophilus influenzae periplasmic zinc-binding protein 1. (2002)
Apic, Gordana; Huber, Wolfgang; Teichmann, Sarah A. Multi-domain
protein families and domain pairs: comparison with known structures
and a random model of domain recombination. Journal of Structural
and Functional Genomics (2003)
Schmidt, Dawn M. Z.; Mundorff, Emily C.; Dojka, Michael; Bermudez,
Ericka; Ness, Jon E.; Govindarajan, Sridhar; Babbitt, Patricia C.;
Minshull, Jeremy; Gerlt, John A. Evolutionary potential of
(b/a)8-barrels: Functional promiscuity produced by single
substitutions in the enolase superfamily. Biochemistry (2003)
Shakhnovich, Boris E.; Harvey, John M.; Comeau, Steve; Lorenz, David;
DeLisi, Charles; Shakhnovich, Eugene. ELISA: structure-function
inferences based on statistically significant and evolutionarily
inspired observations. BMC Bioinformatics (2003)
Shirai, Tsuyoshi; Shionyu-Mitsuyama, Clara; Ogawa, Tomohisa. Changes
of protein structure and function on accelerated evolution: a case of
fish galectin. Tanpakushitsu Kakusan Koso (2003)

>Where? You made a blanket statement but you provided
>no details or references to support this statement.

In practically every message, starting from the first one. In the
message you are responeding to, where I gave you the example of RNA
Polymerase complex.

>However, I highly doubt that your job invokes the use of
>evolution beyond the lowest levels of functional complexity.

Actually, my work has a significant molecular evolution side note to
it. So you may lay your doubts to rest.

>What I am talking about here are those types of beneficial
>functions that require, at minimum, several thousand amino
>acids working together at the same time in a rather specified
>order.

You are talking about things that do not exist in nature?

>References?

Above.

>"Both the genetic and chemical inhibition of fibril-matrix
>construction inhibited the evolved swarming phenotypes."

Ah, indeed. Development of novel complex *functions* isn't what you
are after. You are after development of novel large proteins. Because
you claim, without any evidence or reasoning to suport it, that large
proteins are more unlikely to have beneficial mutations. Sorry, my
mistake.

>Perhaps you don't like to list the actual references
>to research studies like this because you know how very
>limited they are and that they really do not support
>your contentions and most exaggerated statements in
>the least.

Perhaps you will one day stop ignoring evidence and insisting on
erroneous mathematical constructions?

M.

[Chris Merli]

Very nice post.

<sweetnes...@yahoo.com> wrote in message
news:4d71d185.03121...@posting.google.com...

[Harlequin]

"Chris Merli" <clm...@insightbb.com> wrote in
news:ptqEb.424677$275.1301005@attbi_s53:

> Very nice post.

Seconded.

And if it wins, links to free full text copies or to
PubMed entries where appropriate should be made in
the Archive.

--
Anti-spam: replace "usenet" with "harlequin2"

Creationist arguments are like orcs. They are wimpy and
easy to slay. But there just so many of them....

[Rodjk]

"Chris Merli" <clm...@insightbb.com> wrote in message news:<ptqEb.424677$275.1301005@attbi_s53>...
> Very nice post.

Is this a nomination? If so, I second it...

Rodjk #613

>
> <sweetnes...@yahoo.com> wrote in message
> news:4d71d185.03121...@posting.google.com...
> > Sorry for the delay, much more important things got in the way. And
> > apologies for starting a new thread; apologies to everyone else, too,
> > for starting so many threads involving rebutalls of this half-assed
> > theory. But let us continue.

<Snip?

[Bill Rogers]

sweetnes...@yahoo.com wrote in message <snip excellent post>

No need to apologize. I really enjoy your posts.

In addition to the many, many references you already provided I saw a
nice review of the evolution of enzymes, taking advantage of all the
genomic and structural data that's come out over the last few years

Gerlt, JA and Babbit, PC (2001) DIVERGENT EVOLUTION OF ENZYMATIC
FUNCTION: Mechanistically Diverse Superfamilies and Functionally
Distinct SuprafamiliesAnn Rev Biochem 70:209-246

Bill

BTW does sweetness and light come from your having worked on optical
rotation spectroscopy of carbohydrates?

[RobinGoodfellow]

Terrific post! You've actually motivated me to read the Shakhnovich
paper (Protein Evolution within a Structural Space), which I've been
putting off for months now. Thanks!

[Tracy Hamilton]

"Bill Rogers" <bro...@noguchi.mimcom.net> wrote in message
news:8984713a.03121...@posting.google.com...

It could be that he is just another lowly University of Ediacara janitor,
retreiving
discarded copies of journals from the rubbish bin (poor Dr. Silence's
disadvantage
is in the contents of his rubbish bin at the Subway). Sweetness and light
must refer to
the fact that he changes light bulbs while eating left over Krispy Kremes
from the Biochemistry Department coffee counter.

Tracy P. Hamilton
Building Manager, Alco Hall

[Sean Pitman]

sweetnes...@yahoo.com wrote in message news:<4d71d185.03121...@posting.google.com>...

> Sorry for the delay, much more important things got in the way.

Glad to have you back. Hope everything else worked out well . . .

> And
> apologies for starting a new thread; apologies to everyone else, too,
> for starting so many threads involving rebutalls of this half-assed
> theory. But let us continue.
>
> I will begin by exploring yet another catastrophic error in your
> calculation:

Ok . . .

> >You have shown that short amino acid sequences can
> >come together to form a new unified function that
> >is indeed unique and of greater complexity. The
> >only problem you have here is that the minimum amino
> >acid number required for your most complex example
> >is only 3 or 4 hundred amino acids. My argument is
> >that evolution becomes more and more difficult the
> >greater the minimum amino acid requirement until it
> >becomes impossible this side of zillions of years
> >when the minimum requirement reaches a few thousand
> >amino acids in fairly specified order.
>
> I have given many reasons why this is wrong already, but let us go on
> to some new ones, until the state of affairs becomes painfully
> obvious.

Hopefully you will come up with something good this time.

> Let's start with a simple comparison. Let's take an imaginary protein,
> call it A, size 300. It acquires a point mutation, becoming B, still
> size 300. According to you, there is no problem in this happening,
> since the number of amino acids is low.

Actually, there might be a problem depending upon how specified in
sequencing "Function B" is required to be.

> Now, the same protein A acquires a large insert, and its size
> increases to, say, 2000. This is now protein C. According to you,
> there is no way whatsoever that this new protein could have a
> beneficial effect on the cell. I am mystified. Why?

I didn't say this at all. What I said was that if the minimum part
requirement needed to gain the type of function of protein C was a few
thousand fairly specified amino acids, then going from A to C would
indeed require trillions upon trillions of years on average. Of
course, now you are going to try and come back with an example of a
smaller protein attaching to a larger protein to yield a "new" type of
function that involves a few thousand amino acids. Hopefully you do
better than the bcr-abl chimeric protein that a few others have
already tried to float (which is neither beneficial nor does it have a
new type of function - only an up-regulation of abl's previous
tyrosine kinase function).

>
> It is because the strawman you are presenting has large proteins being
> built up by completely random association of amino acids. This, as I
> have pointed out again and again, does not happen at all.

This is not what I'm saying. You can build up your larger specified
proteins with new types of beneficial functions however you want. No
matter how you try to do it, it will not work beyond the lowest levels
of functional complexity.

> Let's say (erring on your side of the argument) that one mutation in

> ten billion produces a significantly beneficial new function.

Now you're talkin . . .

> So, for
> every ten billion point mutations in small proteins, there will be one
> that will produce a new small protein, with a new, beneficial
> function.

Ok, I'm good with this so far . . .

> Also, for every ten billion large insertions, there will be
> one that will produce a new LARGE protein, with a new, beneficial
> function.

But what is the minimum amino acid requirement for this new beneficial
function? Did the new beneficial function really require all of the
amino acids from the large protein or just a small stretch of them?
In other words, you must ask, "What is the shortest amino acid
sequence that would yield this particular type of function working in
this particular way?" For example, say a small protein would gain a
beneficial function if it hand just three particular amino acids in a
particular order. No matter how it gains this amino acids, it will
gain the new type of beneficial function. It doesn't matter if a
short protein only three residues comes along and get put in just the
right place or a large protein comes along and puts the three residues
in just the right place. The same minimum part requirement is still
just 3 additional specified amino acids.

> So even ignoring myriad other factors that go against your theory (I
> will add some more in this post too), just sticking to the most
> general, basic facts of genetic mutation, your entire spiel on how it
> is "more difficult" to produce a large protein then a small one fails
> completely.

Not if you consider the minimum part requirement, not the maximum junk
you can include with this minimum. The extra junk residues that are
not required for minimum function do not count toward the calculation
of minimum functional complexity.

> Let's carry this further. Because of domains, their recombination, and
> some thermodynamic factors (which I will cover below, in answer to
> another particular whopper of yours), larger proteins are far more
> versatile then the small ones - it is *easier* to adapt a large
> proteins for a completely novel function then a small one; you simply
> have more space to work with. This can be seen by wide varieties of
> independently produced structural proteins, membrane proteins,
> signaling complexes, etc; all of which depend on certain protein
> architecture rather then your "amino acids in specific order".

Hmmmm, but isn't architecture dependent upon amino acid order? The
greater the specificity of the required architecture, the greater the
specificity of amino acid order - right?

>The
> truly difficult feats are coming up with small but very
> sequence-specific proteins, such as, say, proteins that facilitate
> glucose transport across membranes.

These are difficult too, but some larger proteins are also fairly
specified. And, you forget, that I'm not just talking about larger
single proteins which often loose specificity requirements, but about
systems that require multiple smaller proteins working together at the
same time (as in various bacterial motility systems like the flagellar
system). Such multiprotein systems require many fairly specified
proteins working together at the same time. This means that several
thousand fairly specified amino acids are working together at the same
time. Such multiprotein functions are quite specified. They are not
as versatile as some of the less specified larger proteins that you
have tried to float as examples against my position. Such specificity
greatly reduces the density of beneficial sequences at such levels of
complexity in sequence space.

> >Where are your examples of evolution requiring such a
> >level of minimum amino acid specificity? I see a lot
> >of hot air coming from you, but no such example.
>
> There is no such thing as minimum amino acid specificity. You invented
> it. I cannot show you something that exists only in your imagination.

LOL - oh really? What then, is the shortest cytochrome c sequence
that would have the cytochrome c type of function? How many different
cytochromes are there at this minimum amino acid requirement?
Likewise, what is the shortest lactase enzyme that will actually work
in a beneficial way in a living creature? You have gone off about how
only a handful of amino acids are really important for such functions
and how the rest of the amino acids are not nearly as specified.
However, there definitely is a degree of specificity even for those
amino acids that are not fully constrained.

Take cytochrome c for example. The shortest functional cytochrome c
sequence that I know of requires a minimum of around 80 or so amino
acids. Of these about 20-30 are pretty much "invariant". Many of the
others (around 25) are highly constrained to within just 2 or 3 amino
acids having similar chemical natures. For example, Lys can change to
Arg, Glu can change to Asp, Ala can change to Val, and Val can change
to Ile etc. Surface AA residues can vary more than those in the
interior. Those in key positions involved in the functionality of the
protein are more constrained and can not vary without disturbing the
3-D shape of the protein and therefore its functionality. Already we
are up to around 40 to 60 highly specified amino acids. Many of the
others can vary only between certain classes of amino acids
(hydrophobic, hydrophilic, acidic, basic, etc). Only a very small
handful of cytochrome c amino acid positions can vary extensively
between all 20 amino acid possibilities.

Now, I'm sure you will bring up several studies that show that
individual amino acid positions, taken by themselves can be changed
and sometimes more significantly than occurs in nature, without a
complete loss of function. This is certainly true just like I can
change any one letter in this paragraph without a complete loss of
function. However, you must admit that not too many letters can be
changed very much at all at the same time without a complete loss of
beneficial function. The constraints on the cytochrome c type of
function are obviously significant. But what does this mean in
regards to my position?

The higher the constraints and the greater the minimum amino acid
requirement, the less the density of that type of function in sequence
space. For the cytochrome c function, in particular, some have
suggested that the total number of functional cytochrome c sequences
in sequence space is around 10e60. The problem here is that 10e60 is
only a minute fraction of the total sequence space of 10e104 at this
level of specified complexity. At such a level of specified
functional complexity the density of all such functions is low enough
that significant neutral gaps, in the form of non-beneficial potential
sequences, start forming that slow down the speed of evolution
significantly. At this level evolution is still possible, but it is a
lot slower than at lower levels of specified complexity. The problem
is that at higher levels of specified complexity, requiring more
fairly specified amino acids at minimum, the abilities of evolution to
come across new types of functions becomes exponentially reduced until
evolution complete stalls out at the level of just a few thousand
fairly specified amino acids working in concert.

> To point out just one piece of evidence you completely ignored, I have
> first asked you to look up some sequence alignments, and then (after
> you ignored my suggestion) I have provided you with a link to one.
> There you have several proteins that perform the same function, but
> agree with each other, sequence-wise, only ~20%. There are few
> exceptions to this rule, and no exceptions that I am aware of among
> large proteins (the larger the protein, less % agreement).

I did not ignore this statement at all. I fully agree with it, expect
it, and have discussed this argument multiple times. Take the lactase
function, for example. I'm sure that there are literally trillions
upon trillions of different potential lactase sequences in sequence
space (many having low sequence identity). For argument's sake let's
say that the shortest functional lactase enzyme requires at least
400aa to start working. Now, there most certainly would be a fair
degree of flexibility to such a function as a lactase enzyme. Not too
much specificity is required. This may result in there being at least
10e400 potentially beneficial lactase enzymes in this sequence space
of 400aa. That sounds like an absolute enormous number, and it is.
Until, until you compare it to the minimum sequence space at this
level of complexity which is 10e520. Now 10e400 doesn't look so
impressive anymore - does it? Finding the various islands of
clustered lactase sequences might be a bit of a challenge for a
particular colony of randomly mutating genomes even though there are
is a large absolute number of very different islands with the lactase
function.

> If a protein needs to have a "minimum" number of amino acids in a
> "specific sequence", how come that so many different-sized proteins
> (sometimes tenfold size difference for same function, same efficiency)
> with wildly diverging sequences can produce the same effect?

Easy. Look around you at all the different types of cars, airplanes,
boats, cameras, etc. Each of these different types of structures may
have certain common functions such as flying, floating, rolling,
picture taking etc. However, these functions, taken by themselves, do
indeed require a minimum number and specified orientation of parts.
The same is true for protein functions. Each type of protein function
requires both a minimum number of amino acids as well as a minimum
specified orientation of these amino acids before that type of
function can be realized. Just because there are many ways to get a
lactase function does not mean that there is not a minimum part
requirement for all of these ways. No matter which way you choose to
create a functionally beneficial lactase enzyme of cytochrome c
electron transport protein, you will not be able to go below a certain
minimum amino acid level. That is a cold hard fact. I really don't
see how you can get around it although I do sympathize with your
efforts to try and think of a way of talking yourself out of this most
interesting little predicament.

> If you
> want a correct definition of what is required, you have to say that it
> is necessary to have certain small stretches of specific amino acids
> occuring somewhere on the protein. Which screws up your calculation,
> and allows for directed evolution to be efficient (see references
> below).

Not at all. You need to have some idea of all the possibilities in
sequence space. You need to have an idea of the minimum part
requirement among all of these possibilities and calculate a density
of beneficial sequences taking all of these possibilities into
consideration before you can have any sort of idea about the abilities
of evolution to evolve or not evolve at such a level of functional
complexity.

> >This is good thinking since it is actually an attempt
> >to controvert my hypothesis whereas you haven't even
> >tried to present the de novo evolution of anything with
> >a minimum function requiring over a few hundred amino
> >acids working together at the same time.
>
> I did, among other things, a search on your previous posts. As far as
> you are concerned, if something is observed in lab, it is not
> sufficiently complex; if it hasn't been observed in the lab, it is
> obviously too complex to arise spontaneously. The approach does not
> suprise me, but let's see if something can be done about your
> credibility.
>
> If you paid any attention to what I said, you would have found that
> many of the proteins I spoke about (including those in pterine
> pathway) require many large proteins to function together. It is just
> that the one which provided a clue was 300 residues long. The next one
> is 800. You could have also noted that novel proteins of sizes
> measured in thousands of residues arise in labs all the time, only
> slightly less frequently then the smaller ones; usually, they evolve
> down in size, since large protein sizes are inefficient (contrary to
> your claim that they are *necessary*; again, see references below).

Exactly true. Many large proteins are not very specified and have
many redundant structural sequences. Again, my argument concerns the
*minimum* amino acid requirement needed to gain a particular type of
function. Certainly you can understand how this is much different
from the finding of the *maximum* number of amino acids that can
perform a particular type of function. I'm sure that you could make a
protein with 10,000 amino acids that would have the lactase function
or the cytochrome c function, but this does not change the fact that
the minimum part requirement needed to achieve such functions is much
much less - as is the level of complexity.

> >What you have shown are simpler systems combining to
> >produce a novel function that requires, at minimum,
> >only 3 or 4 hundred amino acids. You have also provided
> >examples of a simple function requiring less than a few
> >hundred amino acids at minimum evolving a new type of
> >function within the same level of complexity, but not anything
> >much greater. Please, I think you can at least try and
> >do better than this.
>
> Sure. As illustrated above, there is no correlation between size and
> effects of a mutation. Since you repeat the same argument many times
> in your posts, I will repeat mine a few times, in different forms, to
> attempt to minimize the amount of smoke you will try to cover them
> with.
>
> Your model is this: there is a function X, that requires at minimum a
> 10000-residue protein. The evolution works like this: it creates a
> random 10000 residue protein. Then it modifies one amino acid at a
> time, until it produces the protein that performs function X. You say
> that this is impossible, since it would take impossibly long to search
> through all the combinations of amino acids.
>
> As someone has nicely put it before, well, duh.

You and others say this because you misstate my position over and over
again, building a straw man that has no resemblance to my position at
all. I never said that evolution needs to create any sort of "random"
non-functional sequence. You can start with whatever you want. For
example, start with a function already in place that requires 10,000
fairly specified amino acids (such as in a multiprotein system where
all the proteins work together at the same time to create a unified
function). Then, starting with such a functional system, evolve a new
type of function at the same level of complexity or greater using the
mindless mechanisms of evolution (genetic duplication, random point
mutations, insertions etc. combined with natural selection).

> The evolution actually works a bit differently. I will illustrate this
> in two steps. For the first step, let's say that you are right, and
> that there is an absolute minimum for a size of a protein before it
> can perform function X; so 10000 residues or nothing. In this case, it
> is hit and miss. The population will relatively frequently produce
> 10000-residue proteins; occasionaly, some of them will have a
> beneficial function, and will be conserved.

This is where you go wrong. A 10,000 residue protein produced at
random may indeed come across a new function, but NOT a new type of
function that requires 10,000 amino acids AT MINIMUM. A small portion
of the 10,000aa may be in just the right order to produce a function
that only requires a small or less specified sequence of amino acids.
The rest of the 10,000 amino acids really aren't needed for such a
function - they just happen to be there. So, just because the protein
or protein system has 10,000 amino acids does not mean that they are
all required. You haven't done what I'm suggesting is impossible
here. You have started with a high level of complexity that actually
requires 10,000 fairly specified amino acids working at the same time
and you have evolved a new function that requires far fewer amino
acids working together at the same time. You do understand the
difference here - right?

> If one that is close
> enough to peform function X arises by accident, then the cell will be
> able to perform function X. If not, cell will not be able to perform
> function X, and will have to live without it.

Not just "function X", but anything within the level of complexity of
function X. The colony of life forms will not come across anything,
any type of function, within the level of complexity that function X
is a part of. They may come across functions that require a far lower
level of specified amino acid complexity, but they will not come
across any type of function requiring, at minimum, 10,000 fairly
specified amino acids working in concert.

> There are some proteins that fit this description.

Obviously there are some proteins that fit your description, but this
is not a description that counters my position. You must understand
this, so I'm not sure how you can see yourself clear to misstate this
concept you badly.

> For examples, you
> can compare the cell structures of eukariotes with those of
> prokaryotes, or animal cells with plant cells, or diatoms with
> amoebae... Occasionaly, a group will stumble upon something great, and
> will develop it further. Other groups have to do without it. See
> "nested hierarchy" for examples too.

Yeah yeah . . . The nested hierarchy argument. This is so common and
predictable among you evolutionists. When you can't demonstrate the
actual evolution of novel functions beyond the lowest levels of
complexity, you revert to the nested hierarchy argument. This is
nothing more than an argument that assumes that shared similarities
must mean common evolutionary ancestry. The problem with this
argument is that shared similarities do not support the notion of
common descent from a single common evolutionary ancestor over the
notion of common design. Shared similarities are explained equally
well by both positions. In order to rule out my position of
intelligent design, you must show that evolutionary mechanisms can
explain the differences as well as the similarities found within and
between various life forms. The similarities can be equally explained
by both of our respective positions. The differences, on the other
hand, are a different story.

> One thing that is absolutely certain even here is that you *will*
> eventually have *many* very large proteins performing many different
> functions within the cell. You might not have one that performs
> functon X. Luck of the draw.

This is not "absolutely certain" above the lowest levels of minimum
specified complexity. You have made a bold assertion that this is
true, but you have yet to back yourself up with anything other than
assumed historical correlatory comparisons between similar functional
sequences and systems in various living things. You have no real-time
example of evolution in action now do you? Just admit it and move on.
If all you have to rest your faith on is nested hierarchies, wouldn't
it be the most honest thing just to say it like that?

> Second step of my illustration is a bit futher away from your strawman
> and closer to evolutionary genetics. In real world, there is no such
> thing as minimum residue requirement; small polypeptides can perform
> analogous functions quite well (small polupeptides that perform the
> galactosidase function - which should be impossible if your theory is
> correct - have been shown to perform quite efficient catalytic
> function).

Oh really? Where is your real life example where such efficient
catalytic functions have been tried in a living creature and what is
the minimum amino acid requirement for the smallest successful
demonstration? I have asked you this before but you failed to give me
the relevant reference to back up this statement of yours.

> Those small polypeptides get domains added to them, which
> then provide specificity, signaling, and all other functions a large
> protein has. This is, again, *observed to happen* in lab, especially
> during directed evolution experiments (which have the benefit of
> speeding up things to an observable rate).

Your problem is that laboratory experiments such as this which do not
involve any living thing actually using such evolutionary progressions
in some sort of beneficial way do not support your position that they
would work in real life. Each step has to be beneficial to a living
organism in its own environment - not the imagination of what is
beneficial to a laboratory scientist.

<snip>

> >However, with each amino acid increase in the minimum
> >random walk required to achieve the success of any new
> >type of beneficial function within that level of
> >functional complexity, the time required grows exponentially.
>
> If you are going to a specific place. Let us go through yet another
> example.
>
> Imagine the cell as a man who walks around in a room. Each minute, he
> steps in a random direction. Room is, say, one mile by one mile in
> radius. In the room are spread, say, ten thousand functions that 50
> residue proteins can possibly have (imagine them as small dark spots
> where the function is optimal, surrounded with vast grey areas where
> function somewhat works, but not very well).
>
> It is obvious that the cell will frequently step on various grey
> areas, and produce various 50 residue proteins, right?

You forget to think about the black areas in the room that are not
beneficial in the least. How large are the black areas? Sure, once
you hit a gray area, all bets are off. The question is, what is the
relative size of the black areas in the room?

> Now, let's increase the size of the protein to 5000 residues. Now the
> room is the size of a continent. The man continues to walk randomly.
> Your argument is: there is no way that, randomly walking, the man
> could cross from this particular spot in LA to that particular spot in
> NY without taking trillions of years. And you are right.

No. That is not my argument. My argument is that there is no way
that a many could walk randomly from a particular spot and its
surrounding gray area to any other gray areas in the entire room -
which would actually be the size of zillions of universes, not a small
continent. The "sequence space" of a 5000aa protein is over 10e6505
sequences. If the total number of beneficial sequences (to include
all the gray area sequences) at this level of fairly specified
complexity were as great as 10e6000, finding one of them, any one of
them, would be like finding a single atom in zillions of universes the
size of ours.

Are you starting to see the problem now?

> But that has no effect. You see, the functionalities of proteins are
> as dense on a continental scale as on the room scale. If 50 residue
> proteins can have ten thousand different functions, 5000 residue
> proteins can have trillions of different functions. These are spread
> around, covering the whole continent.

You make the mistake of thinking that small rooms with simple
functions can be added up to cover the entire continent and even
universes upon universes. You are basically multiplying your little
rooms than then thinking that the continent has made up of many of
these little rooms. This is simply not the case. The continent space
is not simply a collection of little rooms at all. It is a collection
of ONLY those functions that require a content sized protein at
minimum. A multi-universe sized protein, or protein system, such as
one that requires, at minimum, 5000 fairly specified amino acids, is
in select company. The lesser functions (i.e., smaller rooms) that
could be made with these 5000aa are not part of those functions that
require at least 5000 fairly specified amino acids to be realized.
You cannot simply add together smaller functions in a stepwise manner
to gain the 5000 level either just like you cannot add shorter words
together to make a longer sentence or paragraph where each addition is
selectably beneficial. The odds that your additions will be
meaningful are the same odds that you will hit a gray area at the
higher level of minimum specified sequence complexity.

> So the men walk, multiply, occasionaly fall down cliffs and into
> ditches...one of them ends up in Las Vegas, one of them ends up in
> Miami, one of them ends up in New York. Then I come around and call
> one "plant", the other "bacteria", the third one "animal"...

Yes, if it were all this easy and if the beneficial sequences were as
closely spaced as you seem to think they are. However, I am
suggesting to you that the dots and gray circles are not nearly as
close together as you seem to think. They are universes apart beyond
the lowest levels of functional complexity. That is why you are
having such difficulties showing me a real life example of evolution
in action that does very much beyond the lowest levels of functional
complexity. There simply are no such examples.

> Then you come along, and say "there is no way this man could have ever
> reached New York by walking randomly!"

Again, this is not at all what I said. I said that there is no way
that this man could reach any beneficial city, much less New York, in
a reasonable amount of time since all the beneficial cities average
zillions of miles apart. I'm not talking about a single teleological
goal here, but absolutely any beneficial point in sequence space that
this man might happen upon via random walk.

> Of course not. But he wasn't
> going there in the first place. He was waking randomly. He just
> happened to be there when you looked.

Not at all. You have not shown how any beneficial sequence, not just
a teleological sequence, can be reached within any level of complexity
that goes very far beyond those that require just a few hundred fairly
specified amino acids working together at the same time.

> And, the buildup of complexity is a factor also. Take five people.
> Land them on a new, Earth-like planet. Tell them to build a computer.
> They will fail: they do not have the tools, the expertise or the
> infrastructure. It takes thousands of years to build the
> infrastructure necessary for someone to produce a computer; but this
> does not mean that production of a computer is a momentuous task when
> infrastructure is present. It is difficult, but not impossibly hard.

Not impossibly hard when intelligent minds are involved, but certainly
impossible when there are only mindless processes available. It is
very interesting that you invoke humans as part of the requirement for
the complexity of computer function to be realized. What about the
mindless processes of wind, water, sand, dirt, and heat acting
together mindlessly? How long would it take for such mindless
processes to build a computer or anything else of equal or greater
functional complexity as compared to the intelligent process of human
creativity?

> >Ok - take, for example, a particular function that
> >requires, at minimum 5,000aa at minimum to be realized.
> >Say this sequence happens to get duplicated so that it
> >can undergo various mutations without risking significant
> >loss to the original beneficial function (which has been
> >optimized for its host by now - as far as *level* of
> >function is concerned). The sequence space at this level
> >of complexity is more than 10e6500 sequences.
>
> Let's point out another error in this. Genetic code is degenerate;
> which means that the mutation that produces methionine at a certain
> place is far less likely (what, 64 times less likely?) to come up then
> one that produces glycine. Since mutations happen at genetic level,
> your numbers are utterly meaningless.

Actually, at the level of DNA the numbers would be worse, even taking
into account the degenerate nature of the genetic code. For example,
at minimum 5000aa would require a base pair sequence of 15,000. In a
codon the third base is fairly flexible. Many amino acids have
multiple codons (as many as 6 or 7 different codons) coding for them
via a wobble in this third base. Taking this redundancy into account,
the DNA sequence space in this case would be well over 10e6575 base
sequences. These are not meaningless numbers at all. Do the math
yourself. The problems are only worse for you. For example, what if a
particular function needs more methionine residues than it needs
glycine residues? The odds of gaining such functions are even less
likely. My calculations are only dealing with averages, but more
detailed calculations do not hurt my position in the least.

<big snip which I may have more time for later>

> >Then, adding letters to that word, evolve it were each
> >addition makes beneficial sense in that situation.
>
> Can you understand this sentence?
> Can you understand this sentencence?
> Can you understand this sentencenoe?
> Can you understand this sentence noe?
> Can you understand this sentence now?

Are you trying to tell me that each stepwise change here is supposed
to be more selectably beneficial as compared to the one that came
before? Maintaining function is not the same thing as each change
having a new type of beneficial function. As I see it each of your
changes here were actually functionally detrimental, though still
meaningful, until the last change where a new beneficial function may
have been evolved (depends on the situation/environment). Really, try
doing the same thing where each change is not only understandable but
also more functionally beneficial as compared to what came before.
That is the problem you know since natural selection can only select
in a preferential manner those sequences that are actually more
beneficial in function as compared to the other options.

> Pronounce the last sentence before you answer. And keep in mind that
> words are far less flexible (orders and orders of magnitude) then
> proteins.

This is true, but many protein-based functions are far more
constrained that you seem to understand. The language illustration is
a good way to get people to start to understand the problem. It is
easier to evolve longer sequences of functional proteins (up to a few
hundred), but relatively soon the specificity of higher levels of
protein functions prevents further evolution just like it does with
English sentence evolution where each step must be more beneficial
than the one that came before if it is to be selectable, in a positive
way, by natural selection.

> The second and third mutations are neutral, since the
> information is still kept. The fourth is already beneficial, since
> (when pronounced "noe" sounds similar enough to "now" to occasionaly
> be understood).

Ok . . . So you have a random walk that must cover two mutations that
are addition mutations (rarer than point mutations). The sequence
space involved is 729. That means that a random walk will cover over
700 steps/mutations on average before it comes up with "noe". Now
that little gap of just two steps requires several hundred random walk
steps on average to cross. That requires quite a bit of time, but is
still doable - right? Just keep going. Starting with this beginning
phrase, see how far you can go before you come to more and more
significant neutral gaps. See how many different types of ideas you
can evolve at this level of specified complexity where the meaningful
steps are so close together . . .

<snip>

> Perhaps you will one day stop ignoring evidence and insisting on
> erroneous mathematical constructions?

Perhaps someday you will actually understand the very real problem
that these very relevant mathematical constructs pose for your faith
in the very strange notion of evolutionary progression.

Owing to the length of your post, this is all I have time to respond
to today. You have repeated several errors and misrepresentations of
my position, which may actually be simple misunderstandings on your
part. I have tried to correct these misunderstandings. Perhaps, after
gaining a better understanding of my position, your future posts will
actually address my actual position and challenges to your position?

> M.

Sean
www.naturalselection.0catch.com

[howard hershey]

Sean Pitman wrote:

^^

*If* is a rather crucial word here, Sean. You seem to be quite certain
that there actually are proteins where the "minimum part requirement"
are "a few thousand fairly specified amino acids". What evidence do you
have to support that argument?

> requirement needed to gain the type of function of protein C was a few
> thousand fairly specified amino acids,

Can you describe how you determine the fraction of total amino acids in
a protein that correspond to "the minimum part requirement" and give
examples of proteins or protein systems where you have *demonstrated*
and have *evidence* for the statement that "the minimum part" is a few
thousand amino acids long? And what exactly do you mean by "fairly
specified"? Invariant amino acids? Amino acids that can be substituted
by a few others? Amino acids that only require that the substitute be
able to form a beta sheet or alpha helix or be hydrophobic or
hydrophilic? Amino acids where any other is fine *except* one or two
alternative amino acids (say proline).

> then going from A to C would
> indeed require trillions upon trillions of years on average. Of
> course, now you are going to try and come back with an example of a
> smaller protein attaching to a larger protein to yield a "new" type of
> function that involves a few thousand amino acids.

No. He already said that it was the single mutation that produced the
"new" type of function. It did so whether the protein was 300 or 2300
amino acids long. So, how do you go about determining that the "minimum
part" was only some unspecified fraction of the 300 amino acids than the
2300? What about the possibility that 300 is itself not the "minimum
part"? How do you go about determining the "minimum part"? To date, it
seems that "minimum part" seems to be "total amino acid length" except
when it is pointed out that it is nonsense.

> Hopefully you do
> better than the bcr-abl chimeric protein that a few others have
> already tried to float (which is neither beneficial nor does it have a
> new type of function - only an up-regulation of abl's previous
> tyrosine kinase function).
>
>
>
>>It is because the strawman you are presenting has large proteins being
>>built up by completely random association of amino acids. This, as I
>>have pointed out again and again, does not happen at all.
>
>
> This is not what I'm saying. You can build up your larger specified
> proteins with new types of beneficial functions however you want. No
> matter how you try to do it, it will not work beyond the lowest levels
> of functional complexity.

And how does one measure "functional complexity"? To date, it seems to
be nothing but the "total amino acid length" which is also the "minimum
part" (except when you say it isn't).

>
>
>>Let's say (erring on your side of the argument) that one mutation in
>>ten billion produces a significantly beneficial new function.
>
>
> Now you're talkin . . .
>
>
>>So, for
>>every ten billion point mutations in small proteins, there will be one
>>that will produce a new small protein, with a new, beneficial
>>function.
>
>
> Ok, I'm good with this so far . . .
>
>
>>Also, for every ten billion large insertions, there will be
>>one that will produce a new LARGE protein, with a new, beneficial
>>function.
>
>
> But what is the minimum amino acid requirement for this new beneficial
> function?

According to your mathematical argument, it is the total length in amino
acids.

> Did the new beneficial function really require all of the
> amino acids from the large protein or just a small stretch of them?
> In other words, you must ask, "What is the shortest amino acid
> sequence that would yield this particular type of function working in
> this particular way?" For example, say a small protein would gain a
> beneficial function if it hand just three particular amino acids in a
> particular order. No matter how it gains this amino acids, it will
> gain the new type of beneficial function. It doesn't matter if a
> short protein only three residues comes along and get put in just the
> right place or a large protein comes along and puts the three residues
> in just the right place. The same minimum part requirement is still
> just 3 additional specified amino acids.
>
>
>>So even ignoring myriad other factors that go against your theory (I
>>will add some more in this post too), just sticking to the most
>>general, basic facts of genetic mutation, your entire spiel on how it
>>is "more difficult" to produce a large protein then a small one fails
>>completely.
>
>
> Not if you consider the minimum part requirement, not the maximum junk
> you can include with this minimum. The extra junk residues that are
> not required for minimum function do not count toward the calculation
> of minimum functional complexity.

Given that active sites are quite small, where do you find starting
states where "minimum functional complexity" is thousands of amino acids?

>
>
>>Let's carry this further. Because of domains, their recombination, and
>>some thermodynamic factors (which I will cover below, in answer to
>>another particular whopper of yours), larger proteins are far more
>>versatile then the small ones - it is *easier* to adapt a large
>>proteins for a completely novel function then a small one; you simply
>>have more space to work with. This can be seen by wide varieties of
>>independently produced structural proteins, membrane proteins,
>>signaling complexes, etc; all of which depend on certain protein
>>architecture rather then your "amino acids in specific order".
>
>
> Hmmmm, but isn't architecture dependent upon amino acid order? The
> greater the specificity of the required architecture, the greater the
> specificity of amino acid order - right?

Nope. Most protein architecture is a matter of stretches of alpha
helixes, beta sheets and bends as well as regions of hydrophobicity and
hydrophilicity. None of those involve much sequence specificity wrt
amino acid order. There are many sequences which form those structures.

>>The
>>truly difficult feats are coming up with small but very
>>sequence-specific proteins, such as, say, proteins that facilitate
>>glucose transport across membranes.
>
>
> These are difficult too, but some larger proteins are also fairly
> specified.

Specific examples and references. It would also be nice to tell us how
you determined the "minimum parts".

> And, you forget, that I'm not just talking about larger
> single proteins which often loose specificity requirements, but about
> systems that require multiple smaller proteins working together at the
> same time (as in various bacterial motility systems like the flagellar
> system). Such multiprotein systems require many fairly specified
> proteins working together at the same time. This means that several
> thousand fairly specified amino acids are working together at the same
> time.

Can you tell us how you calculated this number of 'several thousand' for
the E. coli bacterial flagella system?

> Such multiprotein functions are quite specified. They are not
> as versatile as some of the less specified larger proteins that you
> have tried to float as examples against my position. Such specificity
> greatly reduces the density of beneficial sequences at such levels of
> complexity in sequence space.

Could you provide the evidence and references from which you generated
these numbers?

>>>Where are your examples of evolution requiring such a
>>>level of minimum amino acid specificity? I see a lot
>>>of hot air coming from you, but no such example.
>>
>>There is no such thing as minimum amino acid specificity. You invented
>>it. I cannot show you something that exists only in your imagination.
>
>
> LOL - oh really? What then, is the shortest cytochrome c sequence
> that would have the cytochrome c type of function?

What level of function? After all, in the land of the blind, the
one-eyed man is king.

> How many different
> cytochromes are there at this minimum amino acid requirement?
> Likewise, what is the shortest lactase enzyme that will actually work
> in a beneficial way in a living creature?

What makes you think that *length* is the relevant feature?

> You have gone off about how
> only a handful of amino acids are really important for such functions
> and how the rest of the amino acids are not nearly as specified.
> However, there definitely is a degree of specificity even for those
> amino acids that are not fully constrained.
>
> Take cytochrome c for example. The shortest functional cytochrome c
> sequence that I know of requires a minimum of around 80 or so amino
> acids. Of these about 20-30 are pretty much "invariant". Many of the
> others (around 25) are highly constrained to within just 2 or 3 amino
> acids having similar chemical natures. For example, Lys can change to
> Arg, Glu can change to Asp, Ala can change to Val, and Val can change
> to Ile etc. Surface AA residues can vary more than those in the
> interior. Those in key positions involved in the functionality of the
> protein are more constrained and can not vary without disturbing the
> 3-D shape of the protein and therefore its functionality. Already we
> are up to around 40 to 60 highly specified amino acids. Many of the
> others can vary only between certain classes of amino acids
> (hydrophobic, hydrophilic, acidic, basic, etc). Only a very small
> handful of cytochrome c amino acid positions can vary extensively
> between all 20 amino acid possibilities.

That is, there is a range of 'specificity' among the amino acids, not a
number of highly specified amino acids. That, to me, makes calculating
a "minimum parts" number a fool's errand. And makes any calculations
you make utter nonsense.

Moreover, heme does not need any protein to act as a receptor of an
electron. Cytochrome c acts primarily to regulate the exchange of an
electron from one heme's edge to another heme's edge.

And, of course, there is no evidence that cytochrome c arose by starting
with a random sequence of amino acids nor that the original function of
the system was that of being involved in a respiratory chain. Waste
removal is more likely. Cytochrome c, of course, is an ancient protein.
And the older the protein, the more evidence of ancestry gets lost.
The interesting thing is that the 3-D structure of cytochrome c is more
highly conserved than the sequence.

>
> Now, I'm sure you will bring up several studies that show that
> individual amino acid positions, taken by themselves can be changed
> and sometimes more significantly than occurs in nature, without a
> complete loss of function. This is certainly true just like I can
> change any one letter in this paragraph without a complete loss of
> function. However, you must admit that not too many letters can be
> changed very much at all at the same time without a complete loss of
> beneficial function. The constraints on the cytochrome c type of
> function are obviously significant. But what does this mean in
> regards to my position?
>
> The higher the constraints and the greater the minimum amino acid
> requirement, the less the density of that type of function in sequence
> space. For the cytochrome c function, in particular, some have
> suggested that the total number of functional cytochrome c sequences
> in sequence space is around 10e60. The problem here is that 10e60 is
> only a minute fraction of the total sequence space of 10e104 at this
> level of specified complexity.

This, again, assumes that evolution works by starting with a random
sequence and explores sequence space by a random walk.

That certainly is a good reason why common descent with most changes in
those amino acids which are not constrained is a better explanation for
the similarity of betagalactosidases in those related bacteria that have
that activity than is the independent creation of each variant from a
random walk from a random sequence or any other creationist mechanism.
Especially since there is undoubtedly a nested pattern of such enzymes
just like there is for cytochrome c.

> Now 10e400 doesn't look so
> impressive anymore - does it? Finding the various islands of
> clustered lactase sequences might be a bit of a challenge for a
> particular colony of randomly mutating genomes even though there are
> is a large absolute number of very different islands with the lactase
> function.
>
>
>>If a protein needs to have a "minimum" number of amino acids in a
>>"specific sequence", how come that so many different-sized proteins
>>(sometimes tenfold size difference for same function, same efficiency)
>>with wildly diverging sequences can produce the same effect?
>
>
> Easy. Look around you at all the different types of cars, airplanes,
> boats, cameras, etc. Each of these different types of structures may
> have certain common functions such as flying, floating, rolling,
> picture taking etc. However, these functions, taken by themselves, do
> indeed require a minimum number and specified orientation of parts.
> The same is true for protein functions. Each type of protein function
> requires both a minimum number of amino acids as well as a minimum
> specified orientation of these amino acids before that type of
> function can be realized. Just because there are many ways to get a
> lactase function does not mean that there is not a minimum part
> requirement for all of these ways.

Yet the number you use in your calculations is total amino acids, not
"minimum part". And given the variability in specificity, I wonder what
meaning "minimum part" has, if anything beyond its use as a meaningless
buzzword, like "low levels of functional complexity".

> No matter which way you choose to
> create a functionally beneficial lactase enzyme of cytochrome c
> electron transport protein, you will not be able to go below a certain
> minimum amino acid level. That is a cold hard fact. I really don't
> see how you can get around it although I do sympathize with your
> efforts to try and think of a way of talking yourself out of this most
> interesting little predicament.
>
>
>>If you
>>want a correct definition of what is required, you have to say that it
>>is necessary to have certain small stretches of specific amino acids
>>occuring somewhere on the protein. Which screws up your calculation,
>>and allows for directed evolution to be efficient (see references
>>below).
>
>
> Not at all. You need to have some idea of all the possibilities in
> sequence space.

Why?

> You need to have an idea of the minimum part
> requirement among all of these possibilities

How?

> and calculate a density
> of beneficial sequences taking all of these possibilities into
> consideration

Can you show us your calculations, how you determined each number, and
what assumptions you made wrt your calculations for any of the systems
you talk about? Or are your calculations, as I think, GIGO.

How do you go about determining this number?

What determines that a system is "new" and "at the same level of
complexity or greater"? How would I be able to distinguish that from a
function which is "new" but at a lower level of complexity? From a
function which is not 'new', but merely an expansion or variation of a
previous function? Would a protein that is chimeric, say one having a
transmembrane moiety attached to a different allosteric effector binding
site, be 'new' or 'old'? How would I determine the "minimum parts" of
such a change?

> using the
> mindless mechanisms of evolution (genetic duplication, random point
> mutations, insertions etc. combined with natural selection).
>
>
>>The evolution actually works a bit differently. I will illustrate this
>>in two steps. For the first step, let's say that you are right, and
>>that there is an absolute minimum for a size of a protein before it
>>can perform function X; so 10000 residues or nothing. In this case, it
>>is hit and miss. The population will relatively frequently produce
>>10000-residue proteins; occasionaly, some of them will have a
>>beneficial function, and will be conserved.
>
>
> This is where you go wrong. A 10,000 residue protein produced at
> random may indeed come across a new function, but NOT a new type of
> function that requires 10,000 amino acids AT MINIMUM. A small portion
> of the 10,000aa may be in just the right order to produce a function
> that only requires a small or less specified sequence of amino acids.
> The rest of the 10,000 amino acids really aren't needed for such a
> function - they just happen to be there. So, just because the protein
> or protein system has 10,000 amino acids does not mean that they are
> all required. You haven't done what I'm suggesting is impossible
> here. You have started with a high level of complexity that actually
> requires 10,000 fairly specified amino acids working at the same time
> and you have evolved a new function that requires far fewer amino
> acids working together at the same time. You do understand the
> difference here - right?

Yes. The only example you would accept as being evolution is one where
you start out with a random sequence of 10,000 amino acids and produce a
new function that involves changing all 10,000 amino acids. Is that
correct? Any new function that requires less than 10,000 amino acid
changes is not sufficiently complex.

>
>
>>If one that is close
>>enough to peform function X arises by accident, then the cell will be
>>able to perform function X. If not, cell will not be able to perform
>>function X, and will have to live without it.
>
>
> Not just "function X", but anything within the level of complexity of
> function X.

How do you determine the level of complexity of function X? The total
number of amino acids? The "minimum number" (so long as it is 10,000 or
so amino acids long)? Since most proteins are only 200-300 amino acids
long, there can't be too many multi-protein *systems* other than the
entire cell itself that has a "minimum number" of just the active sites
of 10,000 unless you are counting almost every amino acid.

> The colony of life forms will not come across anything,
> any type of function, within the level of complexity that function X
> is a part of. They may come across functions that require a far lower
> level of specified amino acid complexity, but they will not come
> across any type of function requiring, at minimum, 10,000 fairly
> specified amino acids working in concert.
>
>
>>There are some proteins that fit this description.
>
>
> Obviously there are some proteins that fit your description, but this
> is not a description that counters my position. You must understand
> this, so I'm not sure how you can see yourself clear to misstate this
> concept you badly.
>
>
>>For examples, you
>>can compare the cell structures of eukariotes with those of
>>prokaryotes, or animal cells with plant cells, or diatoms with
>>amoebae... Occasionaly, a group will stumble upon something great, and
>>will develop it further. Other groups have to do without it. See
>>"nested hierarchy" for examples too.
>
>
> Yeah yeah . . . The nested hierarchy argument. This is so common and
> predictable among you evolutionists.

Are you saying that the nested hierarchy doesn't exist in Pitman-world?

> When you can't demonstrate the
> actual evolution of novel functions beyond the lowest levels of
> complexity, you revert to the nested hierarchy argument. This is
> nothing more than an argument that assumes that shared similarities
> must mean common evolutionary ancestry. The problem with this
> argument is that shared similarities do not support the notion of
> common descent from a single common evolutionary ancestor over the
> notion of common design. Shared similarities are explained equally
> well by both positions.

Anything can be explained from the position of "Goddidit that way". You
need to explain why God chose this particular pattern of common design
among all the other possibilities.

> In order to rule out my position of
> intelligent design, you must show that evolutionary mechanisms can
> explain the differences as well as the similarities found within and
> between various life forms. The similarities can be equally explained
> by both of our respective positions. The differences, on the other
> hand, are a different story.

The nested hierarchy is an explanation of the pattern of differences.
Selection for function (and time since divergence) are explanations of
similarities.

>
>
>>One thing that is absolutely certain even here is that you *will*
>>eventually have *many* very large proteins performing many different
>>functions within the cell. You might not have one that performs
>>functon X. Luck of the draw.
>
>
> This is not "absolutely certain" above the lowest levels of minimum
> specified complexity. You have made a bold assertion that this is
> true, but you have yet to back yourself up with anything other than
> assumed historical correlatory comparisons between similar functional
> sequences and systems in various living things. You have no real-time
> example of evolution in action now do you? Just admit it and move on.
> If all you have to rest your faith on is nested hierarchies, wouldn't
> it be the most honest thing just to say it like that?
>
>
>>Second step of my illustration is a bit futher away from your strawman
>>and closer to evolutionary genetics. In real world, there is no such
>>thing as minimum residue requirement; small polypeptides can perform
>>analogous functions quite well (small polupeptides that perform the
>>galactosidase function - which should be impossible if your theory is
>>correct - have been shown to perform quite efficient catalytic
>>function).
>
>
> Oh really? Where is your real life example where such efficient

Who said anything about efficient? In the land of the blind...

And even you know that that number is irrelevant garbage, or you
wouldn't be making the distinction between total amino acids and minimum
parts.

> If the total number of beneficial sequences (to include
> all the gray area sequences) at this level of fairly specified
> complexity were as great as 10e6000, finding one of them, any one of
> them, would be like finding a single atom in zillions of universes the
> size of ours.
>
> Are you starting to see the problem now?

Yes. You keep using GIGO numbers that even you know are GIGO.

[Ian Musgrave & Peta O'Donohue]

G'Day All
Address altered to avoid spam, delete RemoveInsert

On Thu, 18 Dec 2003 22:43:09 +0000 (UTC), sweetnes...@yahoo.com
wrote:

>Sorry for the delay, much more important things got in the way. And
>apologies for starting a new thread; apologies to everyone else, too,
>for starting so many threads involving rebutalls of this half-assed
>theory. But let us continue.

[snip]

Bravo! POTM secondment.

I was going to use the alpha-beta barrel superfamily, with lots of
different function sone step from each other, as an example in a long
reply I am slowly writing to Dr. Pitman, its nice to see someonelse
picking up on them.

Mind you have I have already used the histidine synthetase coupled to
GFP as an example of your "big protein couple to ltt;e protein which
is one mutaion away from a new function regardless of size. Dr. Pitman
hasn't replied to that one yet.

Cheers! Ian
=====================================================
Ian Musgrave Peta O'Donohue,Jack Francis,Michael James and Andrew Thomas Musgrave
reynella@RemoveInsret_werple.mira.net.au http://home.mira.net/~reynella/
Southern Sky Watch http://www.abc.net.au/science/space/default.htm

[Sean Pitman]

"Chris Merli" <clm...@insightbb.com> wrote in message news:<ptqEb.424677$275.1301005@attbi_s53>...
>

> Very nice post.
>
> <sweetnes...@yahoo.com> wrote in message
> news:4d71d185.03121...@posting.google.com...

Oh, I think it would be just great if this post won POTM - no joke!

Thirded.

Sean
www.naturalselection.0catch.com

[Glenn]

"Ian Musgrave & Peta O'Donohue" <reynella_R...@werple.mira.net.au> wrote
in message news:un07uvkgga1ovoqik...@4ax.com...

> G'Day All
> Address altered to avoid spam, delete RemoveInsert
>
> On Thu, 18 Dec 2003 22:43:09 +0000 (UTC), sweetnes...@yahoo.com
> wrote:
>
> >Sorry for the delay, much more important things got in the way. And
> >apologies for starting a new thread; apologies to everyone else, too,
> >for starting so many threads involving rebutalls of this half-assed
> >theory. But let us continue.
>
> [snip]
>
> Bravo! POTM secondment.
>
> I was going to use the alpha-beta barrel superfamily, with lots of
> different function sone step from each other, as an example in a long
> reply I am slowly writing to Dr. Pitman, its nice to see someonelse
> picking up on them.
>
> Mind you have I have already used the histidine synthetase coupled to
> GFP as an example of your "big protein couple to ltt;e protein which
> is one mutaion away from a new function regardless of size. Dr. Pitman
> hasn't replied to that one yet.
>

"One amino acid at a time. Here's a thought experiment for you. Take
and enzyme of around 200 aas, say the histidine synthesizing protein
alpha/beta barrel protein HisF, this can be converted tyrosine
synthesizing protein TyrF by one aa mutation. Now add a 200 aa protein
such as Green Fluorescent Protein to the C-terminal end of HisF. THe
combined protein is 400 aa long, and the TryF function should be lost
in the welter of sequences, but you can still get there with one
single mutation."

What new function is realized, what does this new combined
protein do? Perhaps I am reading this wrong,
but it appears that you are left with GFP with added baggage.

[sweetnes...@yahoo.com]

Howard Hershey has already asked most of the relevant questions
(thanks!), so I will just give a few relevant comments. I would,
however, ask you to sit down and read the rest of my original message
before you respond to this.

> I didn't say this at all. What I said was that if the minimum part
> requirement needed to gain the type of function of protein C was a few
> thousand fairly specified amino acids, then going from A to C would
> indeed require trillions upon trillions of years on average.

What I am saying again and again is that there is no such thing as
minimum amino acid requirement. You have invented that out of thin
air. It is as simple as that. You are simply refusing to look at the
evidence, and repeating the same things over and over again.

> This is not what I'm saying. You can build up your larger specified
> proteins with new types of beneficial functions however you want. No
> matter how you try to do it, it will not work beyond the lowest levels
> of functional complexity.

I have provided you with examples and references to proteins doing
exactly that which you say is impossible. Large proteins, thousands of
residues long. There have been directed evolution experiments
involving RANDOM long proteins (random mixing of sheared DNA
fragments) that produced some fairly humongous functional enzymes.
Take a look at the reading list I provided for you.

> But what is the minimum amino acid requirement for this new beneficial
> function? Did the new beneficial function really require all of the
> amino acids from the large protein or just a small stretch of them?
> In other words, you must ask, "What is the shortest amino acid
> sequence that would yield this particular type of function working in
> this particular way?"

Take a look at the literature; there are snippets in the list I gave
you to read, but there is so much more if you bother to read a
fricking journal now and then.

Your question is meaningless, but no matter how many examples and
references I provide, you just keep repeating it over and over. Very
well, let us try a different approach. You choose a large protein -
any large protein over one thousand residues. Just pick one. Then YOU
tell ME how many of its amino acids are REQUIRED for its function, and
WHY you think that they are all REQUIRED.

> Not if you consider the minimum part requirement, not the maximum junk
> you can include with this minimum. The extra junk residues that are
> not required for minimum function do not count toward the calculation
> of minimum functional complexity.

There is no "minimum part requirement". See references in the original
post.

> Hmmmm, but isn't architecture dependent upon amino acid order? The
> greater the specificity of the required architecture, the greater the
> specificity of amino acid order - right?

Wrong.

One, the overall composition of a segment is what determines
architecture (see Chou-Fasman rules). Two, the larger the protein,
less constraint to its exact structural composition (less constraint
on how extensively it can be modified even *structurally*).

> These are difficult too, but some larger proteins are also fairly
> specified.

Examples, please. If you actually bother to *look*, you will be
suprised.

>And, you forget, that I'm not just talking about larger
>single proteins which often loose specificity requirements, but about
>systems that require multiple smaller proteins working together at
the
>same time (as in various bacterial motility systems like the
flagellar
>system). Such multiprotein systems require many fairly specified
>proteins working together at the same time.

Again, LOOK at the references I gave you (and also the examples in the
very first post I sent in response to your theory). I have both given
you examples of exactly such systems arising in lab. AND all of the
abovementioned applies to them too; while proteins are "working
together", their individual amino acids are not.

>Such multiprotein functions are quite specified.

Wrong.

> LOL - oh really? What then, is the shortest cytochrome c sequence
> that would have the cytochrome c type of function?

LOL indeed! Read your question: what is the shortest cytochrome c that
is a chytochrome c? :))) Answer: if it is half the size, and different
in mechanism, we will call it something else; which is why cytochrome
c is cytochrome c.

But let's answer the implied question: what can perform the function
of cytochrome c, i.e. electron transfer function? Heme group on its
own is a decent electron transporter, all you need is any kind of
protein to bind it and prevent two hemes from coming in contact with
each other. Other then that, check flavoproteins, copper proteins,
quinoproteins, molybdenium-containing hydroxylases...

> Likewise, what is the shortest lactase enzyme that will actually work
> in a beneficial way in a living creature? You have gone off about how
> only a handful of amino acids are really important for such functions
> and how the rest of the amino acids are not nearly as specified.
> However, there definitely is a degree of specificity even for those
> amino acids that are not fully constrained.

What is that degree of specificity, Sean? Do tell. Do give the math
for once, instead of just giving powers of twenty. Even better, let's
use some negative controls. Apply your theory to a protein that HAS
evolved in the lab; say, fucosidase - what is the "minimum number of
fairly specified amino acids" that is "required" for fucosidase
function, and what is the chance of it evolving? We know that it has
evolved, but test your theory: does your theory predict that
fucosidase should be easy to evolve?

You keep pulling huge numbers out of nowhere, while I (and others)
keep giving you examples of lactase function evolving in the lab. If
it worked your way, that couldn't happen. I also gave you examples of
short polypeptides acting as beta-galactosidases.

> Take cytochrome c for example. The shortest functional cytochrome c
> sequence that I know of requires a minimum of around 80 or so amino
> acids. Of these about 20-30 are pretty much "invariant". Many of the
> others (around 25) are highly constrained to within just 2 or 3 amino
> acids having similar chemical natures. For example, Lys can change to
> Arg, Glu can change to Asp, Ala can change to Val, and Val can change
> to Ile etc. Surface AA residues can vary more than those in the
> interior. Those in key positions involved in the functionality of the
> protein are more constrained and can not vary without disturbing the
> 3-D shape of the protein and therefore its functionality. Already we
> are up to around 40 to 60 highly specified amino acids.

Are we indeed? Take a look at the genetic code now. Look at how many
combinations code for Val, and how many for Met. Think on it a bit. It
screws up your probability calculations mightily.

Oh, I'll grant you that there are many small, very constrained
proteins (histones are, again, the textbook example). But whenever I
give you an example of a small protein evolving, you laugh at me and
tell me it's to small. Well, let me laugh here, and tell you that
cytochrome c is too small. Why don't you try to analyze a large
protein in the same way? Take lactase: how much can its structure
change (hint: take a look at various 3D structures of various
lactases, and compare sequences). You will find that the larger the
protein gets, less constrained it is. And, in addition, how do you
know how ancient cells performed the function of electron transport?

>This is certainly true just like I can change any one letter in this
paragraph
>without a complete loss of function. However, you must admit that
not too
>many letters can be changed very much at all at the same time without
a
>complete loss of beneficial function. The constraints on the
cytochrome
>c type of function are obviously significant. But what does this
mean in
>regards to my position?

Yes, I will admit that you can't change too many letters in a
paragraph without changing the meaning. This is why language is a poor
example for the way proteins work. In proteins, I can change ALL the
letters in a paragraph, except for one or two, and as long as those
two are in the same approximate relation to each other spatially, it
will still work.

> The higher the constraints and the greater the minimum amino acid
> requirement, the less the density of that type of function in sequence
> space. For the cytochrome c function, in particular, some have
> suggested that the total number of functional cytochrome c sequences
> in sequence space is around 10e60. The problem here is that 10e60 is
> only a minute fraction of the total sequence space of 10e104 at this
> level of specified complexity.

I have a better idea. Let us measure the density of hypocrisy in the
space of the preceding paragraph, starting from the end.

As you could have learned by now, genetic code is degenerate, giving
raise to some mutations far more frequently then other; and the basic
motifs of proteins will reappear again and again, driven by
thermodynamics. Therefore, your statement that the total "sequence
space" is 10e104 is a direct, knowing lie. You *know* that this number
means nothing, yet you repeat it.

Going back, I don't know who suggested that the number of functional
cytochrome c sequences is 10e60. I didn't, and I disagree with that
number. Neither I nor you can easily (i.e. without a lifetime of work)
find out how many functional sequences there are.

Thirdly, you are again doing what you insist that you don't do, yet
keep repeating: you look at the sequence of *currently existing*
protein, that has *evolved to perform a sepcific function*, and
calculate the odds for its existence. A good analogy has been given to
you before: calculate the chance of your parents meeting, among all
the people in the world; do so for your grandparents, then
great-grandparents. Just six or seven generations back, what are the
chances of you being born?

Cytochrome c didn't always exist. It started off as a heme group bound
to some other protein, and then adapted first to fit in the membrane,
and then slowly along with other regulatory pathways that influence it
today. It never had to be designed for scratch. There was no random
walk through 10 to the umpteenth power of possible functions. Look at
the past papers in the Journal of Molecular evolution to see the steps
in its development: and then find one that is highly unlikely, and
explain WHY is it highly unlikely. Then you will have something worth
listening to, not just diatribes how an imaginary process couldn't
happen.

> evolution complete stalls out at the level of just a few thousand
> fairly specified amino acids working in concert.

Oh, I agree: if evolution and molecular synthesis worked as you
imagine them to work, it could never happen. But neither evolution nor
molecular synthesis work like that, so it is a moot point.

>I'm sure that there are literally trillions upon trillions of
different potential
>lactase sequences in sequence space (many having low sequence
identity).

No, you do not understand, and you do not agree. You are just
repeating your argument instead of listening.

>For argument's sake let's say that the shortest functional lactase
enzyme
>requires at least 400aa to start working. Now, there most certainly
would be a fair
>degree of flexibility to such a function as a lactase enzyme.

You know, let's. Let's say, for argument's sake, that your nonsensical
repetitive **** about minimum residue requirement really exists, ok.
Your sequence space still has nothing to do with reality.

Why? Because you don't start from scratch. You start from a fairly
large beta-galactosidase, and add lactase recognition function to it.
That is all. Addition of a domain to a functional protein. Even if we
ignore the adaptive funnel (you yourself agreed, not realizing that
you are sinking your theory right then and there, that proteins can
gain specificity very, very quickly; not that your agreement means
much, there are piles of evidence for it in directed evolution
studies), random walk here includes a hundred or so residues, not 300.

Let's go on. Let's ignore the fact that there are small molecules that
act as quite good beta-galactosidase, and say that the lactose
precursor is 250 residues long. How do we get that? Well, by
adaptation (funnel again) of an unselective glycoside hydrolase. Which
brings us down to about a hundred residues.

This inability to comprehend that evolution builds up on already
existing molecules has always been one of the larger holes in your
theory (although your theory has more holes then substance, so it is
hard to tell which is the largest one).

>However, these functions, taken by themselves, do indeed require
>a minimum number and specified orientation of parts.

Really? Let's look at the *function*, please. Of, say, a car.
Transporting people. Yes, there is a minimum amount of parts a car has
to have. However, a car is a refinement of a carriage, with horse
replaced by a combustion engine. And a carriage is a refinement of a
cart. And you can remove horse from a cart, and pull it yourself, if
it's small enough. And you can even remove the wheels, and drag it.
Improvements make it more efficient, but it is a buildup process.

> The same is true for protein functions. Each type of protein function
> requires both a minimum number of amino acids as well as a minimum
> specified orientation of these amino acids before that type of
> function can be realized.

I am giving up on repeating myself. If you decide to actually check on
what we know about protein functions, you will find out that you are
wrong. If you refuse to, it is entirely your own problem.

>No matter which way you choose to create a functionally beneficial
>lactase enzyme of cytochrome c electron transport protein, you will
>not be able to go below a certain minimum amino acid level.
>That is a cold hard fact. I really don't see how you can get around
it
>although I do sympathize with your efforts to try and think of a way
>of talking yourself out of this most interesting little predicament.

Wow. The level of hypocrisy in this paragraph is...wow.

I have given you REFERENCES to small proteins performing lactase
function. Synthesize them, check it for yourself if you do not believe
it. Electron transport? NADPH, for Odin's sake! You don't even need a
single amino acid. Cold hard facts indeed. There is nothing cold, hard
or factual about them. They are warm, soft, moist, and smelly. The
word for them is bullshit.

> Not at all. You need to have some idea of all the possibilities in
> sequence space.

No, I don't, since "sequence space" has nothing to do with protein
evolution.

>Again, my argument concerns the *minimum* amino acid requirement
>needed to gain a particular type of function.

Pick a protein. Tell me what is the minimum. You picked lactase, I
showed you smaller proteins that perform the job (again, ignoring the
fact that evolution builds up, and that you can start from a small
protein and get to lactase, without going through vast "sequence
spaces" in a random walk).

>I never said that evolution needs to create any sort of "random"
>non-functional sequence. You can start with whatever you want.
>For example, start with a function already in place that requires
10,000
>fairly specified amino acids (such as in a multiprotein system where
>all the proteins work together at the same time to create a unified
>function). Then, starting with such a functional system, evolve a
new
>type of function at the same level of complexity or greater using the
>mindless mechanisms of evolution (genetic duplication, random point
>mutations, insertions etc. combined with natural selection).

See the references. Look up "Directed Evolution", a company that earns
money doing what you say is impossible. Go tell their investors to
pull their money, that they cannot possibly get what they are getting
from the company. Oh, ooops, they *ARE* already getting it? Never
mind.

>This is where you go wrong. A 10,000 residue protein produced at
>random may indeed come across a new function, but NOT a new type of
>function that requires 10,000 amino acids AT MINIMUM.

:))))) Really. An intriguing statement. Why do you think so? For
example, in your twisted biochemistry, a connexin has a quite large
residue "requirement". Take one that transports Ca2+. It can easily
mutate into one that transports K+. The second one also has a large
residue "requirement"; it came from a mutation of another such
protein. So, new function, "minimum residue requirement", it's all
there. Why doesn't it qualify?

Oh, too small of a change. You want a connexin to transform, one step,
into something radically different, of course. Well, that won't
likely happen, of course. Which is a death-blow to the twisted theory
of evolution that exists in your head. And does nothing to the theory
of evolution as it actually is, for reasons that have been explained
above, and before.

>You do understand the difference here - right?

Yes, it is the difference between imaginary biochemistry and actual
biochemistry. Read the references I gave you.

>When you can't demonstrate the actual evolution of novel
>functions beyond the lowest levels of complexity, you revert
>to the nested hierarchy argument. This is nothing more than
>an argument that assumes that shared similarities must mean
>common evolutionary ancestry.

Indeed. How incredibly stupid we all are.

No, Sean. It is not just similarity. It is the fact that you can see
many genes at a simpler stage of development. It takes incredible
chance to produce human RNA polymerase de novo. It takes only slight
adaptation to do so from a simpler primate polymerase. Which also
requires only a slight development from a simpler mammalian
polymerase...and so on, until you reach simple polymerases at the
bacterial level. And so you claim that there is no chance that a
molecule such as human RNA polymerase coud be found by random walk, it
is irrelevant; it wasn't found by a random walk. This is why the
nested hierarchy is important. It isn't just "hey, it looks alike", it
is "this is just a few small modifications away from that".

>Shared similarities are explained equally well by both positions.
>In order to rule out my position of intelligent design, you must
>show that evolutionary mechanisms can explain the differences
>as well as the similarities found within and between various life
forms.

No, they are not explained equally well by both positions.
Evolutionary theory explains them on the basis of testable, visible,
repeatable knowledge on the laws of the universe. Laws that we observe
again and again. Intelligent Design does not explain anything; it puts
a period, and demands a stop to research: there is no point in trying
to understand the world, when the world was created by an entity we
can infer nothing about.

And evolutionary mechanisms do explain the differences quite well
enough. That is why all those biochemists, biologists, biophysicist,
physiologists...etc, take it for granted. They are not stupid, deluded
people who, for some evil reason, seek only to deny the existence of
your God.

> This is not "absolutely certain" above the lowest levels of minimum
> specified complexity. You have made a bold assertion that this is
> true, but you have yet to back yourself up with anything other than
> assumed historical correlatory comparisons between similar functional
> sequences and systems in various living things.

I have given you references to random large proteins being made that
perform complex functions. I have provided you with some introductory
reading as to structural theory of proteins. If you try to produce
proteins yourself, you will find that they will almost invariably bind
quite a few substrates, and that many will spontaneously form complex
polymers in organized, symmetrical fashion. Etc, etc. All of this in
the post you are replying to. And which you didn't read before
starting your response.

>You have no real-time example of evolution in action now do you?
>Just admit it and move on.

I have been providing a few in practically every post I sent in
response to your statements. I have provided quite a list in the last
one. I cannot provide evidence to a person who refuses to look at it.

I will let it rest here, and give you a chance to read through my
post. It is quite obvious by now that you don't read any of the
responses anyone has provided. You hit reply, and start answering by
repeating you position, paragraph by paragraph, not even looking to
see what lays few paragraphs beyond that. I believe my work here is
done. If you provide anything besides repetition in your next post, I
will respond, otherwise I won't bother - I'll just refer you back to
my original post and wait to see whether you'll actually read it
someday.

> Perhaps someday you will actually understand the very real problem
> that these very relevant mathematical constructs pose for your faith
> in the very strange notion of evolutionary progression.

Perhaps you will someday learn some biochemistry, and understand that
no amount of imagination can deny facts of life. And for the end, I
will repeat the two questions I would like you to answer the most:

One, what do you actually believe happened? When did the Earth come to
existance, and how? What is the history of the Earth as you see it?

Two, why don't you publish your devastating theory in a professional
journal? Why do you think they won't accept it?

M.

[Sean Pitman]

sweetnes...@yahoo.com wrote in message news:<4d71d185.03121...@posting.google.com>...
>

> Howard Hershey has already asked most of the relevant questions
> (thanks!), so I will just give a few relevant comments. I would,
> however, ask you to sit down and read the rest of my original message
> before you respond to this.

I read your original message, in its entirety, before I responded the
first time. However, responding to everything in particular takes a
lot of time, especially since many of the ideas you were presenting
were directed against a straw man version of my position. I only had
time to cover a few key points were you misinterpreted my position.
Many if not all of your references do not deal with anything that
challenges my actual position. Showing me reference concerning the
development of very large proteins with various enzymatic functions is
not helpful since much of the size of the proteins in your long lists
are not needed for a particular type of beneficial function to be
realized. I am interested in the minimum size requirement not the
maximum size possible that could produce a given type of function.

So far, you have come up with no more impressive examples of evolution
in action than I have seen dozens of evolutionists before you present
as conclusive evidence. The only problem with these examples of real
time evolution is that the minimum fairly specified amino acid "part"
requirements have been no more than a few hundred amino acids, working
at the same time, at best. Of course, you have also presented examples
of very large proteins that do relatively simple functions that do not
require the protein to be so large before that type of function can be
realized at some level of beneficial selectability. Don't feel bad
though. Others in this forum, such as Ian Musgrave, have tried to
float this same sort of evidence as something rather "devastating" to
my position. He presented the Titan protein, which is truly huge, but
is basically made up of many repeats of the same protein
sequences/structures over and over again - polymer-like. The minimum
genetic information/real estate required to code for this type of
protein function is relatively small. In other words, this type of
protein function requires very low specificity - much lower than say a
collection of proteins working together at the same time in a
flagellar-type motility system (totaling several thousand much more
specified amino acids at minimum working together at the same time for
that type of motility function to be realized). You see, you cannot
simply quote to me reams of papers talking about large enzymes with
low sequence specificity requirements. By now you should know that
such examples are nothing more than the presentation of large enzymes
with informationally simplistic functions that do not require these
enzymes to be nearly that large in order for their basic functions to
be realized in a beneficial way.

Now, of course you will argue that my suggestion that certain
functions, such as the lactase function, have a minimum amino acid
requirement are "ludicrous". Over and over again you suggest that
only a handful of amino acids are really required for the lactase
function to be realized. Well, you know sweetness, until you can show
me a lactase enzyme made up of only a handful of amino acids working
in a beneficial way in a real life form, I'd have to say that I have
my serious doubts about such naked assertions. In real life it seems
like the lactase function requires at least 400+ amino acids working
at the same time. Now granted, the specificity of the lactase
function might not be all that great, but it seems to be much greater
than many who respond to my thread seem to realize. Some have
suggested that the ratio of lactase enzymes in sequence space
(requiring 500aa or less), is as high as 1 in 1000 sequences. Really
now, if this were the true ratio of beneficial lactases vs.
non-lactases, absolutely all average sized bacterial colonies would be
able to evolve the lactase function within a single generation. On
average only 2.3 amino acid positions would need to change in some
specific way for such a gap to be crossed. The E. coli experiments
done by Barry Hall and others prove conclusively that the density of
lactases in sequence space is far lower than 1 in 1000. More likely
it is somewhere beyond 1 in 10e20 or Hall's E. coli would have evolved
lactases over and over again during the tens of thousands of
generations that they were observed. The fact that only one other
sequences was close enough to even one of the potential lactases in
sequence space to find it in tens of thousands of generations is very
good evidence that the lactase function is indeed fairly specified in
both its minimum amino acid requirement and degree of limitation on
the arrangement of these amino acids.

> > I didn't say this at all. What I said was that if the minimum part
> > requirement needed to gain the type of function of protein C was a few
> > thousand fairly specified amino acids, then going from A to C would
> > indeed require trillions upon trillions of years on average.
>
> What I am saying again and again is that there is no such thing as
> minimum amino acid requirement.

Yes, and what I am saying again and again is that there certainly is
such a thing as a minimum amino acid requirement for all types of
cellular functions that are protein dependent. Tell me again, how
many amino acids, at minimum, can you get to actually produce a
beneficial lactase function in a living organism? Can you do it with
a protein made up of just 10 amino acids? I know that you have tried
to make this claim, but where has this bold claim been supported in
the beneficial use of such a small lactase enzyme in an actual living
creature? I know I know, you argue that only a handful of positions
are actually important. The rest of the supporting cast of amino
acids can be arranged in any old way - right? But this simply is not
true. Many of the other amino acids have constraints as well. Not
just any amino acid will work in even the less constrained positions
of the molecule. Some of the positions are restricted to certain
classes of amino acids (i.e., acidic, basic, hydrophilic, hydrophobic,
etc). Others positions are restricted to within only 2 or 3 different
amino acids. Some are more loosely restricted to within a dozen or so
amino acids. Only a few positions can be occupied widely by all 20
amino acids. Of course, taken one at a time, most positions can be
changed significantly, but this is not true if too many positions are
changed at the same time. All of these constraints lessen the density
of such types of beneficial sequences in sequence space vs. the
non-beneficial sequences that greatly outnumber them.

For example, in the June 2003 issue of Nature, Jack Szostak, in a
discussion entitled, "Molecular Messages" observed that functional
aptamers and ribozymes (60-90 bases) had a density of only 1 in 10e10
at best. The sequence space at this level of complexity is only 10e54
sequences max. On average, a randomly mutating genome that would
benefit from a ribozyme sequence, not having one already or needed a
new one, would have to undergo a random walk of 10e10 non-beneficial
steps before it found one that would have this type of function.
Although the average time for an average bacterial colony to cross
this gap would be only one or two generations, the concept here is
important. These functions require relatively short averagely
specified genetic sequences to code for them, and yet they still
require a minimum number of bases as well as a minimum amount of
sequence specificity. If they didn't, then the ratio would not be 1
in 10e10. It would be 1 in 1. Doesn't this make sense to you?

> You have invented that out of thin
> air. It is as simple as that. You are simply refusing to look at the
> evidence, and repeating the same things over and over again.

Hmmmm . . . obviously that is your very clear opinion, but I simply
disagree. I actually do think that I am looking at the evidence that
is being presented in an honest and open way. Certainly you
understand/interpret the evidence in a very different way than I do -
and you may be absolutely correct in your views and I could be
completely off my rocker in my views. However, until my dull wit is
actually enlightened enough to understand things as you do, I honestly
don't get it. I think you are wrong and that your understanding is
confused. Of course, I understand and respect the fact that you see
me as the one that is obviously so confused as to be pretty much
beyond any sort of hope. Maybe you're right, but that doesn't help me
now does it? Those who are following these threads must also judge
for themselves.

> > This is not what I'm saying. You can build up your larger specified
> > proteins with new types of beneficial functions however you want. No
> > matter how you try to do it, it will not work beyond the lowest levels
> > of functional complexity.
>
> I have provided you with examples and references to proteins doing
> exactly that which you say is impossible. Large proteins, thousands of
> residues long. There have been directed evolution experiments
> involving RANDOM long proteins (random mixing of sheared DNA
> fragments) that produced some fairly humongous functional enzymes.
> Take a look at the reading list I provided for you.

I'm not sure how I can make this any more clear to you beyond the way
in which I responded last time. I have read about many such
experiments that you quote and reference here. I fully accept and
agree with the results and many of the conclusions of the researchers
in such cases. However, as I mentioned many times before, just
because a protein is big does not make it more functionally complex
(i.e., requiring more genetic real estate at minimum). A much smaller
protein can be much more informationally complex than a much larger
protein due to a much higher degree of sequence specificity required
to achieve a particular type of function. So, I say again, I am
interested in knowing the minimum amino acid requirement to achieve a
particular type of function at a level where it would be beneficial in
a given life form (not just the ideas of what a lab scientists thinks
might be beneficial - it must actually be tried out in a living
thing).

> > But what is the minimum amino acid requirement for this new beneficial
> > function? Did the new beneficial function really require all of the
> > amino acids from the large protein or just a small stretch of them?
> > In other words, you must ask, "What is the shortest amino acid
> > sequence that would yield this particular type of function working in
> > this particular way?"
>
> Take a look at the literature; there are snippets in the list I gave
> you to read, but there is so much more if you bother to read a
> fricking journal now and then.

Oh, a *frickin* journal?! Why didn't you say so? LOL - you know, I
have read lots and lots of journals - especially those written by
evolutionists such as yourself. I dare say that I read more
publications written by evolutionists than most evolutionists in this
forum, and certainly more than most creationists/IDists that I know
of. The fact of the matter is that I have failed to find anything
that significantly challenges my position as of today. You certainly
haven't referred me to anything that comes even close. Really now, a
few things that you have said do sound intriguing, but the one
observation that you present as being most convincingly against my
position you say has not been published yet. Ok, well maybe when it
is published you will notify me of your success. Until then, you have
come up with a lot of examples of large proteins with very low
specificity and the evolution of a new function requiring quite a few
rather specified amino acids in the form of multiple proteins, which
turned out not to require any new protein structures/sequences at all,
but simply an increased production of what was already there - nothing
new evolved at all (your bacterial swarming example). Come one now,
do you really think such lame examples are enough to hang your hat on?

> Your question is meaningless, but no matter how many examples and
> references I provide, you just keep repeating it over and over.

Hmmm . . . I don't seem to be the only one who likes to repeat himself
. . .

> Very
> well, let us try a different approach. You choose a large protein -
> any large protein over one thousand residues. Just pick one. Then YOU
> tell ME how many of its amino acids are REQUIRED for its function, and
> WHY you think that they are all REQUIRED.

Ok, take the flagellar system of motility. At minimum such a system
requires 20 or so different types of proteins working together at the
same time. The amino acid total involved is over 6,000 fairly
specified amino acids working together at the same time. If you have
less than this minimum, you will not get the flagellar motility
function - period. If you reduce the flagellar system below this
minimum part requirement, the beneficial motility function will vanish
completely.

Oh, but what about the secretory function (TTSS)? Well, the secretory
function may still be maintained since it requires far fewer specified
amino acids working together at minimum in order for its beneficial
function to be realized. The gap between the minimum requirement of
the secretory pore and the motility function of such a secretory
structure to be realized is involves over 3,000 fairly specified amino
acids.

There are many who have proposed various stepping-stone pathways
across this gap, but none of these proposed pathways has been
supported by experimental testing/demonstration. Not even one of the
proposed steps has been demonstrated to evolve - not one. Many
scientists even challenge the idea that the TTSS system was actually
the precursor of the motile flagellum. The current views of most is
that the TTSS system evolved independently and some actually suggest
that the TTSS system evolved from the fully formed motile flagellar
system (Nguyen et al. 2000)

> > Not if you consider the minimum part requirement, not the maximum junk
> > you can include with this minimum. The extra junk residues that are
> > not required for minimum function do not count toward the calculation
> > of minimum functional complexity.
>
> There is no "minimum part requirement". See references in the original
> post.

Ok, I looked at your references. Which one was it again that showed
how all protein functions can be had by any amino acid sequence of any
length and order?

> > Hmmmm, but isn't architecture dependent upon amino acid order? The
> > greater the specificity of the required architecture, the greater the
> > specificity of amino acid order - right?
>
> Wrong.
>
> One, the overall composition of a segment is what determines
> architecture (see Chou-Fasman rules).

Doesn't the composition of a segment require a certain degree of amino
acid specificity?

> Two, the larger the protein,
> less constraint to its exact structural composition (less constraint
> on how extensively it can be modified even *structurally*).

Again, many of your examples of large proteins are not required to be
nearly as large as they are in order to achieve a beneficial level of
the type of function that they currently perform . . . isn't this
true? Also, we are not talking only about single protein functions
here. You must also consider those functions that require multiple
smaller fairly specified proteins working together at the same time -
as in the flagellar motility system. The total number of amino acids
involved in such a case is rather great, as is the level of overall
amino acid order/specificity. So you see, you can get a larger
protein *system* of function without a corresponding decrease in the
percentage/level of constrained positions.

> > These are difficult too, but some larger proteins are also fairly
> > specified.
>
> Examples, please. If you actually bother to *look*, you will be
> suprised.

Oh really? I would suggest to you that the lactase function requires
a higher level of information/genetic real estate than does the
function of the mammoth Titan protein. Then, going beyond this, the
genetic information required to code for the flagellar system is far
beyond either one of these. What do you think of these oft-repeated
examples?

> > And, you forget, that I'm not just talking about larger
> > single proteins which often loose specificity requirements, but
> > about systems that require multiple smaller proteins working
> > together at the same time (as in various bacterial motility
> > systems like the flagellar system). Such multiprotein systems
> > require many fairly specified proteins working together at
> > the same time.
>
> Again, LOOK at the references I gave you (and also the examples in the
> very first post I sent in response to your theory). I have both given
> you examples of exactly such systems arising in lab.

Where? List your best example here again. The only examples that you
have listed of multiple proteins working together at the same time
involve no more than 200 or 300 amino acids total. Please, what did I
miss here? Oh, and don't try and float your M. xanthus swarming
example again at this point. This "evolution" did not involve the
production of anything new structure/protein sequence at all. It was
simply an increased production of exactly the same protein matrix that
was already being made. Come on now, did I miss where you actually
presented something better?

> AND all of the
> above mentioned applies to them too; while proteins are "working

> together", their individual amino acids are not.

Uh, I might be a bit too slow to understand how you could possibly
have just made this statement. So, perhaps you could explain to me
your sizzling rational for stating that proteins can work together but
their subparts (amino acids) having nothing to do with how they work?
Haven't you ever heard of "emergent functions"? Without the
underlying order of interacting parts, the parts themselves will not
work together properly. For example, consider the fact that in order
to walk the legs must be attached and interact properly with the torso
of the body. However, the leg suparts must also be in proper order as
well or the leg simply will not work right even if it is attached
properly to the torso. If the there is something wrong with the actin
or myosin of the muscle groups, they will not contract properly and
the leg will not respond to the nervous signal from the brain. So, in
reality, all levels of order are required and are thus working
together at the same time in a system whos highest levels of emergent
function require the simultaneous interaction of the highest level
parts.

> >Such multiprotein functions are quite specified.
>
> Wrong.

Nice "just-so" statement, but it comes across a bit hollow. You
simply haven't backed yourself up in a convincing manner.

> > LOL - oh really? What then, is the shortest cytochrome c sequence
> > that would have the cytochrome c type of function?
>
> LOL indeed! Read your question: what is the shortest cytochrome c that
> is a chytochrome c? :))) Answer: if it is half the size, and different
> in mechanism, we will call it something else; which is why cytochrome
> c is cytochrome c.

Obviously true. If it is half the size and has a different function,
it will not have the type of function of a cytochrome c sequence
(i.e., a rather specified type of electron transport protein).

> But let's answer the implied question: what can perform the function
> of cytochrome c, i.e. electron transfer function? Heme group on its
> own is a decent electron transporter, all you need is any kind of
> protein to bind it and prevent two hemes from coming in contact with
> each other. Other then that, check flavoproteins, copper proteins,
> quinoproteins, molybdenium-containing hydroxylases...

Oh, so just about any protein sequence would work just fine as an
electron transport protein in place of a cytochrome c sequence in
living creatures? Has this assertion been supported by any sort of
relevant demonstration in a living organism? The fact of the matter
is that many scientists disagree with you here. For example, Yockey
has suggested (and Ian Musgrave brought this estimate to my attention)
that the total number of all functional cytochrome c sequences is
around 10e60 (and this is out of 10e100 possibilities). The fact of
the matter is that the vast majority of possible protein sequences of
similar size will not work at all as a replacement for cytochrome c
since they just don't have a similar enough functional potential to be
beneficial to any given life form as an electron transport protein.

> You keep pulling huge numbers out of nowhere, while I (and others)
> keep giving you examples of lactase function evolving in the lab. If
> it worked your way, that couldn't happen. I also gave you examples of
> short polypeptides acting as beta-galactosidases.

In-vivo examples? You have made assertions that your laboratory
experiments would actually have some sort of correlation in some
potential life form, but you have yet to show that anything shorter
than 400 or so amino acids would actually work in a beneficial manner
as a lactase enzyme in an actual life form. Please, you have try out
your assertions/experiments where they would really mean something
(i.e., in an actual life form). You have only made it through the
first part of the scientific method. You have a fine hypothesis and a
fine prediction based on this hypothesis. Now, all you need to do is
test your predictions and see if they hold up in real life. Once you
have done this, get back to me.

> > Take cytochrome c for example. The shortest functional cytochrome c
> > sequence that I know of requires a minimum of around 80 or so amino
> > acids. Of these about 20-30 are pretty much "invariant". Many of the
> > others (around 25) are highly constrained to within just 2 or 3 amino
> > acids having similar chemical natures. For example, Lys can change to
> > Arg, Glu can change to Asp, Ala can change to Val, and Val can change
> > to Ile etc. Surface AA residues can vary more than those in the
> > interior. Those in key positions involved in the functionality of the
> > protein are more constrained and can not vary without disturbing the
> > 3-D shape of the protein and therefore its functionality. Already we
> > are up to around 40 to 60 highly specified amino acids.
>
> Are we indeed? Take a look at the genetic code now. Look at how many
> combinations code for Val, and how many for Met. Think on it a bit. It
> screws up your probability calculations mightily.

Oh really? How so? There are 4 combinations for valine and one for
methionine. There are also six each for leucine and arginine. On
average though, there are 3.2 codons per amino acid. Say we have 60
highly specified amino acids. On average, the likelihood is that
these amino acids will be a fairly even mix with an average of 3.2
codons per amino acid. So, averages still work out rather nicely
suggesting that getting a particular amino acid position right in a
highly constrained position via random chance is still around 1 in 20.
At the very worst the average might be as high as 1 in 15 or so, but
this does not "mightily" mess up my probability calculations at all.
In fact, I find it rather amusing that you would even suggest such a
thing. You are the very first one to challenge me in this way. I
must give you points for creativity here though.

> Oh, I'll grant you that there are many small, very constrained
> proteins (histones are, again, the textbook example). But whenever I
> give you an example of a small protein evolving, you laugh at me and
> tell me it's to small. Well, let me laugh here, and tell you that
> cytochrome c is too small. Why don't you try to analyze a large
> protein in the same way? Take lactase: how much can its structure
> change (hint: take a look at various 3D structures of various
> lactases, and compare sequences). You will find that the larger the
> protein gets, less constrained it is. And, in addition, how do you
> know how ancient cells performed the function of electron transport?

I have asked this question of several apparently well-informed
individuals. Some, such as Ian Musgrave, have actually given me an
understandable response. Ian suggested to me that the total number of
lactase enzymes in sequence space was probably around 10e100. Of
course, this was before he suggested to me that the minimum lactase
sequence probably required more than 400aa - which would result in a
sequence space of well over 10e500. From his more resent posts, I'm
sure that Ian would markedly increase his estimate so that the ratio
would be closer to 1 in 10e12 or so. However, I have problems with
this ratio considering that at such a high ratio I would expect that
all types of bacteria, with an average sized colony of 10 billion or
so individuals, would quickly evolve a function that had such a high
ratio of sequences in sequence space (a few months at most). The
problem here is that Hall's experiment with E. coli showed that the
lactase function did not evolve in ebg negative E. coli in tens of
thousands of generations. This tells me that the lactase function
requires a bit more sequence specificity than many in this forum have
recently suggested to me.

> Yes, I will admit that you can't change too many letters in a
> paragraph without changing the meaning. This is why language is a poor
> example for the way proteins work. In proteins, I can change ALL the
> letters in a paragraph, except for one or two, and as long as those
> two are in the same approximate relation to each other spatially, it
> will still work.

You overstate your position. For many types of single protein
functions there are a handful of proteins that cannot really change at
all. These are called, "invariant" amino acid positions. At this
point we are both in agreement. But, after this point you overstate
your position. You seem to be suggesting that *all* of the rest of
the amino acid positions can freely change without any real damage to
protein function. This is simply not true, or at best it is a huge
understatement of the actual constraints involved with the rest of the
amino acid positions. Many other amino acids beside those that cannot
change at all can only change between 2 or 3 very similar amino acids.
Many more can change only between certain classes of amino acids as
noted above. Only a very few can change easily between all 20 amino
acids. A limitation to class or to a handful of amino acids is still
a significant limitation. Each limitation decreases, in an
exponential fashion, the density of beneficial sequences with this
type of function in sequence space (i.e., it markedly reduces the size
of the gray areas around each "dot").

Anyway, this is all I have time for today. Thanks again for your time
and the great effort (despite a lot of significant frustration I'm
sure) that you must put into your posts. I really do admire your
sincerity and your energy.

Sean
www.naturalselection.0catch.com

[Howard Hershey]

Sean Pitman wrote:
>
> sweetnes...@yahoo.com wrote in message news:<4d71d185.03121...@posting.google.com>...
> >
> > Howard Hershey has already asked most of the relevant questions
> > (thanks!), so I will just give a few relevant comments. I would,
> > however, ask you to sit down and read the rest of my original message
> > before you respond to this.
>
> I read your original message, in its entirety, before I responded the
> first time. However, responding to everything in particular takes a
> lot of time, especially since many of the ideas you were presenting
> were directed against a straw man version of my position.

It is not a straw man to say that your position is that you cannot or
will not distinguish between the total number of amino acids in a system
and the 'minimum number of amino acids'.

> I only had
> time to cover a few key points were you misinterpreted my position.
> Many if not all of your references do not deal with anything that
> challenges my actual position.

You have an "actual" position?

> Showing me reference concerning the
> development of very large proteins with various enzymatic functions is
> not helpful since much of the size of the proteins in your long lists
> are not needed for a particular type of beneficial function to be
> realized.

That is *always* true, Sean. Enzymes typically act to change or modify
small molecules or small parts of larger molecules (e.g, the
betagalactoside linkage of lactose and many other substrates). The
active site of enzymes is almost always only *a few* specific
(invariant) amino acids in particular positions on the molecule. The
rest of the molecule's function is primarily arranging positions in 3-D
space so that those few molecules form a catalytic surface that lowers
the energy of activation of a reaction so that occurs faster than it
would in the absence of the enzyme. No enzyme violates the laws of
thermodynamics. Typically what the active site does is stretch or bend
a substrate so as to make a bond more likely to break or, alternatively,
bring substrates into specific contact so that a particular bond is
likely to form. Often one of the substrates/products is a water
molecule (which obviously is not bound by "thousands" of amino acids in
a protein). In other cases, one step involves the breaking of a high
energy molecular bond used to make the entire reaction thermodynamically
feasible (all enzymatic reactions like all chemical reactions, in net,
run downhill). More rarely, the active function involves a number of
less specific amino acids touching many positions in a large molecule.
The canonical example of the latter are the histones.

But even protein-protein binding does not involve 'thousands' or
'hundreds' or even, usually, 'tens' of binding sites. If they did,
binding would be irreversible and many proteins and systems depend upon
the reversibility of aggregation into complexes.

The only way you get your "thousands" of amino acids is by counting
every single amino acid in a protein as if it needed to be invariant.
You certainly give no clue that you do otherwise.

> I am interested in the minimum size requirement not the
> maximum size possible that could produce a given type of function.

So, in the bacterial flagella, how do you calculate "minimum size
requirement" and for which of the potential intermediate functional
steps between selectable functions is this number in the "thousands"? I
see a lot of handwaving, but no evidence.

>
> So far, you have come up with no more impressive examples of evolution
> in action than I have seen dozens of evolutionists before you present
> as conclusive evidence. The only problem with these examples of real
> time evolution is that the minimum fairly specified amino acid "part"
> requirements have been no more than a few hundred amino acids, working
> at the same time, at best.

Isn't it amazing, then, that the amount of *change* needed to generate
new function is much smaller than even a 'few hundred' amino acids?
Isn't it amazing that all the real examples of real evolutionary change
do not require change in 'thousands' of amino acids? Isn't it amazing
that one can modify the function of a bacterial flagella from it being
primarily a motility device to being solely a protein transport device
with far fewer than 'thousands' of amino acid changes? That should tell
you that your calculations are utterly irrelevant.

> Of course, you have also presented examples
> of very large proteins that do relatively simple functions that do not
> require the protein to be so large before that type of function can be
> realized at some level of beneficial selectability.

So, how many proteins actually meet your requirements? Gawd damn few, I suspect.

> Don't feel bad
> though. Others in this forum, such as Ian Musgrave, have tried to
> float this same sort of evidence as something rather "devastating" to
> my position. He presented the Titan protein, which is truly huge, but
> is basically made up of many repeats of the same protein
> sequences/structures over and over again - polymer-like. The minimum
> genetic information/real estate required to code for this type of
> protein function is relatively small. In other words, this type of
> protein function requires very low specificity - much lower than say a
> collection of proteins working together at the same time in a
> flagellar-type motility system (totaling several thousand much more
> specified amino acids at minimum working together at the same time for
> that type of motility function to be realized).

Can you tell us how you calculated this "several thousands" number? And
how you determined what degree of 'specificity' was required for an
amino acid to be considered 'fairly specified'? Does the bacterial
flagella *really* meet your requirements? How do you deal with the
objection that there are intermediate states of potential utility
(independent of the teleologic goal of motility) in the evolution of the
bacterial flagella? Enquiring minds want to know if you actually are
presenting evidence or merely blowing smoke by tossing out an
unevidenced 'several thousands' of amino acids number?

> You see, you cannot
> simply quote to me reams of papers talking about large enzymes with
> low sequence specificity requirements. By now you should know that
> such examples are nothing more than the presentation of large enzymes
> with informationally simplistic functions that do not require these
> enzymes to be nearly that large in order for their basic functions to
> be realized in a beneficial way.

Look, Sean. Everyone agrees that bacterial flagella did not evolve by
starting from some random set of amino acid sequences and reaching the
current state by a random walk. Not one of us has even hinted that that
is the way it arose. Only you seem to be suggesting that that is the
way that evolution says it appeared. You keep saying stuff and using
numbers that are only consistent with this straw man while
simultaneously denying that that is what you are saying. In the face of
such blatant denial we (meaning everyone else writing in response) are
unable to decide whether you are stupid enough to believe this straw man
version is what evolutionary biologists think or if you think we are
stupid enough to agree with your straw man evolution. Unless you can
start putting together a *reasonable* rather than silly model for how
you think evolution produced the bacterial flagella, you can hardly say
it can't occur.

> Now, of course you will argue that my suggestion that certain
> functions, such as the lactase function, have a minimum amino acid
> requirement are "ludicrous". Over and over again you suggest that
> only a handful of amino acids are really required for the lactase
> function to be realized.

How many amino acids could *possibly* actively affect a beta-galactoside
linkage? There is no possibility of more than a few being involved in
the active site. All the other amino acids can *only* be involved in
carving out 3-D space so that those weary few can provide the proper
surface.

> Well, you know sweetness, until you can show
> me a lactase enzyme made up of only a handful of amino acids working
> in a beneficial way in a real life form, I'd have to say that I have
> my serious doubts about such naked assertions. In real life it seems
> like the lactase function requires at least 400+ amino acids working
> at the same time.

How many of the 400+ are actually "working" (at the active site; binding
the beta galactoside linkage and water in such a way as to increase the
rate of hydrolysis)? Only those amino acids are *required* for lactase
function. And again, it is physically impossible for more than a few
amino acids to actually be in contact with the substrates. This is not
to belittle the need for a particular 3-D structure that, at least part
of the time, produces a surface with enough affinity for the substrates
to reduce the energy of activation and thereby increase the rate of
hydrolysis to a level that has biological significance. But the really
crucial amino acids, those at the active site, must necessarily be
numbered in the less than 10s. It is *physically* impossible for the
number to be much greater. So either a change that produces a changed
amino acid at an active site or a change so as to bring existing amino
acids in the proper 3-D relationship will be sufficient to produce
activity *regardless* of how many amino acids are present in it.

There are *different* ways to generate function, depending on what the
starting point is. It is important to know that in already evolved
systems, the existing proteins from which the new function evolved do
not represent a random sample of all possible sequences. In particular,
evolved systems have a very low incidence of proteins that do not form a
specific structure during protein synthesis. Proteins of evolved
systems are not floppy, regularly flipping among *many* energetically
equal structures. That means that most evolution, which almost always
involves modification of existing systems and proteins, involves
modification of active sites in pre-existing proteins. These proteins
already have clefts and structure that allow what are called secondary
metabolites to enter and leave.

But occasionally evolution does involve the formation of a new activity
(such as nylonase) from what is essentially a random sequence. What
that means is that the intial protein sequence generated formed a
catalytic surface with the ability to lower the energy of activation of
nylon cleavage at least often enough to have biological significance in
environments with sufficient nylon as a potential carbon source. The
secondary mutations that *improve* upon that chance activity will
typically involve mutations, not so much in the active site, but in
proteins that stabilize and optimize the 3-D structure so that the
active site is more effective. Similarly, a chimeric protein formed by
partial duplication, can often be improved by changes in proteins that
optimize the stability or frequency at which the active sites come
together. This latter, rarer form (given that most evolution is due to
modification of pre-existing proteins that have related functions rather
than the types of pure chance events such as the frameshift that
produced nylonase or the formation of a particular chimera) is closer to
what you think must happen (and probably did during the early shift to
protein enzymes way back some 3+ billion years ago): start with a random
sequence that, by chance, has *some* activity (the one-eyed man is king
in the land of the blind) and then produce a *better* functional protein
by stabilizing and fine-tuning the 3-D structure. Once you have done
this, you will find that, even in enzymes where there is virtually no
sequence similarity above that expected by chance alone, the 3-D
structure will be preserved.

> Now granted, the specificity of the lactase
> function might not be all that great, but it seems to be much greater
> than many who respond to my thread seem to realize.

The *specificity* of beta galactosidase activity does not involve more
than a tiny number (probably countable on one hand) of amino acids,
those directly involved at the active site.

> Some have
> suggested that the ratio of lactase enzymes in sequence space
> (requiring 500aa or less), is as high as 1 in 1000 sequences.

How many times do you have to be told that the number of amino acids in
the protein is irrelevant? Let me re-state. The number of amino acids
in a protein is irrelevant *unless* you are trying to make a straw man
argument. If you are trying to make a straw man argument, conflating
the total number of amino acids and the number of amino acids that need
to change is an absolute necessity.

> Really
> now, if this were the true ratio of beneficial lactases vs.
> non-lactases, absolutely all average sized bacterial colonies would be
> able to evolve the lactase function within a single generation.

Not necessarily. What we are saying is that you need a protein in the
system which is only a few amino acids away. And, in E. coli, there are
such proteins. The size of the proteins is irrelevant.

> On
> average only 2.3 amino acid positions would need to change in some
> specific way for such a gap to be crossed. The E. coli experiments
> done by Barry Hall and others prove conclusively that the density of
> lactases in sequence space is far lower than 1 in 1000.

More like 2 quite different proteins that were no more than 1 amino acid
change away from having beta galactosidase activity (one being the
native enzyme). Plus another that was somewhat further away. That is
what has been found. In addition, it is possible to generate a beta
galactosidase activity in an immunoglobin.

> More likely
> it is somewhere beyond 1 in 10e20 or Hall's E. coli would have evolved
> lactases over and over again during the tens of thousands of
> generations that they were observed.

Again, that was not what was observed. What was observed was that there
was at least one other protein that was no more than a single change
away from beta galactosidase activity. And another was somewhat further
away. Remember that there is a difference between selecting for a beta
galactosidase activity and generating such activity. It is virtually
certain that the ebg gene mutated to some beta galactosidase activity
*many* times. But that would have occurred in a cell with a
pre-existing beta galactosidase that had already evolved to satisfy any
need for such activity. It simply could not compete *until* the
original enzyme is removed under conditions where having beta
galactosidase activity is more important than whatever else ebg does.

> The fact that only one other
> sequences was close enough to even one of the potential lactases in
> sequence space to find it in tens of thousands of generations

And what does the number of generations have to do with anything?

> is very
> good evidence that the lactase function is indeed fairly specified in
> both its minimum amino acid requirement and degree of limitation on
> the arrangement of these amino acids.
>
> > > I didn't say this at all. What I said was that if the minimum part
> > > requirement needed to gain the type of function of protein C was a few
> > > thousand fairly specified amino acids, then going from A to C would
> > > indeed require trillions upon trillions of years on average.
> >
> > What I am saying again and again is that there is no such thing as
> > minimum amino acid requirement.
>
> Yes, and what I am saying again and again is that there certainly is
> such a thing as a minimum amino acid requirement for all types of
> cellular functions that are protein dependent.

How do you numerically determine "minimum amino acid" requirement?

> Tell me again, how
> many amino acids, at minimum, can you get to actually produce a
> beneficial lactase function in a living organism? Can you do it with
> a protein made up of just 10 amino acids?

Are you saying that you think that evolution works by adding amino acids
to a protein one amino acid at a time? If not, what are you saying?
How is this relevant to any rational, non-straw man evolutionary
mechanism? It is virtually certain that only 10 or fewer amino acids
are involved in the active site. How could anyone fit in any more than
that? The beta galactosidase linkage and water are not nearly large
enough to contact 'thousands' or 'hundreds' or even 'tens' of amino acids.

> I know that you have tried
> to make this claim, but where has this bold claim been supported in
> the beneficial use of such a small lactase enzyme in an actual living
> creature? I know I know, you argue that only a handful of positions
> are actually important.

It is physically impossible for more than a handful of positions to be
functionally relevant (be in the active site).

> The rest of the supporting cast of amino
> acids can be arranged in any old way - right?

No. The supporting cast is important oragami. But sequence is much less
important to this function. Any set of supporting cast that manages to
get the few functionally relevant amino acids in reasonable enough
proximity to the right shape to have any significant effect on lowering
the energy of activation is a good supporting cast. Typically, for
modern already evolved organisms, new function is not radically new
unless it is an emergent function via alteration of a previous function
(as in motility in the two different rotary flagella in eubacteria and
archaebacteria) or an amplification of a previously minor function (e.g.
lens crystalins). And usually it involves duplication and modification
to produce a modified member of a gene family. But occasionally one
does get a novel function via chance alone (nylonase), or by chance
aggregation of protein motifs via chimera formation (as in the
antifreeze proteins in some Arctic fish).

> But this simply is not
> true. Many of the other amino acids have constraints as well. Not
> just any amino acid will work in even the less constrained positions
> of the molecule.

So? Is that your justification for your claiming that there is no
effective difference between total amino acid numbers and minimum
required amino acids?

> Some of the positions are restricted to certain
> classes of amino acids (i.e., acidic, basic, hydrophilic, hydrophobic,
> etc). Others positions are restricted to within only 2 or 3 different
> amino acids. Some are more loosely restricted to within a dozen or so
> amino acids. Only a few positions can be occupied widely by all 20
> amino acids. Of course, taken one at a time, most positions can be
> changed significantly, but this is not true if too many positions are
> changed at the same time. All of these constraints lessen the density
> of such types of beneficial sequences in sequence space vs. the
> non-beneficial sequences that greatly outnumber them.

Can you demonstrate the mathematics of the above?

> For example, in the June 2003 issue of Nature, Jack Szostak, in a
> discussion entitled, "Molecular Messages" observed that functional
> aptamers and ribozymes (60-90 bases) had a density of only 1 in 10e10
> at best. The sequence space at this level of complexity is only 10e54
> sequences max. On average, a randomly mutating genome that would
> benefit from a ribozyme sequence, not having one already or needed a
> new one, would have to undergo a random walk of 10e10 non-beneficial
> steps before it found one that would have this type of function.

And what does this have to do with the evolution of new functions from
existing proteins in existing organisms? The above is the situation
that you claim you are not actually proposing: the synthesis of new
genes from a randomly generated sequence by a random walk. Yet here you
propose that straw man for modern evolution once again. I presume that
the above was in the "Letters to the Editor" section rather than in the
commentary section.

> Although the average time for an average bacterial colony to cross
> this gap would be only one or two generations, the concept here is
> important. These functions require relatively short averagely
> specified genetic sequences to code for them, and yet they still
> require a minimum number of bases as well as a minimum amount of
> sequence specificity. If they didn't, then the ratio would not be 1
> in 10e10. It would be 1 in 1. Doesn't this make sense to you?

> > You have invented that out of thin
> > air. It is as simple as that. You are simply refusing to look at the
> > evidence, and repeating the same things over and over again.
>
> Hmmmm . . . obviously that is your very clear opinion,

He is not alone.

> but I simply disagree. I actually do think that I am looking at the evidence that
> is being presented in an honest and open way.

You, OTOH, are.

> Certainly you
> understand/interpret the evidence in a very different way than I do -
> and you may be absolutely correct in your views and I could be
> completely off my rocker in my views.

'Could be' is not the phrase I would use. It is a bit too tenuous.

> However, until my dull wit is
> actually enlightened enough to understand things as you do, I honestly
> don't get it. I think you are wrong and that your understanding is
> confused. Of course, I understand and respect the fact that you see
> me as the one that is obviously so confused as to be pretty much
> beyond any sort of hope. Maybe you're right, but that doesn't help me
> now does it? Those who are following these threads must also judge
> for themselves.
>
> > > This is not what I'm saying. You can build up your larger specified
> > > proteins with new types of beneficial functions however you want. No
> > > matter how you try to do it, it will not work beyond the lowest levels
> > > of functional complexity.
> >
> > I have provided you with examples and references to proteins doing
> > exactly that which you say is impossible. Large proteins, thousands of
> > residues long. There have been directed evolution experiments
> > involving RANDOM long proteins (random mixing of sheared DNA
> > fragments) that produced some fairly humongous functional enzymes.
> > Take a look at the reading list I provided for you.
>
> I'm not sure how I can make this any more clear to you beyond the way
> in which I responded last time. I have read about many such
> experiments that you quote and reference here. I fully accept and
> agree with the results and many of the conclusions of the researchers
> in such cases. However, as I mentioned many times before, just

> because a protein is big does not make it more func tionally complex

> (i.e., requiring more genetic real estate at minimum).

Yes. You always have an excuse for why any proposed protein or system
that evolved without changing 'thousands' of amino acids by a random
walk doesn't fit, including the fact that you require the protein/system
to *be* thousands of amino acids in total length (but not be a cascade
or a pathway). There be very few of those systems. That means that God
has been reduced to being the God of the bacterial flagella, but not the
nucleosome, nor the ribosome, nor the immunoglobins, nor the pretty much
anything else. My guess is that you cannot even make a good honest case
for the bacterial flagella using your criteria.

> A much smaller
> protein can be much more informationally complex than a much larger
> protein due to a much higher degree of sequence specificity required
> to achieve a particular type of function.

So how *do* you determine informational complexity? Specifically.
Mathematically. With operational definitions. Not by hand waving
gestures and tossing out the total number of amino acids.

> So, I say again, I am
> interested in knowing the minimum amino acid requirement to achieve a
> particular type of function at a level where it would be beneficial in
> a given life form (not just the ideas of what a lab scientists thinks
> might be beneficial - it must actually be tried out in a living
> thing).

The number of amino acids in the active site of enzymes is necessarily
quite small. Most of the amino acids in an enzyme never comes in
contact with the substrates.

>
> > > But what is the minimum amino acid requirement for this new beneficial
> > > function? Did the new beneficial function really require all of the
> > > amino acids from the large protein or just a small stretch of them?
> > > In other words, you must ask, "What is the shortest amino acid
> > > sequence that would yield this particular type of function working in
> > > this particular way?"
> >
> > Take a look at the literature; there are snippets in the list I gave
> > you to read, but there is so much more if you bother to read a
> > fricking journal now and then.
>
> Oh, a *frickin* journal?! Why didn't you say so? LOL - you know, I
> have read lots and lots of journals - especially those written by
> evolutionists such as yourself. I dare say that I read more
> publications written by evolutionists than most evolutionists in this
> forum,

I daresay you are wrong wrt those who are responding to you.

> and certainly more than most creationists/IDists that I know
> of.

Talk about an easy task! One article would be sufficient for most of them.

> The fact of the matter is that I have failed to find anything
> that significantly challenges my position as of today.

If you chose a position that is as clear as smoke and as hard as jello
and then ask that evolution be the straw man of your choosing, I dare
say that will be quite true.

> You certainly
> haven't referred me to anything that comes even close.

Since no one here is about to claim that some teleologic current system
evolved from a completely random sequence via a random walk, we aren't
even trying to present what you claim we need to show.

> Really now, a
> few things that you have said do sound intriguing, but the one
> observation that you present as being most convincingly against my
> position you say has not been published yet. Ok, well maybe when it
> is published you will notify me of your success. Until then, you have
> come up with a lot of examples of large proteins with very low
> specificity and the evolution of a new function requiring quite a few
> rather specified amino acids in the form of multiple proteins, which
> turned out not to require any new protein structures/sequences at all,
> but simply an increased production of what was already there - nothing
> new evolved at all (your bacterial swarming example). Come one now,
> do you really think such lame examples are enough to hang your hat on?

It was a new function that did not previously exist. Just as eye
crystallins are a new function that did not previously exist. In
neither case did the new function result from a structure that evolved
by a random walk from a random sequence. In each case, it evolved by a
modification in structure (in a regulatory gene or genes). That is how
evolution typically works. By modifying existing structure. Not by
inventing them from scratch. Or adding amino acids one at a time.

>
> > Your question is meaningless, but no matter how many examples and
> > references I provide, you just keep repeating it over and over.
>
> Hmmm . . . I don't seem to be the only one who likes to repeat himself
> . . .
>
> > Very
> > well, let us try a different approach. You choose a large protein -
> > any large protein over one thousand residues. Just pick one. Then YOU
> > tell ME how many of its amino acids are REQUIRED for its function, and
> > WHY you think that they are all REQUIRED.
>
> Ok, take the flagellar system of motility. At minimum such a system
> requires 20 or so different types of proteins working together at the
> same time. The amino acid total involved is over 6,000 fairly
> specified amino acids working together at the same time.

How did you determine this number? Is 6,000 the total amino acid
number? If not, how did you reduce the number? What criteria did you
use? There are some proteins in this group that look like modified
duplicates of others. Did you reduce the "minimum number"
appropriately. Some of them are similar to proteins that are not
involved in motility, but in protein export. How did you reduce the
number of amino acids to account for the possibility that some of the
proteins may simply be modified from pre-existing proteins? How did you
generate your number of 6,000. By waving your hands frantically?

> If you have
> less than this minimum, you will not get the flagellar motility
> function - period. If you reduce the flagellar system below this
> minimum part requirement, the beneficial motility function will vanish
> completely.

I still don't see where you actually determined this number as the
"minimum required amino acid sequence".

>
> Oh, but what about the secretory function (TTSS)? Well, the secretory
> function may still be maintained since it requires far fewer specified

*Far* fewer? How many fewer? Does the *difference* between TTSS and a
motile flagella equal 6000 amino acids? Can you put a number on this?
How did you determine this number? Where is it to be found in the
literature? Or are you merely waving your hand desperately?

> amino acids working together at minimum in order for its beneficial
> function to be realized. The gap between the minimum requirement of
> the secretory pore and the motility function of such a secretory
> structure to be realized is involves over 3,000 fairly specified amino
> acids.

Where did you get this number? The same foul air as the other numbers?
Which half of the motility system's "minimum required amino acids" are
not required for protein secretion? Stop blowing smoke and start
telling us where you get these numbers.

> There are many who have proposed various stepping-stone pathways
> across this gap, but none of these proposed pathways has been
> supported by experimental testing/demonstration. Not even one of the
> proposed steps has been demonstrated to evolve - not one. Many
> scientists even challenge the idea that the TTSS system was actually
> the precursor of the motile flagellum. The current views of most is
> that the TTSS system evolved independently and some actually suggest
> that the TTSS system evolved from the fully formed motile flagellar
> system (Nguyen et al. 2000)

Yada. Yada. Yada. If TTSS evolved independently, that would be a death
blow to your explanation. If the current TTSS were a secondary
evolution from a motile flagella, that would hardly be proof that the
reverse were impossible.

>
> > > Not if you consider the minimum part requirement, not the maximum junk
> > > you can include with this minimum. The extra junk residues that are
> > > not required for minimum function do not count toward the calculation
> > > of minimum functional complexity.
> >
> > There is no "minimum part requirement". See references in the original
> > post.
>
> Ok, I looked at your references. Which one was it again that showed
> how all protein functions can be had by any amino acid sequence of any
> length and order?

Who is arguing that? Not anyone here. But what I cannot understand is
how, given the undeniable fact that active sites are necessarily going
to involve only a small (and often a very small) number of amino acids,
one manages to wind up with minimum required numbers of amino acids that
are indistinguishable from total number of amino acids? Even in the
cases of enzymes that you agree have evolved new or modified function by
changing a single amino acid?

> > > Hmmmm, but isn't architecture dependent upon amino acid order? The
> > > greater the specificity of the required architecture, the greater the
> > > specificity of amino acid order - right?
> >
> > Wrong.
> >
> > One, the overall composition of a segment is what determines
> > architecture (see Chou-Fasman rules).
>
> Doesn't the composition of a segment require a certain degree of amino
> acid specificity?

No. See Chou-Fasman rules. It does not require *sequence* specificity.
And sequence specificity based on a particular amino acid being
*invariant* is what you are assuming when you make your calculations
using the 1/20 as a probability factor. Hint: Even though your
calculation would still be fatally ignorant using more correct values,
the probability of an amino acid which can be changed to any other amino
acid but one would be 19/20, not 1/20. The probability of amino acid
which can be changed to any similar amino acid (say also hydrophilic) is
actually higher than the proportion of hydrophylic amino acids out of 20
because, by the nature of the genetic code, most single mutations
produce similar amino acids rather than disparate amino acids. That
means you are not even calculating your bogus math right. Instead you
are assuming that each of the "minimum number of amino acids" is
necessarily invariant and also assuming that there is little or no
difference between your "minimum number of amino acids" and the "total
number of amino acids" in a protein. The reason, of course, is that
making such silly assumptions generates the large numbers you want. In
short, your entire calculation system is GIGO.

>
> > Two, the larger the protein,
> > less constraint to its exact structural composition (less constraint
> > on how extensively it can be modified even *structurally*).
>
> Again, many of your examples of large proteins are not required to be
> nearly as large as they are in order to achieve a beneficial level of
> the type of function that they currently perform . . . isn't this
> true? Also, we are not talking only about single protein functions
> here.

Apparently, like typical proteins and cascades and pathways and most
multi-protein enzymes, large single proteins also don't meet your
criteria for being unevolvable. There simply are no single proteins
that are both large enough to have more than 6000 amino acids that
aren't also composed of internally repetitive domains. So that means
your "impossible to evolve" systems are necessarily limited to a very
few sub-structures in a cell or the entire cell itself. Can you name
*all* the structures in a cell that *do* meet your criteria for being unevolvable?

> You must also consider those functions that require multiple
> smaller fairly specified proteins working together at the same time -
> as in the flagellar motility system.

Good. Lets start with this system. First, what is the minimum number of
parts and how did you calculate that number?

> The total number of amino acids
> involved in such a case is rather great, as is the level of overall
> amino acid order/specificity.

Where is your evidence for this? How did you calculate this? Are there
no proteins which show similarity of sequence which would require the
reduction in the number of amino acids? My assumption, of course, is
that neither you nor *any* creationist or IDiot have done *any* real
analysis of the bacterial flagella. All you have done is to point out
that it is big, involves a number of proteins, and has a current (which
you mistake for teleological) function that requires most of the parts.
You (and the IDiots) then *assume* that every amino acid in every
protein you can reasonably rope into the 'system' is necessary for
funtion and must be invariant for function (which you *assume* can only
be motility). That allows you to calculate a big number that would be
required if evolution worked by taking a random sequence of that length
and had to reach the single possible invariant sequence by a random
walk. That that is not how evolutionary biologists think the bacteria
flagella evolved doesn't seem to matter to you or the IDeologists who
repeatedly toss out the large numbers.

> So you see, you can get a larger
> protein *system* of function without a corresponding decrease in the
> percentage/level of constrained positions.

Where is your evidence in support of this? I keep not finding it. Is
it written in the Bible Code within your answer?

> > > These are difficult too, but some larger proteins are also fairly
> > > specified.
> >
> > Examples, please. If you actually bother to *look*, you will be
> > suprised.
>
> Oh really? I would suggest to you that the lactase function requires
> a higher level of information/genetic real estate than does the
> function of the mammoth Titan protein. Then, going beyond this, the
> genetic information required to code for the flagellar system is far
> beyond either one of these. What do you think of these oft-repeated
> examples?

Not much until you put some meat upon these bare bones instead of the
empty smoke you cover them with.

> > > And, you forget, that I'm not just talking about larger
> > > single proteins which often loose specificity requirements, but
> > > about systems that require multiple smaller proteins working
> > > together at the same time (as in various bacterial motility
> > > systems like the flagellar system). Such multiprotein systems
> > > require many fairly specified proteins working together at
> > > the same time.
> >
> > Again, LOOK at the references I gave you (and also the examples in the
> > very first post I sent in response to your theory). I have both given
> > you examples of exactly such systems arising in lab.
>
> Where? List your best example here again. The only examples that you
> have listed of multiple proteins working together at the same time
> involve no more than 200 or 300 amino acids total. Please, what did I
> miss here? Oh, and don't try and float your M. xanthus swarming
> example again at this point. This "evolution" did not involve the
> production of anything new structure/protein sequence at all.

Doesn't change in the regulatory sequences of genes count as
evolutionary change either? Apparently the only thing that counts as
'evolutionary change' to you is if one started with a ranodm sequence
and produced a new function that required every amino acid to change to
a pre-specified different result by a random walk. That is, your idea
of what 'evolution' is and how it works not remotely similar to the idea
of 'evolution' that everyone talking to you has. You keep pulling that
straw man out *despite* the fact that you have been given example after
example of how evolution really works -- via modification or change
rearrangements of pre-existing functional structures. [Nylonase is the
rare counterexample or exception that probes or proves this rule.
Nylonase is also probably how most of the current basic functional
motifs first arose. It shows that random sequence can generate
function.] That most evolution involves modification or rearrangement
of existing protein motifs (there are probably less than a hundred
distinct motifs) is not surprising. Most of an organism's needs involve
small organic molecules created by living creatures or simple ions like
SO4 or PO4 or involve binding to other proteins or other organics.

> It was
> simply an increased production of exactly the same protein matrix that
> was already being made. Come on now, did I miss where you actually
> presented something better?

Again, it *was* a change in *function*. The very point is that
evolution does not need to change *structure* significantly (the change
in the regulatory sequence is a change in structure) to produce a change
in *function*. This is entirely relevant to the case of the emergence
of the bacterial flagella's motility function. It, too, was likely the
result of a small change in structure in a pre-existing 'system' (with
TTSS being the obvious earlier *proximal* functionality -- other
structures would be proximal to the TTSS structure) and not due to a
process where one starts with a random sequence of x number of proteins
with a total of 6000 or whatever amino acids and reaching the end result
by a random walk. It is probable that the proximal step involved a
co-opted function of another protein. Your straw man view of evolution,
no matter how you try to disguise it and deny that that is what you are
proposing is exactly what your calculation of the odds says. Of course,
you may honestly be ignorant of that fact, despite everyone else
pointing it out to you.

>
> > AND all of the
> > above mentioned applies to them too; while proteins are "working
> > together", their individual amino acids are not.
>
> Uh, I might be a bit too slow to understand how you could possibly
> have just made this statement. So, perhaps you could explain to me
> your sizzling rational for stating that proteins can work together but
> their subparts (amino acids) having nothing to do with how they work?

Protein-protein recognition is almost always the result of a few sites
on the two proteins involved and not a consequence of binding at many
sites. The latter would not be nearly reversible enough for many
protein-protein interactions. When cells want protein-protein
interaction to be quasi-irreversible, they tend to use cross-linkage by
cysteine. One does not need a large amount of amino acid specificity
involving 'thousands' or 'hundreds' or even 'tens' of amino acids to
generate even the protein-protein specificity of the immune system.
Very often the protein-protein recognition is based on things like
patches of hydrophylicity or hydrophobicity rather than on specific
amino acids at all.

> Haven't you ever heard of "emergent functions"?

Yes. But what you describe below is not an "emergent function". A fin
useful for swimming that has a bone and muscle mechanism like that seen
in the lobe-finned fish becoming a structure useful for walking is an
example of a new function (walking on land) emerging from a system
originally use for a different function, swimming. Similarly, the air
bladder of fish used for bouancy is an emergent function of a lung
evolved for a different function, extracting O2 from the atmosphere.

What you describe below is a developmental sequence.

> Without the
> underlying order of interacting parts, the parts themselves will not
> work together properly. For example, consider the fact that in order
> to walk the legs must be attached and interact properly with the torso

> of the body. However, the leg subparts must also be in proper order as

> well or the leg simply will not work right even if it is attached
> properly to the torso. If the there is something wrong with the actin
> or myosin of the muscle groups, they will not contract properly and
> the leg will not respond to the nervous signal from the brain. So, in
> reality, all levels of order are required and are thus working
> together at the same time in a system whos highest levels of emergent
> function require the simultaneous interaction of the highest level
> parts.
>
> > >Such multiprotein functions are quite specified.
> >
> > Wrong.
>
> Nice "just-so" statement, but it comes across a bit hollow. You
> simply haven't backed yourself up in a convincing manner.
>
> > > LOL - oh really? What then, is the shortest cytochrome c sequence
> > > that would have the cytochrome c type of function?
> >
> > LOL indeed! Read your question: what is the shortest cytochrome c that
> > is a chytochrome c? :))) Answer: if it is half the size, and different
> > in mechanism, we will call it something else; which is why cytochrome
> > c is cytochrome c.
>
> Obviously true. If it is half the size and has a different function,
> it will not have the type of function of a cytochrome c sequence

> (i.e., a rather specified type of electon transport protein).

>
> > But let's answer the implied question: what can perform the function
> > of cytochrome c, i.e. electron transfer function? Heme group on its
> > own is a decent electron transporter, all you need is any kind of
> > protein to bind it and prevent two hemes from coming in contact with
> > each other. Other then that, check flavoproteins, copper proteins,
> > quinoproteins, molybdenium-containing hydroxylases...
>
> Oh, so just about any protein sequence would work just fine as an
> electron transport protein in place of a cytochrome c sequence in
> living creatures?

No. But there are clearly more than one type of protein that can bind
heme. The way that cytochrome c works is to basically hide most of the
heme except for the one end where the electrons enter and leave. This
means that most of the protein is in contact with one or the other
surface of the heme. A very few amino acids are strongly linked to the
heme, but most are simply required to be hydrophobic rather than
hydrophilic. As a percentage of the total protein, and unlike most
larger enzymes, more of cytochrome c is in contact with the substrate
and constrained by that. That, in fact, is why cytochrome c is useful
for deep phylogenetic analysis -- more of its amino acids are
evolutionarily constrained and change only rarely than amino acids in,
say, hemoglobin alpha chain. In general, the larger the protein, the
smaller the percentage of the protein in contact with its substrates and
the less evolutionarily constrained it is, as evidenced by the rate of
change in sequence (generally in the non-constrained sequence) over
time. It is, in fact, this discovery that makes sequence comparison so
useful as a way of discovering which amino acids are involved in the
active site. Most evolutionary change in sequence is selectively
neutral change and not selectively beneficial change. That is why a
human protein can often be substituted for a Drosophila protein that
performs the same function.

> Has this assertion been supported by any sort of
> relevant demonstration in a living organism? The fact of the matter
> is that many scientists disagree with you here. For example, Yockey
> has suggested (and Ian Musgrave brought this estimate to my attention)
> that the total number of all functional cytochrome c sequences is
> around 10e60 (and this is out of 10e100 possibilities). The fact of
> the matter is that the vast majority of possible protein sequences of
> similar size will not work at all as a replacement for cytochrome c
> since they just don't have a similar enough functional potential to be
> beneficial to any given life form as an electron transport protein.

Yet the relevance of this is not clear. The cytochrome c's in current
organisms did not get invented anew by starting with a random sequence
in that organism. Current organisms have cytochrome c by virtue of
common descent from an organism which did evolve a cytochrome c. We do
not *know* how that initial cytochrome c arose because it arose so early
in the history of life on the earth, but it is likely that it arose from
an initial heme binding protein that had *some* capacity to direct
electron transfer to a specific edge on the heme and prevent heme-heme
face to face interaction. Mutation and selection for improved function,
to the extent that it benefited the organism to prevent heme-heme face
to face interactions and favored the organism to have edge to edge
interactions, would quickly (in geological terms) lead up the path to
one possible optimal heme binding protein for those functions. Vertical
transmission and neutral change would then produce the present observed pattern.

>
> > You keep pulling huge numbers out of nowhere, while I (and others)
> > keep giving you examples of lactase function evolving in the lab. If
> > it worked your way, that couldn't happen. I also gave you examples of
> > short polypeptides acting as beta-galactosidases.
>
> In-vivo examples? You have made assertions that your laboratory
> experiments would actually have some sort of correlation in some
> potential life form, but you have yet to show that anything shorter
> than 400 or so amino acids would actually work in a beneficial manner
> as a lactase enzyme in an actual life form.

The point remains that the number of actual amino acids involved in the
*function* is not 400. Given the number 400, as the minimum number of
amino acids needed to generate a betagalactosidase *function*, what are
the odds that another protein would exist in E. coli that was one
mutation away from that function? And why didn't it have to have the
400 other amino acids that the original beta-galactosidase enzyme had?

> Please, you have try out
> your assertions/experiments where they would really mean something
> (i.e., in an actual life form). You have only made it through the
> first part of the scientific method. You have a fine hypothesis and a
> fine prediction based on this hypothesis. Now, all you need to do is
> test your predictions and see if they hold up in real life. Once you
> have done this, get back to me.

So tell me what the number 400 has to do with beta galactosidase activity?

> > > Take cytochrome c for example. The shortest functional cytochrome c
> > > sequence that I know of requires a minimum of around 80 or so amino
> > > acids. Of these about 20-30 are pretty much "invariant". Many of the
> > > others (around 25) are highly constrained to within just 2 or 3 amino
> > > acids having similar chemical natures. For example, Lys can change to
> > > Arg, Glu can change to Asp, Ala can change to Val, and Val can change
> > > to Ile etc. Surface AA residues can vary more than those in the
> > > interior. Those in key positions involved in the functionality of the
> > > protein are more constrained and can not vary without disturbing the
> > > 3-D shape of the protein and therefore its functionality. Already we
> > > are up to around 40 to 60 highly specified amino acids.
> >
> > Are we indeed? Take a look at the genetic code now. Look at how many
> > combinations code for Val, and how many for Met. Think on it a bit. It
> > screws up your probability calculations mightily.
>
> Oh really? How so? There are 4 combinations for valine and one for
> methionine. There are also six each for leucine and arginine. On
> average though, there are 3.2 codons per amino acid. Say we have 60
> highly specified amino acids. On average, the likelihood is that
> these amino acids will be a fairly even mix with an average of 3.2
> codons per amino acid. So, averages still work out rather nicely
> suggesting that getting a particular amino acid position right in a
> highly constrained position via random chance is still around 1 in 20.
> At the very worst the average might be as high as 1 in 15 or so, but
> this does not "mightily" mess up my probability calculations at all.

Notice that this means, despite your repeated denials, that you are
making the assumption that evolution works by a random walk from a
random sequence to possible useful results. Otherwise, your numbers
don't make any sense. The only way that the number generated as the
denominator can be the product of x number of 1/20 probabilities is the
assumption that the useful results are generated by a random walk
through all of possible sequence space for a sequence x amino acids long
(since there seems to be no difference between your "minimum number of
amino acids" and "total amino acid length" in any example you want to
use -- you only claim a difference between these numbers when you don't
want a protein to be used as an example, such as all the large single
proteins). Any time your denominator is all of sequence space, your
assumption is that evolution works by a random search through all of
sequence space. That is not the case for the evolution of any of the
systems we are talking about. All of the systems we are talking about
evolved from pre-existing systems with functional utility to living
organisms by modification or rearrangement of those pre-existing
systems. This is a much, much smaller subset of sequence space and is
not a random subset either. It is not possible to calculate the
probability of a new function on the basis of simplistic total number of
amino acids *unless* you are assuming that the starting point is some
random sequence.

> In fact, I find it rather amusing that you would even suggest such a
> thing. You are the very first one to challenge me in this way. I
> must give you points for creativity here though.
>
> > Oh, I'll grant you that there are many small, very constrained
> > proteins (histones are, again, the textbook example). But whenever I
> > give you an example of a small protein evolving, you laugh at me and
> > tell me it's to small. Well, let me laugh here, and tell you that
> > cytochrome c is too small. Why don't you try to analyze a large
> > protein in the same way? Take lactase: how much can its structure
> > change (hint: take a look at various 3D structures of various
> > lactases, and compare sequences). You will find that the larger the
> > protein gets, less constrained it is. And, in addition, how do you
> > know how ancient cells performed the function of electron transport?
>
> I have asked this question of several apparently well-informed
> individuals. Some, such as Ian Musgrave, have actually given me an
> understandable response. Ian suggested to me that the total number of
> lactase enzymes in sequence space was probably around 10e100.

That number is, of course, irrelevant unless you have a denominator.
The use of total sequence space as a denominator is what is bogus, since
the starting point is not a system composed of random sequences. The
starting point is an organism with a very non-random subset of total
sequence space. In particular, it is a subset of total sequence space
that includes sequences and protein motifs of functional utility. In
addition to the presence of many functional motifs, the subset of
sequences in real organisms does not include, for example, proteins that
do not form, by self-assembly, a *particular* 3-D structure (i.e.,
wildly floppy proteins are not often seen in organisms). Moreover, this
non-random subset of sequence space is slightly to more significantly
*different* in every single non-clonal *organism* (and, depending on
numbers, even in clonal organisms) that exists on the surface of the
earth. Note that I said *organism* and not *species*. That is because,
for example, you and I, similar though we may be, are not genetically
identical, as a rapid exchange of organs without using immunosuppression
drugs would probably show.

> Of
> course, this was before he suggested to me that the minimum lactase
> sequence probably required more than 400aa - which would result in a
> sequence space of well over 10e500. From his more resent posts, I'm
> sure that Ian would markedly increase his estimate so that the ratio
> would be closer to 1 in 10e12 or so.

And every one of these numbers is irrelevant in the way you use them.
GIGO is still GIGO.

> However, I have problems with
> this ratio considering that at such a high ratio I would expect that
> all types of bacteria, with an average sized colony of 10 billion or
> so individuals, would quickly evolve a function that had such a high
> ratio of sequences in sequence space (a few months at most).

Could you show us your calculations? Didn't do any beyond the
simplistic calculation of total sequence space, eh?

> The
> problem here is that Hall's experiment with E. coli showed that the
> lactase function did not evolve in ebg negative E. coli in tens of
> thousands of generations. This tells me that the lactase function
> requires a bit more sequence specificity than many in this forum have
> recently suggested to me.

Again, the fact remains that, out of a few thousand genes, two were
within a single nucleotide change of beta galactosidase activity. And
another was somewhat further away. That is an indication that sequences
with potential for beta galactosidase activity are at least sporadically
available in some bacteria. *If* a bacteria were to have an absolute
need for such activity, at least *some* bacteria would be able to evolve
it. The others would go extinct in that environment, but could continue
to exist in other environments. If there were currently no bacteria
able to cleave lactose, then lactose would be an unuseable carbon source
(just as cellulose is an unuseable carbon source to humans). Genetic
drift could occur in such populations for generation after generation
after generation in genes in these populations without producing any
protein surface that significantly lowers the energy of activation of
lactose cleavage -- the only cleavage is due to spontaneous cleavage.
Lactose would remain a basically unuseable resource. At some point,
however, *because* the active sites of beta galactosidase function is
small, some protein will, by chance, produce a surface that lowers the
energy of activation significantly. Because that bacteria, in this
lactose rich environment, will have a selective advantage over its
peers, it will increase to a greater extent. Moreover, to the extent
possible (there may be a minimax problem if the protein site involved
also has other utility), other mutations can now build on this original
event. Eventually, this will lead to a protein (or a duplicate of the
original to allow specialization) that is optimized to cleave lactose.
The clone that first invents such a protein will have an advantage in
certain environments, those with lactose. The key is the initial event
that produces a 'significant' lowering of the energy of activation of
lactose cleavage. That event need not be optimal lowering. It only
needs to be 'significant'. Once that happens, the ratchet of selection
for improvement of function is activated. Until that happens, there is
nothing to build on and lactose will be essentially an unuseable
resource. But note that *both* a genetic change *and* a particular
environment need to be present. One without the other doesn't produce a
functional beta galactosidase by selection.

> > Yes, I will admit that you can't change too many letters in a
> > paragraph without changing the meaning. This is why language is a poor
> > example for the way proteins work. In proteins, I can change ALL the
> > letters in a paragraph, except for one or two, and as long as those
> > two are in the same approximate relation to each other spatially, it
> > will still work.
>
> You overstate your position. For many types of single protein
> functions there are a handful of proteins that cannot really change at
> all. These are called, "invariant" amino acid positions. At this
> point we are both in agreement. But, after this point you overstate
> your position. You seem to be suggesting that *all* of the rest of
> the amino acid positions can freely change without any real damage to
> protein function. This is simply not true, or at best it is a huge
> understatement of the actual constraints involved with the rest of the
> amino acid positions. Many other amino acids beside those that cannot
> change at all can only change between 2 or 3 very similar amino acids.
> Many more can change only between certain classes of amino acids as
> noted above.

Yet your calculations completely ignore this fact. Just as it ignores
the fact that total sequence space is irrelevant to all current
evolution.

> Only a very few can change easily between all 20 amino
> acids. A limitation to class or to a handful of amino acids is still
> a significant limitation. Each limitation decreases, in an
> exponential fashion, the density of beneficial sequences with this
> type of function in sequence space (i.e., it markedly reduces the size
> of the gray areas around each "dot").

The problem is that the active site remains very small. And we do not
generate new functions from all of sequence space. Rather we do it from
a specific (and functional) subset of sequence space. We may be able to
roughly calculate the number of sequences *related to* the current
sequence that could have a certain minimal level of functional utility.
But what we cannot predict is how many completely different sequences
and 3-D structures could also put the active site amino acids in the
proper positions. The Hall experiment shows that there are at least two
entirely different sets of structures that can be no more than a single
amino acid change away from having beta galactosidase activity.

>
> Anyway, this is all I have time for today. Thanks again for your time
> and the great effort (despite a lot of significant frustration I'm
> sure) that you must put into your posts. I really do admire your
> sincerity and your energy.
>
> Sean
> www.naturalselection.0catch.com

r

[Sean Pitman]

Howard Hershey <hers...@indiana.edu> wrote in message news:<3FE8BC16...@indiana.edu>...

> Sean Pitman wrote:

> > I read your original message, in its entirety, before I responded the
> > first time. However, responding to everything in particular takes a
> > lot of time, especially since many of the ideas you were presenting
> > were directed against a straw man version of my position.
>
> It is not a straw man to say that your position is that you cannot or
> will not distinguish between the total number of amino acids in a system
> and the 'minimum number of amino acids'.

The "minimum number" of fairly specified amino acids required for a
particular type of function to be realized at a beneficial level, in a
given life form, in a given environment, IS the "total number" of
amino acids needed to achieve that particular function and all other
functions within that level of functional complexity.

Then you have clearly, and I dare say deliberately, misunderstood my
position. I have clearly said over and over again that I do not
consider every amino acid position in a protein to be "invariant" -
not at all! I agree with everything that you said in the above
paragraphs. Your problem though is that you left out a very important
concept from your discussion here. You seem to have forgotten about
"partially variant" amino acid positions. Even though most amino
acids positions in a given protein with a particular function, such as
the lactase, cytochrome, or histone function, are variable, there are
different degrees of variability for each amino acid position. Like
you said, only a very few of these positions are completely invariant.
However, many more can only be changed between 2 or 3 positions. In
the cytochrome c protein, for example, at least 80 amino acids are
required at minimum to achieve this kind of function. Of these, at
least half of the amino acid positions are constrained to within 2 or
3 amino acids. Another 20 positions are restricted to within certain
classes of amino acids (hydrophilic, hydrophobic, basic, acidic, etc).
Only a very few positions can vary widely without some sort of
restriction. Based on these limitations, the total number of
sequences in sequence space that can have this type of function have
been estimated and quoted by some in this forum to be around 10e60.
The problem with this estimate is that the minimum sequence space at
this level of complexity (minimum of 80aa working at the same time in
a fairly specified order) is over 10e100. This creates a ratio of
potential cytochrome c sequences vs. non-cytochrome c sequences of
less than 1 in 10e40.

Histone sequences are even more restrained. Lactase sequences are
certainly less restrained, but they are still more restrained than a
handful of invariant amino acids would suggest. They have many
partially variant positions that are none-the-less restricted to one
degree or another. There is also a minimum amino acid requirement for
lactase enzymes to be beneficial - seemingly well over 400aa at
minimum. Of these 400aa only a handful are invariant, but many of the
rest are still restricted to a small group of 2 or 3 amino acids.
Many more are restricted to certain classes of amino acids. Each
restriction placed on an amino acid position lessens, in an
exponential fashion, the density of those sequences with this type of
function in the minimum sequence space at this level of functional
complexity. At such levels of specificity, those functions that
require longer and longer minimum amino acid sequences become
exponentially rarer and rarer in sequence space.

Obviously many very large single protein sequences are even less
specified than a lactase sequence. Consider that many lactase
sequences use over 1000 amino acids per subunit. However, this is not
the minimum requirement, which seems to be less than half this number
at around 480aa. The same thing is true of mammoth proteins, such as
the Titan protein, that most likely require far fewer amino acids to
achieve a minimum level of beneficial function than are normally used.
However, there are functions that do require thousands of fairly
highly specified amino acid positions working together at the same
time.

Multiprotein functions often require fairly high specificity of amino
acid positions in each of their individual protein parts. Since each
protein part works together at the same time to achieve a unified
function, each of the amino acids in all of the protein parts are also
working together at the same time to achieve this unified function.
The minimum amino acid requirement for many of these multiprotein
functions is in the thousands and even tens of thousands. For
example, the minimum number of different kinds of proteins required
for the flagellar motility function seems to be more than 20. On
average each of these proteins requires a minimum amino acid sequence
of more than 200 to 300aa. At minimum then the motility function of a
flagellar system requires at least 4,000 to 6,000 fairly specified

amino acids working together at the same time.

Evolving a new type of function within such a level of minimum
informational complexity would simply take zillions of years to
achieve via mindless processes alone. You have failed to explain how
any functional or non-functional starting position could achieve such
functional diversity at such levels of informational complexity that
we see today in all life forms (via mindless evolutionary processes
alone).

This is all I have time for today.

Sean
www.naturalselection.0catch.com

[dkomo]

Sean Pitman wrote:

[snip]

> Evolving a new type of function within such a level of minimum
> informational complexity would simply take zillions of years to
> achieve via mindless processes alone. You have failed to explain how
> any functional or non-functional starting position could achieve such
> functional diversity at such levels of informational complexity that
> we see today in all life forms (via mindless evolutionary processes
> alone).
>

No, it wouldn't take zillions of years because it uses the WEASEL
principle described by Richard Dawkins. Anyone who has played around
with a WEASEL program knows how powerful are the operations of random
mutation followed by selection and how rapidly such a program can
converge to even a long target string. In a computer simulation of
the build-up of your target protein starting from strings of random
amino acids, the convergence wouldn't even take zillions of
nanoseconds.

I described such an experiment earlier this year using RNA sequences
instead of amino acid sequences. The principle is exactly the same.
Here's a repost:

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Subject: Playing with WEASEL
Date: Tue, 04 Mar 2003 12:11:38 -0700
From: dkomo <dkomo...@cris.com>
Newsgroups: talk.origins

I thought it would be interesting to see how rapidly a WEASEL program
converges to a target phrase that is a segment of an RNA sequence
rather than the stock phrase "Methinks it is like a weasel".

I went to Ian Musgrave's web site, which contains links to a number of
WEASEL programs:

http://www.health.adelaide.edu.au/Pharm/Musgrave/essays/whale.htm

Being short on time, I didn't want to download and play around with
different stand-alone programs to see which ones ran the fastest and
could handle very long target strings. I was fortunate to quickly
find an excellent Java applet which runs directly in a web browser and
seems to be very fast:

http://home.pacbell.net/s-max/scott/weasel.html

On the stock phrase "Methinks it is like a weasel" I got these
results:

Tries <= 76800
Best Critter Methinks it is like a weasel
Score (0 is best) 0

Checked 76800 critters in 20 seconds == 3840 tries/sec.

This WEASEL applet has the exemplary ability to be able to input goal
phrases of virtually unlimited length and to run with those phrases.

On the Musgrave site I found this RNA fragment:

GUAGAACGAUAGUAACUGAUUAAAGAAUAAGCAUAAUGGCAUGCACUGGAA

The codons in this sequence specify amino acids whose standard
abbreviations spell out "VERY LIKE A WHALE". I took the basic
sequence and duplicated it twice to produce a sequence that was 153
letters long:

GUAGAACGAUAGUAACUGAUUAAAGAAUAAGCAUAAUGGCAUGCACUGGAAGUAGAACGAUAGUAACUGAUUAAAGAAUAAGCAUAAUGGCAUGCACUGGAAGUAGAACGAUAGUAACUGAUUAAAGAAUAAGCAUAAUGGCAUGCACUGGAA

I then fed it into the textbox Goal Phrase of the applet and started
it running. The convergence was amazingly rapid considering the size
of the goal phrase. After about 6 minutes of run time I had:

Tries <= 308224
Best Critter

GUAGAACGAUAGUAACUGAUUAAAGAAUAAGCAUAAUGGCAUGCACUGGAAGUAGAACGAUAGUAACUGAUUaAAGAAUAAGCAUAAUGGTAUGCACUGGAAGUAGAACGAUAGUAACUGAUUAAAGAAUAAGCAUAAUGGCAUGCACUGGAA
Score (0 is best) 2

The score indicates that only two letters of this partial result are
incorrect. The applet seemed to overflow its result box so I couldn't
tell if it eventually converged all the way. I stopped it after I got
the above result.

What's interesting is that the WEASEL program as described by Dawkins
in _The Blind Watchmaker_ is nothing more than a crippled genetic
algorithm. The basic algorithm followed by most of these WEASEL
implementations is:

1. randomly generate an initial population of character strings

2. determine the fitness of each string by computing the difference
between it and the target string

3. select the fitest string and generate a new population from it,
each mutant string differing from the parent in one randomly selected
letter; each letter differs from its original by one position earlier
or later in the alphabet

4. iterate steps 2 and 3 until the difference reaches 0

An ordinary genetic algorithm would allow more than one string to live
into the next generation and would add a crossover operation between
two randomly selected strings to produce progeny. Moreover, the rate
of mutation is normally set quite low in order not to disrupt fit
strings too much. Finally, an "elitest" strategy is often employed
where the fitest string in a particular generation is *always*
preserved to the next generation.

If WEASEL had employed a full genetic algorithm, convergence to the
target would have been much more rapid. Also, the WEASEL program I
used was handicapped for the RNA sequence because it searched the full
space of 68 characters (upper and lower case plus punctation
characters and spaces), rather than just the 4 letters G, C, U and A
for RNA. And then consider that a Java applet is dogshit slow because
it runs on a virtual Java machine. Compiled C would run much faster.

Still, I'm impressed with how rapidly just simple mutation and
selection can generate results. I think it's a proof of principle
that molecular evolution can come up with new genes with surprising
ease.

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

--dk...@cris.com

[Frank J]

seanpi...@naturalselection.0catch.com (Sean Pitman) wrote in message news:<80d0c26f.03121...@posting.google.com>...
(snip)

>
> Yeah yeah . . . The nested hierarchy argument. This is so common and
> predictable among you evolutionists. When you can't demonstrate the
> actual evolution of novel functions beyond the lowest levels of
> complexity, you revert to the nested hierarchy argument. This is
> nothing more than an argument that assumes that shared similarities
> must mean common evolutionary ancestry.

And as you know, Michael Behe has no problem with that.

> The problem with this
> argument is that shared similarities do not support the notion of
> common descent from a single common evolutionary ancestor over the
> notion of common design.

This is so common and predictable among you anti-evolutionists. When
you can't support "independent abiogenesis," you revert to the
hopelessly ambiguous "common design" terminology.

> Shared similarities are explained equally
> well by both positions.

But unless you specifically define "common design" as "independent
abiogenesis" there is no "both positions." The best alternative you
can do is Behe's.

> In order to rule out my position of
> intelligent design, you must show that evolutionary mechanisms can
> explain the differences as well as the similarities found within and
> between various life forms.

Actually that can't rule out "intelligent design" for the simple
reason that nothing can.

> The similarities can be equally explained
> by both of our respective positions. The differences, on the other
> hand, are a different story.
>

Does that mean that we can expect a hypothesis of "independent
abiogenesis" based on the differences, or are we back to the same old
word games?

(snip)

[Howard Hershey]

Sean Pitman wrote:
>
> Howard Hershey <hers...@indiana.edu> wrote in message news:<3FE8BC16...@indiana.edu>...
> > Sean Pitman wrote:
>
> > > I read your original message, in its entirety, before I responded the
> > > first time. However, responding to everything in particular takes a
> > > lot of time, especially since many of the ideas you were presenting
> > > were directed against a straw man version of my position.
> >
> > It is not a straw man to say that your position is that you cannot or
> > will not distinguish between the total number of amino acids in a system
> > and the 'minimum number of amino acids'.
>
> The "minimum number" of fairly specified amino acids required for a
> particular type of function to be realized at a beneficial level, in a
> given life form, in a given environment, IS the "total number" of
> amino acids needed to achieve that particular function and all other
> functions within that level of functional complexity.

You may *think* you are saying something here. You are not. The *only*
relevant number in evolution is the number of changes needed to convert
an ancestral or previously existing protein with a number of selectable
*functions* (note the plural useage) to some 'new' selectable
(selectable for the new functionality that is the next selectable step)
version of that protein. This new protein can have fewer (new functions
can arise by loss of primary function and focus on secondary function),
more (a new site or motif gets added to a protein, or a new protein
associates with a pre-existing complex changing its function), or
quantitatively different amounts of the same functions, or modification
of one function to a slightly different one (say binding one sugar to
binding a slightly different one better). One cannot determine this
number by looking *only* at the end result protein (which is what you
are doing) and how many amino acids are needed for that protein. One
must *also* look at the protein(s) from which it evolved. Your repeated
failure to do this means that your numbers are irrelevant GIGO that only
have meaning *if* one assumes that the current protein evolved
*directly* and *immediately* from a random sequence by a random and
functionless (without intermediate functionality) walk.

*Your* mathematical argument (which you then deny is what you mean when
it is put in words) says that *any* currently existing protein, the beta
globin of hemoglobin, for example, arises by evolution from a *random*
sequence of amino acids or some *random* protein sequence in some
organism. All the functions of beta globin must arise *indepedently*
and in complete *isolation* starting with this *random* sequence and the
evolution of beta globin procedes by a *random* walk with no
intermediate stages of *any* utility, until it just happens to land upon
the beta globin sequence of hemoglobin with its function of binding heme
and affecting the oxygen pressures that allow the heme to bind or
release oxygen and, also, the function of binding to alpha globins to
form the heterotetrameric hemoglobin. That is precisely what your
mathematical ratio gives: The odds of a functional beta globin sequence
(say the embryonic beta globin which has a 'new' or 'different' function
than adult beta globin) as a fraction of total sequence space for a
protein of that length (again, I see no significant difference between
your 'minimum amino acid length' and 'total amino acid length', but you
have repeatedly failed to tell us how to calculate 'minimum amino acid length'.

This is in stark contrast to what I (and I daresay everyone who looks at
beta globin sequences) see as the model of how beta globin arose by
evolution, namely by modification of a duplicate of an ancestral alpha
globin sequence. If you start with that starting point, one needs very
few changes to affect function. And your 'minimum amino acid length' or
'total amino acid length' is utterly without relevance as a denominator
precisely *because* one does not start with a random sequence or a
random protein. One starts with a protein that already has most of the
relevant properties of beta globin (alpha globin binds heme and itself)
and that protein, not some random sequence or some random protein, gets modified.

What is interesting is that the same argument applies to features like
the bacterial flagella or the evolution of beta-galactosidase activity.
Only *you* are assuming that the starting point is some random sequence
or set of random proteins that have to pass, via a random walk of
complete non-function, to just happen to light upon the end result.
Only you are assuming that the ratio you calculate, with its meaningless
(for any real evolutionary model) denominator, has any real meaning.

I am looking at your numbers, not what you say. And your numbers do not
make any distinction between an 'invariant' position, a 'mildly variant'
position, or a position that can be any amino acid but one. And your
numbers essentially assume that there can be no significant difference
between total amino acid length and what you call 'minimum' amino acid
length that is more than a factor of 2. Otherwise, you would not find
*any* system that met your "thousands" numbers.

> I agree with everything that you said in the above
> paragraphs. Your problem though is that you left out a very important
> concept from your discussion here. You seem to have forgotten about
> "partially variant" amino acid positions.

No, I haven't. You, in your calculation, OTOH, have. Every amino acid
(or effectively every amino acid) in the protein seems to be equally
weighted in *your* calculation of 'minimum' or total amino acid length
used to determine the 'total sequence space'. And, as far as I can
tell, you treat each amino acid in a protein as if it were either
invariant or equally necessary to the final function. Otherwise there
would be no way of finding *any* structure that had several thousand
'fairly specified' amino acids, given that the average protein is about
300 amino acids long, in biological systems. How did you calculate your
denominator again? Did you use 1/20 for every *invariant* amino acid
and 5/20 for sites where any of 5 different amino acids can work and
19/20 for sites where only a single aa would not work? Or did you use
1/20 for every site? [It really doesn't matter, because any such
denominator is utterly irrelevant to any real evolutionary model
precisely because it does not consider the possible ancestral protein's sequence.]

> Even though most amino
> acids positions in a given protein with a particular function, such as
> the lactase, cytochrome, or histone function, are variable, there are
> different degrees of variability for each amino acid position. Like
> you said, only a very few of these positions are completely invariant.
> However, many more can only be changed between 2 or 3 positions. In
> the cytochrome c protein, for example, at least 80 amino acids are
> required at minimum to achieve this kind of function. Of these, at
> least half of the amino acid positions are constrained to within 2 or
> 3 amino acids. Another 20 positions are restricted to within certain
> classes of amino acids (hydrophilic, hydrophobic, basic, acidic, etc).
> Only a very few positions can vary widely without some sort of
> restriction.

And cytochrome c, like histones, are small proteins in which a
relatively large fraction of the sequence is in contact with the
co-factor heme. It is, in fact, for that very reason (its relatively
slow rate of change over evolutionary time) that cytochrome c is useful
for analysing deep phylogeny. Other proteins, and especially larger
proteins, are not as evolutionarily constrained (have a much smaller
percentage of amino acids under selective constraint). This, of course,
is another point you got flat-out wrong -- your assumption is that the
larger the protein the larger the *proportion* of amino acids that must
be 'fairly specified'. The facts are quite the opposite. In general,
the larger the protein the less evolutionary constraint on sequence is
seen. Remember that, once optimal function is obtained, selection is
largely a conservative force retaining function rather than a
destabilizing one changing function. That means that most of the
observed differences in protein sequences are selectively neutral or
near neutral (with occasional cases of mutation followed by intragenic
reversion). That means that many proteins, and especially the larger
ones, that are long diverged have less than 25% identity. And plots of
the rate of change in sequence is low for cytochrome c's (but not as low
as certain histones), high for fibrinogen peptide (where essentially no
amino acid constraint is present), and various rates in between for
other proteins (for example, hemoglobin is about half way between
cytochrome c and fibrinogen peptide).

> Based on these limitations, the total number of
> sequences in sequence space that can have this type of function have
> been estimated and quoted by some in this forum to be around 10e60.
> The problem with this estimate is that the minimum sequence space at
> this level of complexity (minimum of 80aa working at the same time in
> a fairly specified order) is over 10e100. This creates a ratio of
> potential cytochrome c sequences vs. non-cytochrome c sequences of
> less than 1 in 10e40.

And, to repeat the obvious (to everyone but you), that ratio is
predicated upon the assumption that current protein evolved directly
from a random sequence by a random walk. The very fact that nearly all
proteins are part of protein families gives lie to that assumption.
This is especially evident when you look at ebg beta-galactosidase.

> Histone sequences are even more restrained. Lactase sequences are
> certainly less restrained, but they are still more restrained than a
> handful of invariant amino acids would suggest.

But that is constraint *from* a fully active ancestral protein during
its common descent in all the organisms that have it. It does not tell
us squat about the original source of these proteins nor the selective
pressures that led to its original evolution. That other, completely
different sequences could also have evolved functions like motility or
lactase is obvious, one need only look at the several different
solutions to those pressures in different organisms.

> They have many
> partially variant positions that are none-the-less restricted to one
> degree or another. There is also a minimum amino acid requirement for
> lactase enzymes to be beneficial - seemingly well over 400aa at
> minimum. Of these 400aa only a handful are invariant, but many of the
> rest are still restricted to a small group of 2 or 3 amino acids.

The point remains, moreover, that your math still *always* presumes that
all current structure evolved directly from total sequence space (from a
random protein sequence or a random protein whose sequence is unrelated
to the final sequence by a random walk with no possible intermediate
states of utility). That is a direct consequence of your using 'total
sequence space' or 'total sequence space limited to the half of the aa's
I consider minimal' as your denominator.

No model of evolution held by *real* biologists assumes that that is
what happened. That ratio completely ignores the *real* models of
evolution, which are like the evolution of beta globin from alpha globin
rather than like nylonase (the rare exception to the rule). That is
what the structure of genomes and the proteins contained within tells
us. We have many genes that are members of gene families. Others are
clearly chimeric proteins. Others are clearly slightly modified
duplicates. Whenever a new 'function' or modified function arises, it
always occurs via a few changes in a protein that existed in the
organism. New functions that arise by chance creation of a crude
selectable function in a random sequence (ala nylonase) is the rarity
(but obviously not so rare as to be impossible, as nylonase shows).
Beta globin or beta galactosidase is the common face of evolution.

> Many more are restricted to certain classes of amino acids. Each
> restriction placed on an amino acid position lessens, in an
> exponential fashion, the density of those sequences with this type of
> function in the minimum sequence space at this level of functional
> complexity. At such levels of specificity, those functions that
> require longer and longer minimum amino acid sequences become
> exponentially rarer and rarer in sequence space.

And all this is irrelevant so long as you insist on a denominator of
total sequence space. That is where the bogosity is. Evolution does
not (very often) proceed by starting with a random sequence (when it
does, it is because that random sequence already has some selectable
activity, ala nylonase). It does not proceed by a random walk from a
random sequence or a random protein with no possibility of intermediate
utility.

It starts, usually, with a related protein already performing a related
function, even if it has NO selectable level of the activity or function
you are selecting for. This is certainly the case with
beta-galactosidase activity in E. coli. Occasionally, changes can
produce a new or emergent selectable function (as is the case with
bacterial flagella). The number of intermediate steps of functional
utility that must be crossed is what determines the amount of time
needed for a current function to have arisen, since an emergent function
cannot be a teleologic function. The intermediate functions must be
completely and independently useful all by themselves.

> Obviously many very large single protein sequences are even less
> specified than a lactase sequence. Consider that many lactase
> sequences use over 1000 amino acids per subunit.

E. coli beta-galactosidase is 1024 amino acids long. The final human
lactase is about the same size:

All articles can be found in PubMed. Unfortunately, the following is
not available on-line as full text, so I am unable to determine (until I
get to our library) how much of the *internal* duplication involves the
final active enzyme and how long the intramembrane and C terminal
portions of the protein that are unrelated to hydrolase activity are.
The pro-enzyme portion appears to act as an intramolecular chaparone,
facilitating the correct folding of the active part of the molecule.
Without reading the articles, I would suspect that the correct folding
basically involves the equivalent of tetramer formation (the two
internal duplicates in the pro portion and the two internal duplications
in the final portion) stabilizing the final internal 'dimer' structure
at a local energy minimum. My comments on this article are in [].

*****
EMBO J. 1988 Sep;7(9):2705-13.
Complete primary structure of human and rabbit lactase-phlorizin
hydrolase: implications for biosynthesis, membrane anchoring and
evolution of the enzyme.

Mantei N, Villa M, Enzler T, Wacker H, Boll W, James P, Hunziker W,
Semenza G.

Laboratorium fur Biochemie II, ETH-Zentrum, Zurch, Switzerland.

We report the primary structures of human and rabbit brush border
membrane beta-glycosidase complexes (pre-pro-lactase-phlorizin
hydrolase, or pre-pro-LPH, EC 3.2.1.23-62), as deduced from cDNA
sequences. The human and rabbit primary translation products contain
1927 and 1926 amino acids respectively. Based on the data, as well as on
peptide sequences and further biochemical data, we conclude that the
proteins comprise five domains: (i) a cleaved signal sequence of 19
amino acids; (ii) a large 'pro' portion of 847 amino acids (rabbit),
none of which appears in mature, membrane-bound LPH;

[Note that this means that the active protein is not 1927 amino acids
long, but is barely 1061 amino acids long. But, using the same logic
you have used yourself wrt duplications in large proteins, even this is
probably not the final case.]

(iii) the mature LPH, which contains both the lactase and phlorizin
hydrolase activities in a single polypeptide chain; (iv) a
membrane-spanning hydrophobic segment near the carboxy terminus, which
serves as membrane anchor; and (v) a short hydrophilic segment at the
carboxy terminus, which must be cytosolic (i.e. the protein has an
Nout-Cin orientation).

[Note that the human protein has several features that are quite
different from the E. coli enzymes. The human enzyme is a transmembrane
protein; the E. coli enzyme is not. The human enzyme also has phlorizin
hydrolase activity. Obviously, you have to eliminate all the parts of
the enzyme involved as a transmembrane anchor as being important for the
lactase function. The human enzyme is active as a monmer; the activity
of the E. coli enzyme requires tetramer formation. Does that mean that
the the human enzyme is less complex?]

The genes have a 4-fold internal homology, suggesting that they evolved
by two cycles of partial gene duplication. This repetition also implies
that parts of the 'pro' portion are very similar to parts of mature LPH,
and hence that the 'pro' portion may be a water-soluble beta-glycosidase
with another cellular location than LPH. Our results have implications
for the decline of LPH after weaning and for human adult-type alactasia,
and for the evolutionary history of LPH.

[I haven't read the full article because of the holidays, but if any of
the parts of the active enzyme that are actually involved in lactase
function (the transmembrane parts are not) are duplicated, the number of
'minimum' amino acids will have to be reduced accordingly, since
duplicated regions are not counted twice in large proteins, according to you.]

Here is the sequence of the human pre-pro-enzyme:

ORIGIN
1 melswhvvfi allsfscwgs dwesdrnfis tagpltndll hnlsgllgdq ssnfvagdkd
61 myvchqplpt flpeyfsslh asqithykvf lswaqllpag stqnpdektv qcyrrllkal
121 ktarlqpmvi lhhqtlpast lrrteafadl fadyatfafh sfgdlvgiwf tfsdleevik
181 elphqesras qlqtlsdahr kayeiyhesy afqggklsvv lraedipell leppisalaq
241 dtvdflsldl syecqneasl rqklsklqti epkvkvfifn lklpdcpstm knpasllfsl
301 feainkdqvl tigfdinefl scsssskksm scsltgslal qpdqqqdhet tdsspasayq
361 riweafanqs raerdaflqd tfpegflwga stgafnvegg waeggrgvsi wdprrplntt
421 egqatlevas dsyhkvasdv allcglraqv ykfsiswsri fpmghgssps lpgvayynkl
481 idrlqdagie pmatlfhwdl pqalqdhggw qnesvvdafl dyaafcfstf gdrvklwvtf
541 hepwvmsyag ygtgqhppgi sdpgvasfkv ahlvlkahar twhhynshhr pqqqghvgiv
601 lnsdwaepls perpedlras erflhfmlgw fahpvfvdgd ypatlrtqiq qmnrqcshpv
661 aqlpefteae kqllkgsadf lglshytsrl isnapqntci psydtiggfs qhvnhvwpqt
721 ssswirvvpw girrllqfvs leytrgkvpi ylagngmpig esenlfddsl rvdyfnqyin
781 evlkaikeds vdvrsyiars lidgfegpsg ysqrfglhhv nfsdssksrt prksayffts
841 iiekngfltk gakrllppnt vnlpskvraf tfpsevpska kvvwekfssq pkferdlfyh
901 gtfrddflwg vsssayqieg awdadgkgps iwdnfthtpg snvkdnatgd iacdsyhqld
961 adlnmlralk vkayrfsisw srifptgrns sinshgvdyy nrlinglvas nifpmvtlfh
1021 wdlpqalqdi ggwenpalid lfdsyadfcf qtfgdrvkfw mtfnepmyla wlgygsgefp
1081 pgvkdpgwap yriahavika harvyhtyde kyrqeqkgvi slslsthwae pkspgvprdv
1141 eaadrmlqfs lgwfahpifr ngdypdtmkw kvgnrselqh latsrlpsft eeekrfirat
1201 advfclntyy srivqhktpr lnppsyeddq emaeeedpsw pstamnraap wgtrrllnwi
1261 keeygdipiy itengvgltn pntedtdrif yhktyineal kayrldgidl rgyvawslmd
1321 nfewlngytv kfglyhvdfn ntnrprtara saryytevit nngmplared eflygrfpeg
1381 fiwsaasaay qiegawradg kglsiwdtfs htplrvenda igdvacdsyh kiaedlvtlq
1441 nlgvshyrfs iswsrilpdg ttryineagl nyyvrlidtl laasiqpqvt iyhwdlpqtl
1501 qdvggwenet ivqrfkeyad vlfqrlgdkv kfwitlnepf viayqgygyg taapgvsnrp
1561 gtapyivghn likahaeawh lyndvyrasq ggvisitiss dwaeprdpsn qedveaarry
1621 vqfmggwfah pifkngdyne vmktrirdrs laaglnksrl peftesekrr ingtydffgf
1681 nhyttvlayn lnyataissf dadrgvasia drswpdsgsf wlkmtpfgfr rilnwlkeey
1741 ndppiyvten gvsqreetdl ndtariyylr tyinealkav qdkvdlrgyt vwsamdnfew
1801 atgfserfgl hfvnysdpsl pripkasakf yasvvrcngf pdpatgphac lhqpdagpti
1861 spvrqeevqf lglmlgttea qtalyvlfsl vllgvcglaf lsykyckrsk qgktqrsqqe
1921 lspvssf

Summary: The protein encoded by this gene belongs to the
family 1
of glycosyl hydrolases. The protein is integral to plasma membrane
and has both phlorizin hydrolase activity and lactase activity.

*******

So, Sean, how do you go about determining the 'minimum amino acid
number' of 'fairly specified amino acids' in this protein? Obviously it
will come in as a smaller number of amino acids than either ebg or
beta-galactosidase of E. coli if you focus on the *function* of lactase
activity. But that means that you are ignoring the basically
independent activities of membrane anchorage, carboxy terminal sequence
inside the cell, the chaparone activity of the pro-enzyme portion, and
even the probably related to lactase phlorizin hydrolase activity, etc.

And, of course, you must also ignore the fact that this enzyme belongs
to a gene family and surely did not arise via a starting point that was
a random point in total sequence space or a random pre-existing protein.
Instead it almost certainly evolved via a modification of a
pre-existing glycosyl hydrolase belonging to family 1 glycosyl
hydrolases. That means that the number of events needed to get *some*
*selectable* level of the particular glycosyl hydrolase activity we are
looking at (be it phlorizin hydrolase or lactase) is much, much, much
smaller than the ratio you would calculate, which assumes that this
protein evolved from a random sequence by a random walk.

So, let's talk about ebg (1030 + 149 aa) and lacZ (.

*****
J Biol Chem. 1983 Sep 10;258(17):10204-7.

The active site regions of
lacZ and ebg beta-galactosidases are homologous.

Fowler AV, Smith PJ.

The active site-directed inhibitor
4-nitrophenyl-beta-D-galactopyranosylmethyltriazene, previously shown
(Fowler, A. V., Zabin, I., Sinnott, M. L., and Smith, P. J. (1978) J.
Biol. Chem. 253, 5283-5285) to alkylate methionine 502 in lacZ
beta-galactosidase, was used to label the second naturally occurring
beta-galactosidase of Escherichia coli (ebgo). The reagent was also used
to label two mutant forms of the enzyme (ebga and ebgb) selected for
enhanced lactase activity. In the case of ebgo and ebga, 75 and 85% of
the label, respectively, was incorporated into a tryptic peptide which
is homologous (38% identity) to residues 483-503 of the lacZ
beta-galactosidase sequence. In the ebgo and ebga enzymes, a serine
probably is alkylated. In the case of the ebgb enzyme, 61% of the label
is found on a tryptic peptide homologous (69% identity) with residues
457-468 of the lacZ beta-galactosidase. In this peptide, a glutamic acid
and a tyrosine residue are both alkylated.

[Note that the *active site* is much smaller than 'thousands of fairly
specified amino acids'. In fact, even *at* or near the active site,
substantial variation can exist.]

J Biol Chem. 1992 Jun 5;267(16):11126-30.


Glu-537, not Glu-461, is the nucleophile in the active site of (lac Z)
beta-galactosidase from Escherichia coli.

Gebler JC, Aebersold R, Withers SG.

Department of Chemistry, University of British Columbia, Vancouver, Canada.

The covalent intermediate formed during catalysis by the lac Z
beta-galactosidase from Escherichia coli can be trapped by reaction of
the enzyme with 2',4'-dinitrophenyl
2-deoxy-2-fluoro-beta-D-galactopyranoside, thereby inactivating the
enzyme. Kinetic parameters for this inactivation process with the holo-
and apo-enzymes have been determined. The intermediate so formed turns
over only very slowly (t1/2 = 11.5 h) resulting in reactivation of the
enzyme. The nucleophilic amino acid involved has been identified as
Glu-537 by using a tritium-labeled inactivator to label the enzyme, then
cleaving the labeled protein into peptides and purifying and sequencing
the labeled peptide. This residue is conserved in five homologous
beta-galactosidases and is different from that (Glu-461) proposed to be
the nucleophile (Herrchen, M., and Legler, G. (1984) Eur. J. Biochem.
138, 527-531) on the basis of affinity labeling studies with conduritol
C cis-epoxide. A role for glutamic acid residue 461 as the acid/base
catalyst is proposed and justified.

[That is, the crucial part of the activity involves a single amino acid
acting as a nucleophile.]
*******

Equally important, we find that beta galactosidase is not some 'random'
sequence unrelated to other proteins. Instead we see that it belongs to
family 2 of glycosyl hydrolases. The same is true for the ebg operon
(which includes ebga, the alpha subunit of about 1030 aa, ebgc, the beta
subunit of 149 aa, and ebgr, the repressor of ebg activity, which is
about the same size as the lacI repressor).

******
Genetics. 1989 Dec;123(4):635-48.


Erratum in: Genetics 1990 Mar;124(3):791.

DNA sequence analysis of artificially evolved ebg enzyme and ebg
repressor genes.

Hall BG, Betts PW, Wootton JC.

Molecular and Cell Biology, University of Connecticut, Storrs 06268.

The ebg system has been used as a model to study the artificial
selection of new catalytic functions of enzymes and of inducer
specificities of repressors. A series of mutant enzymes with altered
catalytic specificities were previously characterized biochemically as
were the changes in inducer specificities of mutant, but fully
functional, repressors. The wild type ebg operon has been sequenced, and
the sequence differences of the mutant enzymes and repressors have been
determined. We now report that, contrary to our previous understanding,
ebg enzyme contains 180-kD alpha-subunits and 20-kD beta-subunits, both
of which are required for full activity. Mutations that dramatically
affect substrate specificity and catalytic efficiency lie in two
distinct regions, both well outside of the active site region.

[That is, the changes in ebg that convert it from a protein with no
effective or significant lactase ability into one with selectable
lactase activity involves no changes in the active site that does the
actual hydrolysis, but only in the portions of the molecule that affect
substrate binding and catalytic efficiency. This is in agreement with
the idea that the substrate binding *function* and the catalytic
*function* are separable and can and do change independently of each
other. Sean's argument is that all these functions are inseparable and
one cannot evolve (or modify) one of these features without disrupting
the interaction of the needed 'thousands of fairly specified amino acids
working together' that makes up the whole.]

Mutations that affect inducer specificity of the ebg repressor lie
within predicted sugar binding domains.

[The lacI repressor works by binding a sugar as an allosteric effector,
changing the 3-D structure so that it now binds the promoter of the lac
operon (lacYZA). Changing the sugar bound can have a dramatic effect on
binding of the repressor. My guess is that the mutations that
positively affect lactase activity of ebg essentially results in a loss
of the 'repressor' *function* that ebgr had.]

Comparisons of the ebg beta-galactosidase and repressor with homologous
proteins of the Escherichia coli and Klebsiella pneumoniae lac operons,
and with the galactose operon repressor, suggest that the ebg and lac
operons diverged prior to the divergence of E. coli from Klebsiella. One
case of a triple substitution as the consequence of a single event is
reported, and the implications of that observation for mechanisms of
spontaneous mutagenesis are discussed.

[Again supporting the idea that lactase activity does not evolve by
starting with a random sequence and taking a random walk toward lactase
as a teleologic goal. Instead, we have a duplication of an operon in an
ancestral organism. One of these operons evolved into the lacI-lacYZA
system. The other evolved into the ebgR-ebgAC system, which, in its
native condition does not have lactase activity, but retains much of the
*evolutionary potential* to produce lactase that existed in the original
ancestral operon. This despite having only about 50% DNA sequence
homology and only 35% amino acid sequence homology to lacYZ.]
******

> However, this is not
> the minimum requirement, which seems to be less than half this number
> at around 480aa.

How did you arrive at this number? Are these 'invariant' amino acids?
Somewhat variant amino acids? Amino acids for which at least one amino
acid will NOT work? What criteria were used to identify whether an
amino acid was one of the 50% of amino acids that was 'fairly highly
specified' and which were not? Or is 480 a WAG (wild-assed guess) based
on nothing at all? And, once you specified 480 as the number, how did
you calculate the denominator of your ratio? Did you take 1/20 to the
480th power? Or did you consider that most of these 480 amino acids
need not be invariant? Not that it matters, of course, since any such
denominator, even calculated correctly, is GIGO which is irrelevant to
real evolutionary mechanisms.

> The same thing is true of mammoth proteins, such as
> the Titan protein, that most likely require far fewer amino acids to
> achieve a minimum level of beneficial function than are normally used.
> However, there are functions that do require thousands of fairly
> highly specified amino acid positions working together at the same
> time.

Bullsquat. Name them and justify the idea that they require 'thousands'

of 'fairly highly specified' amino acid positions working together at
the same time.

You need to look again at the lactases and tell me how you arrived at
your 480 amino acid number. The length, in amino acids, of these three
lactases (the final rather than pro- or prepro- transmembrane enzyme in
humans, and the two different, but evolutionarily related E. coli
enzymes, one being a two peptide system) is about the same, a relatively
long 1000 amino acids. You assert (without giving a shred of evidence
that I have ever seen) that about half of the amino acids in each of
these proteins must be 'fairly highly specified' in order for these
proteins to have *any possible* lactase activity. That implies that
there must be some very easy to identify long stretches of sequence
similarities in the sequences of these proteins, namely those specific
amino acids that must be 'fairly highly specified' in order to have any
lactase activity. The 480 'fairly specified' amino acid sites that
stand out as the lonely island of lactase activity in the ocean of
nothingness (no functional activity). They should, because of the
rarity of lactase activity in the ocean of nothingness, be as similar in
the human as in the E. coli enzymes.

> Multiprotein functions often require fairly high specificity of amino
> acid positions in each of their individual protein parts. Since each
> protein part works together at the same time to achieve a unified
> function, each of the amino acids in all of the protein parts are also
> working together at the same time to achieve this unified function.
> The minimum amino acid requirement for many of these multiprotein
> functions is in the thousands and even tens of thousands. For
> example, the minimum number of different kinds of proteins required
> for the flagellar motility function seems to be more than 20. On
> average each of these proteins requires a minimum amino acid sequence
> of more than 200 to 300aa. At minimum then the motility function of a
> flagellar system requires at least 4,000 to 6,000 fairly specified
> amino acids working together at the same time.

But *even if* you were to show that some 200-300 aa's were absolutely
necessary for a protein to function as a beta-galactosidase, that does
not mean that 200-300 amino acids needed to *change* to produce a
beta-galactosidase from an ancestral family 1 glycosyl hydrolase (the
human lactase) or from an ancestral family 2 glycosyl hydrolase (the
lacZ and ebg systems), respectively. Evolution does not involve (except
rarely, as in nylonase) starting with a random sequence (and even when
it does, the random sequence is not devoid of selectable activity --
some selectable nylonase activity was immediately present) or a random
protein or the duplicate of a random protein. It modifies an existing
protein that already has the potential to generate some favorable
selectable function with only a few changes. That selectable function
need not be the teleologic function you determine, in arrogant
hindsight, as *the* function which must be evolved.

*******
Mol Biol Evol. 1985 Nov;2(6):478-83.

Sequence of the ebgR gene of
Escherichia coli: evidence that the EBG and LAC operons are descended
from a common ancestor.

Stokes HW, Hall BG.

Molecular and Cell Biology Department, University of Connecticut, Storrs 06268.

The sequence of ebgR, the gene that encodes the EBG repressor, was
determined. There is 44% DNA sequence identity between ebgR and lacI,
the gene that encodes the LAC repressor. There is also 25% identity
between the amino acid sequence of lacI and the deduced amino acid
sequence of ebgR. The sequence of 596 bp distal to ebgA, the structural
gene for EBG beta-galactosidase, was also determined. Within that region
there were two sequences, 74 and 100 bp long, that showed 46% and 50%
identity, respectively, to sequences in the first 600 bp of lacY, the
structural gene for the lactose permease.

[That is, the structure of the ebg operon has some sequence homology
that indicates that it started out with 'lacY' equivalent, in the right
order, that was later lost. Note that 25% identity indicates a deep
point of divergence.]

The organization and direction of transcription of the repressor and
structural genes of the two operons are identical. Taken together with
the homology between ebgA and lacZ (as demonstrated in the companion
article in this issue), this provides strong evidence that the EBG and
LAC operons are descended from a common ancestor. The map position of
these two operons supports the notion that these operons diverged
following a genome duplication event in an ancestor of Escherichia coli.

Mol Biol Evol. 1985 Nov;2(6):469-77.

Sequence of the ebgA gene of
Escherichia coli: comparison with the lacZ gene.

Stokes HW, Betts PW, Hall BG.

Molecular and Cell Biology Department, University of Connecticut, Storrs 06268.

We have sequenced the ebgA (evolved beta-galactosidase) gene of
Escherichia coli K12. The sequence shows 50% nucleotide identity with
the E. coli lacZ gene, demonstrating that the two genes are related by
descent from a common ancestral gene. Comparison of the two sequences
suggests that the ebgA gene has recently been under selection. A
significant excess of identical, rather than synonymous, codons used to
encode identical amino acids at the same positions in the aligned
sequences implies that some form of selection is operating directly at
the DNA level. This selection is independent of, and in addition to,
selection based on codon usage or on function of the gene products.
*******

That is, the evidence shows that ebg activity evolved originally by a
duplication of an ancestral family 2 glycosyl hydrolase function of some
kind that also led to the lac operon (in fact, a chance duplication of
*an entire genome* -- a rather common phenomenon in the higher
eucaryotes we call plants, and important in the history of vertebrates
as well -- rather than just the ancestral operon in question). That
means that neither enzyme evolved via a start from a random sequence by
a selectively neutral random walk across a vast ocean of inutile
sequence space. Both arose by short, simple modification of an
ancestral glycosyl hydrolase which may well have had *neither* lactase
activity nor the different activity of the ebg gene. It certainly means
that it is possible to generate lactase activity from a protein that had
no lactase activity. But the mechanism has no relationship to the model
you propose as 'evolution'.

The mechanism of generating a *new* function (lactase activity) in a
protein that did not have selectable lactase activity did not involve
starting with a random sequence or random, unrelated protein, and does
not proceed via a long random walk of non-functional states. It started
with a related protein (that has much of the structure and active sites
of a glycoside hydrolase) and proceeded via a very short chain of change
that only affected the *function* of substrate binding.

I, in fact, agree with you that *no* 'new' function *ever* arises by the
mechanism you propose (the starting point being a random sequence
without related function -- either a random sequence or a random protein
-- followed by a long random walk to the only functional possibility).
**Nor have you presented any *real* evidence that any new function
*needs* to arise that way.**

Focus on that last sentence. That is why I label your calculations
irrelevant GIGO. You are continually posting numbers that only make
sense if you are arguing that evolution works by starting with a random
sequence or a random protein and proceeds by a random walk without any
possibility of intermediate function until it reaches some teleologic
'function' you choose as the only possible 'function' it or any
intermediate states could ever have. Those are numbers you think you
can calculate solely by looking at the final or, in your mind,
teleologic structure -- the island in the ocean of nothingness --
without taking any possible ancestral sequence or function into
consideration by assuming that the starting point is a random sequence.
But the fact that you can make up a GIGO ratio does not make it
biologically relevant. And it isn't.

> Evolving a new type of function within such a level of minimum
> informational complexity would simply take zillions of years to
> achieve via mindless processes alone. You have failed to explain how
> any functional or non-functional starting position could achieve such
> functional diversity at such levels of informational complexity that
> we see today in all life forms (via mindless evolutionary processes
> alone).

The above is a meaningless mantra until you specify what you mean by
'new' type of function, tell us how you quantitate 'level of minimum
informational complexity' and which 'mindless processes' you think are
those involved in evolution. And you need to tell us why you repeatedly
assume that evolution involves a random starting point and a random walk
with no possibility of intermediate functionality when all the
*evidence* says that that is not, in fact, how evolution works.

[Sean Pitman]

dkomo <dk...@concentric.net> wrote in message news:<3FF1D809...@concentric.net>...

> Sean Pitman wrote:
>
> [snip]
>
> > Evolving a new type of function within such a level of minimum
> > informational complexity would simply take zillions of years to
> > achieve via mindless processes alone. You have failed to explain how
> > any functional or non-functional starting position could achieve such
> > functional diversity at such levels of informational complexity that
> > we see today in all life forms (via mindless evolutionary processes
> > alone).
> >
>
> No, it wouldn't take zillions of years because it uses the WEASEL
> principle described by Richard Dawkins. Anyone who has played around
> with a WEASEL program knows how powerful are the operations of random
> mutation followed by selection and how rapidly such a program can
> converge to even a long target string. In a computer simulation of
> the build-up of your target protein starting from strings of random
> amino acids, the convergence wouldn't even take zillions of
> nanoseconds.

This is a common misconception among many otherwise intelligent
people, to include Richard Dawkins. Such algorithms, as were used by
Dawkins in his, "Methinks it is like a weasel" sequence evolution are
not following evolutionary mechanisms at all. Here is what is really
going on with these so-called "evolutionary" algorithms:

In his 1986 book called "The Blind Watchmaker" Dawkins described an
experiment of his that showed how evolution is supposed to work. He
programmed a computer to generate random sequences of letters to see
if the computer would, over time, generate the line from Hamlet,
"METHINKS IT IS LIKE A WEASEL." This line has 28 characters (including
spaces), so the computer was programmed to make 28 selections using
the 26 letters of the alphabet plus a space to make 27 possible
characters to pick from. A typical output was "MWR
SWTNUZMLDCLEUBXTQHNZVJQF." With this information, a calculation of the
probability of picking the "correct" sequence can be made, as well as
how long it would take, on average, to find this correct sequence.
Dawkins figured that it would take his computer a million million
million million million million years (or a trillion trillion trillion
years… 1 x 10e35 years), on average. Well, this is clearly way too
long for the current theory. So, how could evolution possibly take
place? Dawkins now put some "natural selection" into the computer
program to simulate "real life" more closely. The computer made
multiple copies of "MWR SWTNUZMLDCLEUBXTQHNZVJQF" (Offspring) while
introducing random "errors" (mutations) into the copies. The computer
then examined all the mutated "offspring" and selected the one that
had the closest match to, "METHINKS IT IS LIKE A WEASEL." This
selection by the computer (nature) was now used to make new copies and
random mutations (in a "new generation"), from which the best copy was
selected again . . . and so on. By ten "generations" the sequence had
"evolved" to read something like, "MDLDMNLS ITJISWHRQEZ MECS P." By
the thirtieth generation it read something like, "METHINGS IT ISWLIKE
B WECSEL." Instead of taking many trillions and zillions of years this
time, the computer came up with the "fittest" phrase in about 40
generations. Of course Dawkins made a disclaimer that this experiment
was not intended to show how real evolution works, but that it does
illustrate the advantages gained by a selection mechanism in an
evolutionary process.

Certainly this is a fine illustration except for one subtle flaw.
Dawkins's computer did not make its selection based on phrase
function, but on phrase sequence comparisons to an "ideal" phrase. Why
is this a problem? After all, it's just an illustration. Perhaps it is
an illustration, but it is not illustrating anything even close to
what natural selection is capable of. The theory of evolution is based
on natural selection and natural selection selects based only on
sequence function. If two genetic sequences are both non-functional or
if they both have the same function, then natural selection cannot
select between them. In other words, nature is blind to their genetic
differences if they both have the same function. If Dawkins had wished
to mirror the type of selection proposed by the theory of evolution,
he would have based his computer model on functional phrase selection.
The problem here is that "MDLDMNLS ITJISWHRQEZ MECS P" doesn't mean
anything. This phrase has no language function. A selection mechanism
that only recognized changes in function would look at this phrase and
compare its function to the function of the phrase, "SDLDMNLS
ITJISWHRQEZ MECS P" where an M was mutated into an S. Of course, both
phrases have the same non-functional function. A selection mechanism
that is based on function will not be able to tell the difference
between them. Therefore, one will not be selected over the other for
"survival" in the next generation. Therefore, there will be no
"directed" evolutionary change toward some sort of improvement or new
function.

Another problem with Dawkins's illustration is that the computer
already had the "ideal" phrase programmed into it by an intelligent
designer (Dawkins) to begin with. The evolution of something that is
already there is not the evolution of anything new at all. If nature
already has what it wants or needs, then it does not need to "evolve"
it.

Dawkins uses a selection mechanism that does seem to work, but
mindless natural processes do not have access to such a selection
mechanism. Dawkins uses a selection mechanism that is capable of
comparing non-functional sequences with an ideal sequence. A mindless
nature only recognizes functional differences, not sequence
differences since mindless nature has no ideal sequence to compare
other sequences with.

Mindless nature does not "see" the actual letters of words (in DNA or
Protein). All that a mindless nature can see is what function results.
Since function is arbitrarily attached to words by an outside source
of information, such as a system of function or definition, a gradual
change in the letters of the words themselves is not necessarily going
to result in a gradual evolution of their meaning or function. A
gradual change in a recognized word or phrase will most likely destroy
its original meaning well before any new word or phrase is recognized
as having meaning. Without function the entire way, natural selection
is blind and even Richard Dawkins will admit that without natural
selection to guide evolution, evolution is statistically impossible.

Yes, blind evolution might result in change to the spelling of genetic
sequences, but the changes would not necessarily be functional
changes. Changing one nonfunctional word into another nonfunctional
word is a "change", but it is not a functional change since both words
remain, well, nonfunctional. Two nonfunctional words both have the
same nonfunctional function. So, all algorithms, to include both yours
and Dawkins's, that are based on the ability to select between
non-functional or non-beneficial sequences are worthless as an
explanation for describing how evolution is supposed to create
functional diversity at higher and higher levels of functional
complexity.

For a more detailed discussion of this concept see:

http://naturalselection.0catch.com/Files/methinksitislikeaweasel.html

Sean
www.naturalselection.0catch.com

[sweetnes...@yahoo.com]

(Several other people have asked some questions that are both
interesting and worth responding to; unfortunately, I currently simply
have no time, and I have to limit myself to the discussions that are
already in progress. Sorry :\ )

>The only problem with these examples of real time evolution
>is that the minimum fairly specified amino acid "part" requirements
>have been no more than a few hundred amino acids, working
>at the same time, at best.

This is pointless. And, I'm sorry, but on the basis of what you have
written, I cannot believe that you've read either my message in
entirety, or even a small fraction of the references I've listed. You
just repeat yourself over and over again, often asking questions that
are answered a few paragraphs later in the post you're responding to
("minimum length of cytochrome c", when there is a discussion of other
small molecules that perform the same function, and a discussion of
evolution building up from small to larger; talking of minimum
requirements of bacterial motility, when I mentioned both not only in
text, but gave a few references to the work that puts everything you
said on it in quite unfavorable light; etc.).

Furthermore, you have certainly not even glanced at any of the texts
on directed evolution, or on structural evolution. You would have
found that proteins that you claim cannot possibly evolve have been
generated in lab, by completely random processes. You would also find
that structural evolution papers detail the development of many large
proteins, and you could then point out the particularly improbable
steps (which you, actually, would not be able to find; but that is
another story altogether).

So, as far as proving that evolution does occur, and that your
mathematical model has nothing to do with reality, I'm done. It's all
here, people can read it, and reach their own conclusions. All that's
left is examining some other "interesting" aspects of your theory, and
of your worldview in general. So, let's reverse roles for a second,
and I'll ask the questions - mostly, same questions you ignored before
(arranged in order of simplicity):

1. What do you actually believe happened? When did the Earth come to

existance, and how? What is the history of the Earth as you see it?

How differentiated was life when it was "designed" (for example, were
horses and donkeys created at the same time, or are they variations of
the same "kind")?

2. Why don't you publish your devastating theory in a professional
journal? If you think they won't accept it, why do you think that is
the case?

3. You keep reffering to "specified complexity" and "x amino acids
working together". Please define those terms. To be more specific, you
stated that "because a protein is big does not make it more

functionally complex (i.e., requiring more genetic real estate at

minimum)." How do you differentinate between size and complexity?

(You did give an indirect answer to the previous question: if a large
protein is shown to evolve in the lab, then "there is no evidence"
that that particular protein "requires" its size for its "specified
complexity", and therefore it is not complex enough. If evolutionary
pathway of a protein (or a system of proteins) is still unknown, then
it is obviously too complex to ever evolve on its own, and therefore
proves your "theory". I would, however, like a more quantitative
answer, something that can be applied accross the board.)

4. Read the several paragraphs appended after the questions. You do
not need to answer it paragraph by paragraph (unless you want to), but
give me your answer to the core problems of ID that are presented
there.

5. Unanswered from a previous post:
>If they are not very common, perhaps millions and
>even billions of years would not be enough time.

Why don't you try to prove it? But instead of calculating meaningless
numbers of total combinations of amino acids, do it right. Take, for
example glycogen synthase (a quite distant molecule, mind you), and
give me a model how many mutations would be required to move from it
to a lactase. Or if you wish to be fair (and to reduce the amount of
work you have to do), take any disaccharide hydrolase, and estimate
the amount of time it would take to evolve a lactose recognition
domain.

6. There are many large, complex proteins whose evolutionary pathway
has been explained in quite a bit of detail. Pick one, and explain why
is the proposed evolutionary pathway impossible, which steps are
unlikely to occur, and why?

There, six direct questions. Since the last two require some
calculations, I will settle for answers to the first four, and you can
leave the last two for later on.

******************
[What follows is my response to your claim that "You don't understand
how, but you still have faith that all original causes in this
universe are mindless". You never answered it before, and I'm really
interested in seeing what an ID-er thinks about questions presented
here.]

But evolution applies from the first roughly functioning cell onwards;
you might be talking about abiogenesis, so let's analyze that for a
moment. How did the first cell come about?

This is something that is quite unknown, and entirely theoretical. So,
yes, unlike the case of evolution, one cannot make one's mind based
entirely on evidence. What do we do?

You might propose that, since we don't know, it is equally rational to
simply say that there was an Intelligent Designer who designed the
first cell, as it is to assume that the mindless processes did it. You
might say that it takes faith in both cases, so there is no real
difference.

If you did so (and I don't know if you would, but I have a general
idea), you would be wrong. There is a simple reason to assume that the
first cell arose by mindless processes rather then by intelligent
design. The reason is evidence of existence.

We *know*, without doubt, that mindless processes exist. Everything
that happens, happens according to the laws of the universe. We have
ascertained without much doubt that the same laws (without any large,
visible changes) applied for quite a few billions of years. We know
that those laws were around when the life first began. Therefore, we
have good reason to think that the laws of the universe were involved
in the begining of life.

As for Intelligent Designer...we have no idea if a supernatural entity
such as he exists or not. We see no evidence of his existence, in the
present or in the past. He is entirely in the domain of imagination,
and as such we can imagine anything about him, and therefore measure
nothing. Not only have we no reason to assume his existence, but
taking him as an "option" would literally mean end of any further
research [...]

How did a star form? If we assume it formed by mindless processes, we
can use our knowledge of the laws of physics, and see if we can
discover a solution. After many years of work, we may produce a
theory, compare it to the reality, and see if it stands up to
scrutiny. If we assume that a Designer made it, that is it. We give
up, and say "goddidit", the end. We cannot know anything further.

Through most of the Dark Ages, people took the second approach;
whatever it is, God made it to be that way. This is, in fact, the
reason why that period of time is called "The Dark Ages". Modern
science uses the first approach, and it got as quite a bit further.

So, in case of abiogenesis, I will look at both options: accept a
Designer without *any* reason to do so, plop on a chair, and say
"done". Or assume that the processes that did [and do] EVERYTHING ELSE
around me also produced life, and try to figure out how they did it.

[Rodjk]

seanpi...@naturalselection.0catch.com (Sean Pitman) wrote in message news:<80d0c26f.03123...@posting.google.com>...

> years? 1 x 10e35 years), on average. Well, this is clearly way too

Why don't you just say:
1) I don't have a clue what Natural Selection is about.
2) I tend to talk a lot about things I do not understand.
This would accomplish what your long post said, and be a lot quicker
and easier to read.

Rodjk #613

>
> Sean
> www.naturalselection.0catch.com

[Sean Pitman]

rjk...@yahoo.com (Rodjk) wrote in message news:<dbe402.040113...@posting.google.com>...

> Why don't you just say:
> 1) I don't have a clue what Natural Selection is about.

Perhaps you could fill me in?

> 2) I tend to talk a lot about things I do not understand.

Then perhaps you could explain to me where I'm going wrong? I mean
really, it is very easy to tell someone that they are wrong, but it is
quite another thing to explain, in a logical way, why they are wrong.

> This would accomplish what your long post said, and be a lot quicker
> and easier to read.

"Quicker" and "easier" are not necessarily "better". For example,
your post here is a common example of what many brainless contributors
to this forum think are "helpful". Certainly such a post is "quick
and easy", but obviously brainless. I expect more from you
evolutionists. After all, you are supposed to be the enlightened
ones.

> Rodjk #613

Sean
www.naturalselection.0catch.com

[david ford]

Sean Pitman <seanpi...@naturalselection.0catch.com> on 31 Dec
2003:
dkomo <dk...@concentric.net> wrote:

If you accept the fact of "change," you accept "evolution."
Well, at least according to some definitions.

> but it is not a functional change since both words
> remain, well, nonfunctional. Two nonfunctional words both have the
> same nonfunctional function. So, all algorithms, to include both yours
> and Dawkins's, that are based on the ability to select between
> non-functional or non-beneficial sequences are worthless as an
> explanation for describing how evolution is supposed to create
> functional diversity at higher and higher levels of functional
> complexity.

Nice critique. In the "Advent of the Head Monkey" section of his
_Commentary_ article on the Discovery Institute website, Berlinski has
some thoughts on Dawkins's weasel musings.
http://tinyurl.com/ygrf

And 1998 Steele, Lindley, & Blanden had some kind words to say about
Berlinski.
http://tinyurl.com/3c5gs

[Sean Pitman]

sweetnes...@yahoo.com wrote in message news:<4d71d185.04011...@posting.google.com>...

> >The only problem with these examples of real time evolution
> >is that the minimum fairly specified amino acid "part" requirements
> >have been no more than a few hundred amino acids, working
> >at the same time, at best.
>
> This is pointless. And, I'm sorry, but on the basis of what you have
> written, I cannot believe that you've read either my message in
> entirety, or even a small fraction of the references I've listed.

I have read several of the references that you listed and commented
extensively on some of them (to include your M. xanthus example). You
just fail to understand why I don't think they support what you say
they support - such as the notion that evolution beyond the lowest
levels of complexity has ever occurred. Sure, large proteins evolve,
but the examples that you've listed are certainly not very specified
at all. They are little more than repeating units of the same thing,
polymer-like. Such protein systems/function require relatively little
genetic real estate to code for them.

When I talk about the flagellar motility system as an example of
system that requires a few thousand fairly specified amino acids
working together at the same time, you scoff and say that this problem
has already been solved. You quote a few more references, but after
reading several such references about the supposed evolution of the
flagellum I am struck by the fact that not even one of the proposed
intermediate steppingstone systems has been shown to evolve from one
to any other. In fact, it seems to me that these steppingstone
systems are too few and are very far apart in sequence space. The
neutral gaps between these steppingstone systems are truly enormous.

Now obviously you disagree, but you seem unwilling to detail to me,
specifically, how such a specified, multiprotein system as a flagellar
motility system involving several thousand amino acids working
together at minimum, could evolve in some sort of stepwise fashion.
How are the intervening neutral gaps reduced to a manageable size so
that they can be crossed this side of a zillion years?

> You
> just repeat yourself over and over again, often asking questions that
> are answered a few paragraphs later in the post you're responding to

Come on now Sweetness. I'm not the only one who repeats himself over
and over again. However, I do deliberately try not to ask a question
that you have already answered in your post. I always read the entire
post through before I try and respond to it. If I end up asking a
question that you feel was already answered by your post, most likely
I did not feel the same way or else I wouldn't have asked it.

> ("minimum length of cytochrome c", when there is a discussion of other
> small molecules that perform the same function,

You did mention other smaller molecules that could *potentially*
perform the electron transport function, but none of these
hypothetical molecules has ever been shown to work in a living
creature in such a capacity. Something like this might seem very good
on paper, but it fails miserably when tried in real life. I'm sorry,
but it seems like there certainly is a minimum length requirement for
the type of function that the cytochrome c molecule does that is more
than the half a dozen or so residues that you seem to be proposing as
the minimum. The high degree of restriction of this protein (60% of
its positions limited to within 3aa) and the lower limits of size of
more than 70aa seem to be about as trimmed down as you are going to
get in a real life form of any kind. If you know of a smaller protein
that actually performs this job in an actual living thing, please let
me know.

> and a discussion of
> evolution building up from small to larger;

Yes, you have done this. You have shown that evolution can build up
very large *non-specified* structures. In fact, you yourself note
that the large a protein gets the less specified it becomes on
average. Isn't that so? Well, I am in perfect agreement with this
concept. I agree that evolution can build up very large proteins if
and only if they have a very low specificity. What evolution cannot
seem to do is build up a system of function that has a fairly high
level of specificity beyond a few hundred amino acids working at the
same time.

That challenge/prediction is what I have repeated over and over again
and you have continually tried to misstate and evade that challenge.
You try to float examples of large proteins with very low specificity
as disproving my position. You try floating examples of functions
evolving where no new structure actually evolved, only an
up-regulation/increased production of a structure that was already
there was required.

Come on now Sweetness. You haven't answered my challenge and you know
it. As frustrating as that must be for you, the best you have done is
shown that large systems with low specificity evolve and that small
systems with high specificity evolve, but you haven't even come close
to showing how large systems with high specificity (i.e., the
flagellar system of motility) can evolve.

That is my challenge. What ya got? Where is your reference for this
one? Just one reference is all you need to shut me up. Go ahead
then. I *repeat myself* - List your best reference that demonstrates
the evolution of a system that requires, at minimum, at least a few
thousand fairly specified amino acids working together at the same
time. This system can be composed of a single protein or multiple
proteins. Of course, since single proteins tend to loose specificity
with size, the fair degree of specificity that I am talking about will
most likely be met by a multiprotein system - such as the flagellar
motility system or some other equally specified system of similar
size. If you think you have already listed such a reference, list it
again here because I obviously missed it.

> Furthermore, you have certainly not even glanced at any of the texts
> on directed evolution, or on structural evolution. You would have
> found that proteins that you claim cannot possibly evolve have been
> generated in lab, by completely random processes.

The reference you listed discussed large proteins with low specificity
or small protein systems with high specificity. As far as I was able
to ascertain, none of the listed references discussed the evolution of
a system that was both large as well as fairly specified in its amino
acid order.

> You would also find
> that structural evolution papers detail the development of many large
> proteins, and you could then point out the particularly improbable
> steps (which you, actually, would not be able to find; but that is
> another story altogether).

The evolution of large proteins with low specificity is not
"improbable" at all. Why would I even try to challenge this notion?
I'm not talking about this sort of evolution now am I?

> So, as far as proving that evolution does occur, and that your
> mathematical model has nothing to do with reality, I'm done. It's all
> here, people can read it, and reach their own conclusions.

They sure can. And, if those who read what you write have an ounce of
skepticism in their bodies they will quickly see that you haven't
presented anything but an bunch of hot air and just-so stories.
You've got nothing to back yourself up in your challenge of my actual
position.

> All that's
> left is examining some other "interesting" aspects of your theory, and
> of your worldview in general. So, let's reverse roles for a second,
> and I'll ask the questions - mostly, same questions you ignored before
> (arranged in order of simplicity):
>
> 1. What do you actually believe happened? When did the Earth come to
> existance, and how?

It seems to me that the Earth itself could be and is most likely quite
old. Many billions of years old is not at all out of the question in
my thinking. However, the evidence as I see it seems to indicate that
life on earth and much of the geologic column is very young - less
than 10,000 years old in fact.

> What is the history of the Earth as you see it?

http://naturalselection.0catch.com/Files/geologiccolumn.html
http://naturalselection.0catch.com/Files/fossilrecord.html

> How differentiated was life when it was "designed" (for example, were
> horses and donkeys created at the same time, or are they variations of
> the same "kind")?

I think that horses and donkeys did share a common ancestor.

http://naturalselection.0catch.com/Files/donkeyshorsesmules.html

Certainly there have been a great many changes over time, but not
significant enough to cross certain boundaries that I think were
established in the beginning, suddenly, by intelligent design. Many of
the functional differences between certain groups of animals, such as
between humans and apes, are not known in sufficient detail to
definitely rule out or rule in a common ancestry via the study of
genetics alone. However, many other genetic functions, especially
those found in bacteria, various insects, and the like, are known in
much more detail so that the limits of evolution can be more easily
seen in such creatures.

http://naturalselection.0catch.com/Files/earlyman.html

> 2. Why don't you publish your devastating theory in a professional
> journal?

Someday I plan on publishing my ideas either in a professional journal
or a book published by a secular publisher.

> If you think they won't accept it, why do you think that is
> the case?

Although the bias is very heavy against anything that challenges the
fundamental "truth" of evolution is popular science journals, I think
that there might still be a chance a being accepted in such a journal.
I might be too optimistic here to think that the editorial staff
would be open minded enough to put some of my ideas in their journal,
but I'm certainly willing to give it a shot when I get the chance.
Currently I'm far too busy, at least for the next couple years, with
my current job and research in medical pathology. Things should slow
down a bit after a couple of years however. So, perhaps I will work
on something for publication in a serious journal at that time.

> 3. You keep reffering to "specified complexity" and "x amino acids
> working together". Please define those terms. To be more specific, you
> stated that "because a protein is big does not make it more
> functionally complex (i.e., requiring more genetic real estate at
> minimum)." How do you differentinate between size and complexity?

You already know the answer to this question. You yourself note that
larger proteins tend to be less specified than smaller proteins. So,
how do you define the term "specified"? What is your own definition?
Likely, it is the same definition as I have.

In short, system specificity is the range to which change within a
system can occur without a complete loss of the original type of
beneficial function. Specificity is basically a measure of
irreducible complexity. Since all types of functions are irreducibly
complex at some point or another (a minimum part requirement exists
for all types of functions), all systems have a degree of
specificity/irreducible complexity.

Obviously then, the greater the variability/changeability, the less
the specificity, and visa versa. Again, small proteins can be very
specified (very limited in the number and variability of amino acid
positions) or loosely specified. Likewise, large proteins can have
very little specificity (have very few limits on which positions can
change or what amino acid can occupy a given position) as well.
However, there can also be large protein systems that maintain a
fairly high level of specificity (allowing for relatively little
change, percentage wise, among their amino acid positions). Such
systems include the multiprotein system of flagellar motility and the
like. The flagellar system of motility has a high minimum amino acid
requirement and also maintains a fairly high degree of combined/total
specificity.

> (You did give an indirect answer to the previous question: if a large
> protein is shown to evolve in the lab, then "there is no evidence"
> that that particular protein "requires" its size for its "specified
> complexity", and therefore it is not complex enough.

Not at all. There has to be some sort of independent evidence
suggesting that this protein is or is not specified beyond a certain
degree. You yourself seem to have a very good idea about how to judge
the specificity of proteins independently since you have commented
several times that you know that larger proteins tend to have lower
degrees of specificity.

> If evolutionary
> pathway of a protein (or a system of proteins) is still unknown, then
> it is obviously too complex to ever evolve on its own, and therefore
> proves your "theory".

Again, this is a false statement. I have predicted ahead of time what
kind of systems I think cannot evolve and why. The specificity of
such systems as the flagellar motility system is fairly well known and
established by more independent means (such as the variability found
between the same protein doing the same job in the same system of
function, in different creatures).

> I would, however, like a more quantitative
> answer, something that can be applied accross the board.)

If the variability of a particular subsystem part/protein is low
between various creatures that utilize such a system, then it seems
reasonable to assume that the specificity of this part is fairly high
and visa versa. How would you estimate protein specificity?

> 4. Read the several paragraphs appended after the questions. You do
> not need to answer it paragraph by paragraph (unless you want to), but
> give me your answer to the core problems of ID that are presented
> there.

Ok - I will answer them paragraph by paragraph . . .

> 5. Unanswered from a previous post:
> >If they are not very common, perhaps millions and
> >even billions of years would not be enough time.
>
> Why don't you try to prove it? But instead of calculating meaningless
> numbers of total combinations of amino acids, do it right. Take, for
> example glycogen synthase (a quite distant molecule, mind you), and
> give me a model how many mutations would be required to move from it
> to a lactase. Or if you wish to be fair (and to reduce the amount of
> work you have to do), take any disaccharide hydrolase, and estimate
> the amount of time it would take to evolve a lactose recognition
> domain.

Since these are all relatively simple functions that require
relatively short proteins with fairly low specificity, why should I
knock myself out showing how they could evolve from one place to the
other? I don't think such evolution would be that much of a problem.
However, for systems such as the flagellar motility system, the gaps
between the proposed steppingstones seem to me to be truly enormous -
involving over several thousand non-beneficial steps that must be
crossed somehow. Such gaps, without any obviously beneficial
steppingstone functions in-between, would certainly be impossible for
the mindless processes of random mutation and natural selection to
cross all by themselves. The only logical means to explain their
crossing, that I can see, comes in the form of intelligent design.
There simply is no other logical means that I know of to explain how
such vast gaps, as obviously exists in all living things, were in fact
crossed.

For example, although many scientist suggest that the TTSS system
evolved either independently of or directly from a fully formed
flagellar system (Nguyen et. al., 2000), it is commonly suggested by
evolutionists, such as yourself, that the flagellar system evolved
from a TTSS-like subsystem. The only problem with this is that the
minimum TTSS system requirement is around 6 different types of
proteins while the minimum flagellar system requires over 20 different
types of fairly specified proteins. This leaves a gap of several
thousand fairly specified amino acids, in the form of 12 to 14
proteins in-between the TTSS function and the motility function of the
flagellum. How is such a gap crossed? Even for a huge population of
bacteria the random walk involved here would required trillions upon
trillions of years of average time. Where are the intermediate
steppingstone functions? Several have tried to suggest various
possible steppingstones, but absolutely none of these proposed steps
has been demonstrated to actually evolve in a living thing - not one
step, much less the crossing of the entire gap from TTSS to a
flagellum or any other equivalent level of specified complexity.
Certainly it should be easy enough to set up an experiment to
demonstrate the crossing of one "short" step from one proposed
steppingstone to another? - right? Why hasn't this been done if the
steppingstones are really this close together?

> 6. There are many large, complex proteins whose evolutionary pathway
> has been explained in quite a bit of detail. Pick one, and explain why
> is the proposed evolutionary pathway impossible, which steps are
> unlikely to occur, and why?

I have picked one - the evolution of the flagellar system. If you
want to detail the steps for me here, I will reply to each of your
supposed steppingstones along your pathway and detail why I think they
can or cannot be crossed.

> There, six direct questions. Since the last two require some
> calculations, I will settle for answers to the first four, and you can
> leave the last two for later on.

Before I answer the last two, you will have to do some work yourself.
Since you already claim to know detailed "hypothetical" pathways to
produce such levels of complexity as are found in systems like the
bacterial flagellum, please detail your favorite one here and then I
will respond to it. I have already done this sort of thing many times
with others, such as Ian Musgrave and his ideas and several of the
references that he referred to dealing with the supposed pathway of
flagellar evolution. So, I'd be willing to do the same thing again
for you with an equivalent system of your own choosing.

> ******************
> [What follows is my response to your claim that "You don't understand
> how, but you still have faith that all original causes in this
> universe are mindless". You never answered it before, and I'm really
> interested in seeing what an ID-er thinks about questions presented
> here.]
>
> But evolution applies from the first roughly functioning cell onwards;
> you might be talking about abiogenesis, so let's analyze that for a
> moment. How did the first cell come about?

I'm not talking about abiogenesis at all, but about the limits of
evolution given the pre-established presence of life.

> This is something that is quite unknown, and entirely theoretical. So,
> yes, unlike the case of evolution, one cannot make one's mind based
> entirely on evidence. What do we do?

Hmmmm . . . Yes, abiogenesis is a real quandary now isn't it?

> You might propose that, since we don't know, it is equally rational to
> simply say that there was an Intelligent Designer who designed the
> first cell, as it is to assume that the mindless processes did it. You
> might say that it takes faith in both cases, so there is no real
> difference.

Actually, there is a real difference as I see it. Statistically what
we see is best explained by comparison to what we see intelligent
minds creating (in the form of our own minds). Abiogenesis is simply
ludicrous outside of some sort of deliberate intelligent activity.

> If you did so (and I don't know if you would, but I have a general
> idea), you would be wrong. There is a simple reason to assume that the
> first cell arose by mindless processes rather then by intelligent
> design. The reason is evidence of existence.

Just because something exists does not mean that the automatic
assumption as to its origin or original "cause" has to be "mindless".
That is simply a philosophical position, not a logical necessity at
all.

> We *know*, without doubt, that mindless processes exist.

We also know, without a doubt, that mindful processes exist.

> Everything
> that happens, happens according to the laws of the universe. We have
> ascertained without much doubt that the same laws (without any large,
> visible changes) applied for quite a few billions of years. We know
> that those laws were around when the life first began. Therefore, we
> have good reason to think that the laws of the universe were involved
> in the begining of life.

Involved does not mean entirely responsible. There is nothing that we
see mindless processes doing now that can even remotely begin to
explain the mindless origin of life. The only thing that remotely
produces the level of informational complexity found in living things
is the intelligent mind. Nothing else even comes close. Abiogenesis
is not a logical position at all. It simply does not follow that since
mindless processes exist that they must explain everything or that
they are the most logical answer given the evidence available.

For example, lets say that we create computer robots to the point
where they become self-aware. Say, in a few million years all humans
die off and just these robots are left populating the Earth. Say they
forget about their true origins at the hands of humans. Would it be
automatically logical for them to conclude that their origins were the
result of the mindless interactions of the sands, waters, electricity,
etc., of the Earth? Not at all since no known non-directed processes
ever produce anything beyond the lowest levels of
functional/informational complexity. How then would it be reasonable
to discount the only process that even comes close - their own
intelligence/creative powers? Since this is the only creative force
that even comes close to making something as complex as themselves,
why would it be more unreasonable to propose that a similar or even
greater creative force, in the form of greater mindful complexity,
gave rise to them?

> As for Intelligent Designer...we have no idea if a supernatural entity
> such as he exists or not. We see no evidence of his existence, in the
> present or in the past.

This is the entire debate now isn't it? How blind can you get? The
evidence of design is found in the complete inability of any known
mindless process to even get close! I mean really, if you were to
visit a far distant galaxy and land on an alien planet, what would you
think if you saw a fully functional UFO-style spaceship there? Or, to
be a bit more mysterious, what would you think if you found a
perfectly cut 58-faceted 20-carate diamond there sitting on the top of
a perfectly symmetrical granite cube measuring 1 x 1 x 1 meter? Would
it be the most reasonable thing for you to assume a mindless cause or
a mindful cause. I mean really, you don't see any aliens and you know
no humans have ever been to this planet. All you see are these items
- a think that looks like a fully functional space ship, a granite
cube, and a perfectly cut diamond. How did these things "originate"
on this far distant planet? What is the most reasonable answer your
understanding of your own creative abilities and your knowledge about
the limits of mindless processes to produce such things?

> He is entirely in the domain of imagination,
> and as such we can imagine anything about him, and therefore measure
> nothing. Not only have we no reason to assume his existence, but
> taking him as an "option" would literally mean end of any further
> research [...]

Not at all. Understanding such a being or beings as the only viable
option should only cause the researchers to see more and more clearly
the brilliance of such a mind or minds as they discover more and more
of the astounding complexity of the creations around themselves and
even within themselves.

> How did a star form? If we assume it formed by mindless processes, we
> can use our knowledge of the laws of physics, and see if we can
> discover a solution. After many years of work, we may produce a
> theory, compare it to the reality, and see if it stands up to
> scrutiny. If we assume that a Designer made it, that is it. We give
> up, and say "goddidit", the end. We cannot know anything further.

No. That is not my thinking anyway. In order to understand how an
intelligence might have done it, or at least appreciate such an
intelligence, we need to examine a given phenomenon in very possible
way, to include all the mindless processes involved with that
phenomenon, to include all the possible variations and even potential
creation of that phenomenon by mindless processes. Don't forget,
brilliant men who believed fervently in a God, where still capable of
some of the greatest discoveries in science of all time. This belief
that their was firm evidence for the existence of God throughout
nature did not dull their scientific interests or abilities in the
least. Certainly some people's ideas of God may have in fact stunted
their scientific curiosity as you suggest, but this certainly did not
happen for everyone nor does this need to happen now.

Also, you forget that evolutionism does the same thing as God does for
some. Many evolutionists, when confronted with a problem that they
cannot readily understand, simply say that, "evolutiondidit", the end.
There are those who try to beat their brains out to find out how
evolution did it without ever doubting that it did, but many do not.
They simply take it as a matter of faith that evolution had to have
done it, end of story. A real scientist should not blind him or
herself to any reasonable possibility by deciding, "a priori" that a
mindless cause is the only cause or that a mindful cause is the only
cause. Such a priori assumptions are nothing more than laziness in my
opinion. Each phenomenon should be approached with a mind open to
either or both possibilities as reasonable explanations as the
evidence leads. A priori assumptions are NOT part of the scientific
method and are in fact contrary to it. They are nothing more than
philosophical blinders, not science.

> Through most of the Dark Ages, people took the second approach;
> whatever it is, God made it to be that way. This is, in fact, the
> reason why that period of time is called "The Dark Ages". Modern
> science uses the first approach, and it got as quite a bit further.

Where do you think "modern science" started? With those who were
fervent believers that God did a whole lot of things, just not
everything. Be careful now when you take on the blinders of the
notion that mindless processes do everything and God does nothing.
Such a notion is not better than the God does everything notion. It is
just the opposite extreme to that position and is therefore just as
bad. The best position, in my opinion, is to keep ones self open to
both options. One should think that God may have done it or a
mindless process may have done it - and then go from there to see
where the evidence leads. It is possible to support either position
given the evidence. However, if you exclude one of these
possibilities before you even consider the evidence, you will not be
able to find the truth if the truth happens to be behind your
pre-assumed blinders.

> So, in case of abiogenesis, I will look at both options: accept a
> Designer without *any* reason to do so, plop on a chair, and say
> "done". Or assume that the processes that did [and do] EVERYTHING ELSE
> around me also produced life, and try to figure out how they did it.

The problem with this notion of yours is that there is no one process
that does "everything else" around you as you suggest. Mindless
processes do not give rise to all that you see. Much of what you see
can and is ONLY done by deliberate intelligent design. For example,
look at your house or the cars on the street. Were they formed via
mindless processes? Or, are they evidence of deliberate design?
Likewise, what do the functional systems within living things most
look like? Does anything made by mindless processes (outside of a
pre-established coded system of information or another living thing)
resemble anything found within a living thing beyond the lowest levels
of functional complexity? When was the last time you saw a mindless
process create a flagellum or anything even close? When was the last
time you saw a human create something at least similar to the function
of a flagellum? - such as an outboard motor? Really then, if no
non-living mindless process even comes close while a known mindful
process comes a lot closer to some of these more "simple" levels of
function found in all living things, how is it unreasonable to think
that a mindful cause, at our own level of creativity or beyond, might
actually have been behind such marvelous works?

Sean
www.naturalselection.0catch.com

[zosdad]

Snipping out just a few bits of Sean Pitman's post:

> In short, system specificity is the range to which change within a
> system can occur without a complete loss of the original type of
> beneficial function. Specificity is basically a measure of
> irreducible complexity. Since all types of functions are irreducibly
> complex at some point or another (a minimum part requirement exists
> for all types of functions), all systems have a degree of
> specificity/irreducible complexity.
>
> Obviously then, the greater the variability/changeability, the less
> the specificity, and visa versa. Again, small proteins can be very
> specified (very limited in the number and variability of amino acid
> positions) or loosely specified. Likewise, large proteins can have
> very little specificity (have very few limits on which positions can
> change or what amino acid can occupy a given position) as well.
> However, there can also be large protein systems that maintain a
> fairly high level of specificity (allowing for relatively little
> change, percentage wise, among their amino acid positions). Such
> systems include the multiprotein system of flagellar motility and the
> like. The flagellar system of motility has a high minimum amino acid
> requirement and also maintains a fairly high degree of combined/total
> specificity.

So, how many amino acids are shared between the flagella of bacteria
and archaea? Remember, on your argument, no conservation of aas
implies no specificity...

> For example, although many scientist suggest that the TTSS system
> evolved either independently of or directly from a fully formed
> flagellar system (Nguyen et. al., 2000), it is commonly suggested by
> evolutionists, such as yourself, that the flagellar system evolved
> from a TTSS-like subsystem.

The question is open. There is one paper for (Nguyen et. al., 2000)
and one against (Gophna et al. 2003).

> The only problem with this is that the
> minimum TTSS system requirement is around 6 different types of
> proteins while the minimum flagellar system requires over 20 different
> types of fairly specified proteins. This leaves a gap of several
> thousand fairly specified amino acids, in the form of 12 to 14
> proteins in-between the TTSS function and the motility function of the
> flagellum. How is such a gap crossed?

Simple. The axial proteins (5 rod, 1 hook, 1 flagellin, 2 linkers,
2-3 cap) all share a common ancestor as part of the axial protein
family. ~12 parts reduced to one right there. Flagellin proteins are
still duplicating and diverging, many bacteria have more than one,
some are known to have 6. The L- and P-rings are derived from a
secretin (a large, widespread family involved in many secretion
systems) and its lipoprotein chaperone. Scratch another 2. The motor
proteins (MotA and MotB) are derived from homologs of the
independently-functioning and widely distributed ion channel proteins
homologous to ExbBD and TolQR. Scratch another 2.

We're already to ~15 proteins down, more than your "gap" of 12-14.
For a kicker we can throw in the fact that numerous other parts
included in, say, Dembski's calculation (based on 40-50 proteins) also
have nonflagellar homologs -- FliM is homologous to FliN+CheC (one
down), the Che proteins themselves have homologs in numerous
nonflagellar systems (5+ proteins down), the flagellar sigma factor
has nonflagellar homologs (another one down), MCPs (another 5+ with a
common ancestor and nonflagellar homologs), axial protein chaperones
(probably share a common ancestor if the axial proteins did; another
couple reduced to one). So that's another 14+ proteins of the
"irreducibly complex" flagellum accounted for.

Sure, we don't understand *exactly* how it evolved, but the bits and
pieces of the process are lying all over the place, just like IC says
is impossible because any IC system lacking a part would be
nonfunctional. It looks pretty bad for ID.

> Even for a huge population of
> bacteria the random walk involved here would required trillions upon
> trillions of years of average time.

Please show your work.

> Where are the intermediate
> steppingstone functions?

See above.

> Several have tried to suggest various
> possible steppingstones, but absolutely none of these proposed steps
> has been demonstrated to actually evolve in a living thing - not one
> step, much less the crossing of the entire gap from TTSS to a
> flagellum or any other equivalent level of specified complexity.

Are you telling us that these homologs don't exist? The biochemists
are lying in their peer-reviewed papers?

> Certainly it should be easy enough to set up an experiment to
> demonstrate the crossing of one "short" step from one proposed
> steppingstone to another? - right? Why hasn't this been done if the
> steppingstones are really this close together?

Hmm, maybe we're still trying to figure out how exactly the flagellum
works before we figure out every last detail of how it evolved. You
have to walk before you can run. Unless, that is, you happen to know
what the mechanism of the flagellar motor is, or how type III
secretion functions (with the specific roles of each protein
delineated). This would be a surprise since no one knows these things
in the leading biochemistry departments of the world.

> I have picked one - the evolution of the flagellar system. If you
> want to detail the steps for me here, I will reply to each of your
> supposed steppingstones along your pathway and detail why I think they
> can or cannot be crossed.

http://www.talkdesign.org/faqs/flagellum.html

>
> > There, six direct questions. Since the last two require some
> > calculations, I will settle for answers to the first four, and you can
> > leave the last two for later on.
>
> Before I answer the last two, you will have to do some work yourself.
> Since you already claim to know detailed "hypothetical" pathways to
> produce such levels of complexity as are found in systems like the
> bacterial flagellum, please detail your favorite one here and then I
> will respond to it. I have already done this sort of thing many times
> with others, such as Ian Musgrave and his ideas and several of the
> references that he referred to dealing with the supposed pathway of
> flagellar evolution. So, I'd be willing to do the same thing again
> for you with an equivalent system of your own choosing.
>

While you're at it:

http://www.talkdesign.org/faqs/Evolving_Immunity.html

Focus on the complement system and the origin of rearranging
antibodies. Be sure to examine the long list of peer-reviewed
articles on the origin of each.

[howard hershey]

zosdad wrote:

But that just shows that we have two gaps in which to insert the God of
the Ignorant.

>>For example, although many scientist suggest that the TTSS system
>>evolved either independently of or directly from a fully formed
>>flagellar system (Nguyen et. al., 2000), it is commonly suggested by
>>evolutionists, such as yourself, that the flagellar system evolved
>>from a TTSS-like subsystem.
>
>
> The question is open. There is one paper for (Nguyen et. al., 2000)
> and one against (Gophna et al. 2003).
>
>
>>The only problem with this is that the
>>minimum TTSS system requirement is around 6 different types of
>>proteins while the minimum flagellar system requires over 20 different
>>types of fairly specified proteins. This leaves a gap of several
>>thousand fairly specified amino acids, in the form of 12 to 14
>>proteins in-between the TTSS function and the motility function of the
>>flagellum. How is such a gap crossed?
>
>
> Simple. The axial proteins (5 rod, 1 hook, 1 flagellin, 2 linkers,
> 2-3 cap) all share a common ancestor as part of the axial protein
> family. ~12 parts reduced to one right there. Flagellin proteins are
> still duplicating and diverging, many bacteria have more than one,
> some are known to have 6. The L- and P-rings are derived from a
> secretin (a large, widespread family involved in many secretion
> systems) and its lipoprotein chaperone. Scratch another 2. The motor
> proteins (MotA and MotB) are derived from homologs of the
> independently-functioning and widely distributed ion channel proteins
> homologous to ExbBD and TolQR. Scratch another 2.

Oh, damn. I was just about to make essentially the same points and you
went ahead and beat me to it. [What am I complaining about? I get to
sit around and drink coffee and let you do the heavy lifting.] Of
course, I was going to use a certain article cited below by some guy
named Matzke as a major source in any case.

[zosdad]

howard hershey <hers...@indiana.edu> wrote in message news:<bu9i72$rcj$1...@hood.uits.indiana.edu>...

Howdy there Howard Hershey. In case I didn't mention it before, many
of your t.o. posts over the years were quite helpful in getting me
started on the Big Flagellum Article.

BTW if Sean is feeling ambitious, after tackling the flagellum and
immune system articles he should then rebut the literature on the
evolutionary origins of:

Photosynthesis
http://www.antievolution.org/cgi-bin/ikonboard/ikonboard.cgi?s=4009054319abffff;act=ST;f=9;t=5

Blood-clotting
http://www.antievolution.org/cgi-bin/ikonboard/ikonboard.cgi?s=4009054319abffff;act=ST;f=9;t=3

Multiple-parts required toxin catabolism pathways
http://www.antievolution.org/cgi-bin/ikonboard/ikonboard.cgi?s=4009054319abffff;act=ST;f=9;t=17

And of course the many articles on the origin of the immune system(s)
that have come out since Matt Inlay's article:
http://www.antievolution.org/cgi-bin/ikonboard/ikonboard.cgi?s=4009054319abffff;act=ST;f=9;t=16

Also acceptable would be if Sean gave us references where Behe or
Dembski (1) acknowledged the existence of the literature cited on the
above webpages, (2) critically reviewed it, and (3) successfully
rebutted it. Heck, it would be a substantial improvement if they
would just do (1). Instead, Dembski at least feels that massively
deluded chest-thumping about the lack of literature is the way to go:

=========
Irreducible Complexity Revisited. [14Jan04] Update on where the debate
over irreducible complexity is eight years after the publication of
Michael Behe's book Darwin's Black Box.
http://www.designinference.com/documents/2004.01.Irred_Compl_Revisited.pdf
=========

[Sean Pitman]

niiic...@yahoo.com (zosdad) wrote in message news:<74227462.04011...@posting.google.com>...

> Snipping out just a few bits of Sean Pitman's post:
>
> > In short, system specificity is the range to which change within a
> > system can occur without a complete loss of the original type of
> > beneficial function. Specificity is basically a measure of
> > irreducible complexity. Since all types of functions are irreducibly
> > complex at some point or another (a minimum part requirement exists
> > for all types of functions), all systems have a degree of
> > specificity/irreducible complexity.
> >
> > Obviously then, the greater the variability/changeability, the less
> > the specificity, and visa versa. Again, small proteins can be very
> > specified (very limited in the number and variability of amino acid
> > positions) or loosely specified. Likewise, large proteins can have
> > very little specificity (have very few limits on which positions can
> > change or what amino acid can occupy a given position) as well.
> > However, there can also be large protein systems that maintain a
> > fairly high level of specificity (allowing for relatively little
> > change, percentage wise, among their amino acid positions). Such
> > systems include the multiprotein system of flagellar motility and the
> > like. The flagellar system of motility has a high minimum amino acid
> > requirement and also maintains a fairly high degree of combined/total
> > specificity.
>
> So, how many amino acids are shared between the flagella of bacteria
> and archaea? Remember, on your argument, no conservation of aas
> implies no specificity...

Consider that in some of the most distantly related bacteria (Aquifex
aeolicus, Bacillus subtilis, Escherichia coli, and Treponema pallidum,
Thermotoga maritima), who are thought to share a common ancestor
several billions of years ago, the following 21 genes are all shared:
MotA, MotB, FliG, FliF, FliG, FliM/N, FlhB, FliQ, FliR, FliP, FliI,
FlhA, FlgB, FlgC, FlgG, FliE, FlgE, FlgL, FlgK, FlgD, FliC, and FliD.
Despite billions of years of supposed evolutionary divergence in very
different environments, these 21 genes have not experience significant
permutations. Each of them still shows a fair degree of amino acid
specificity. Also, one of the subsystems of the eubacterial flagellum
is homologous to the TTSS.

On the other hand, the genes that code for the archaea-type flagellum
are very different. The resulting flagellum is somewhat smaller and
resembles class IV pili. All pili, from the more "simple" type I pili
to the more complex type IV pili, are all built from the bottom-up.
So is the archaea-type flagellum. However, the eubacterial flagellum
is built the other way around - from the top down.

Just about everything about these two flagellar systems of motility is
different. They are not related to each other in the least. Even
from the evolutionary perspective, it seems rather hard to argue that
they had any sort of common ancestor since they share almost no gene
homogeny with each other. So, even from your perspective it would
seem that the best that could be argued is that these flagellar
systems evolved with complete independence of each other.

> > For example, although many scientist suggest that the TTSS system
> > evolved either independently of or directly from a fully formed
> > flagellar system (Nguyen et. al., 2000), it is commonly suggested by
> > evolutionists, such as yourself, that the flagellar system evolved
> > from a TTSS-like subsystem.
>
> The question is open. There is one paper for (Nguyen et. al., 2000)
> and one against (Gophna et al. 2003).
>
> > The only problem with this is that the
> > minimum TTSS system requirement is around 6 different types of
> > proteins while the minimum flagellar system requires over 20 different
> > types of fairly specified proteins. This leaves a gap of several
> > thousand fairly specified amino acids, in the form of 12 to 14
> > proteins in-between the TTSS function and the motility function of the
> > flagellum. How is such a gap crossed?
>
> Simple.

This should be good . . .

> The axial proteins (5 rod, 1 hook, 1 flagellin, 2 linkers,
> 2-3 cap) all share a common ancestor as part of the axial protein
> family. ~12 parts reduced to one right there.

Ok, this is the problem as I see it with this little hypothesis of
yours. Starting with a TTSS like system, what axial protein, added to
the TTSS system, by itself, would give rise to the motility function
of the flagellum? Oh, but perhaps all the other required proteins are
already there by some means or another and the only thing left to gain
the motility function are the axial proteins. Lets say that the
"linker" proteins happen to evolve first. Are they, by themselves,
going to be able to gain the motility function of the flagellum? Or,
would a few of the other types of axial proteins be needed as well? I
mean as I try to imagine the problem, I just cannot seem to envision
how a rod protein, without at least a few of the other types of axial
proteins is going to get me to the motility function. How is this
problem solved?

Also, beyond this problem, you seem to significantly oversimplify the
differences between these "12 axial proteins" Which would, I suppose,
be represented by everything from the drive-shaft to the cap,
including FliD, FliC, FlgE, FlgL, FlgK, FlgB, FlgC, FlgG, and FliE.

Lets start with the cap protein (FliD) by first looking at some
studies done with Shigella-type bacteria. Shigella are non-motile
pathogens. Even though Shigella do not express flagella, they do
possess the flagellar operons, suggesting this non-motile state was
recently acquired. Of four strains analyzed, two strains were missing
was FliD, the gene that codes for the protein that caps the flagellar
filament. Obviously then, without this one protein (FliD) no filament
forms. FliD-deficient Shigella mutants are immotile because they lack
flagellar filaments and yet they continue to leak flagellin monomers
out into the medium!

FliD is a very interesting protein indeed, and obviously essential for
filament formation. What happens is that the FliD gene products form
a five-member pentagon-shaped ring that caps the hollow filament
formed by the flagellin (FliC) subunits (which, although "axial" are
very different in form and function). Each member of this pentamer
has a leg-like extension that points downward and interacts snuggly
with the filament. However, there is a symmetry mismatch between the
cap and the filament. The cap is formed from five protein subunits,
but the helical end of the filament itself is formed from 5.5
flagellin subunits. This means that when one protein of the cap
pentamer is at the dislocation point (think of a split washer), it
will be in a very different environment from the other four members of
the pentamer. In other words, a significant crevice is associated with
the cap and end of the filament. And it is proposed that the next
flagellin subunit that gets added to the filament is added to this
crevice. The addition of the new flagellin subunit is then coupled
with the cap itself rotating along the filament axis to open up a new
adjacent crevice. Think of the cap as a split washer (where the center
is filled) sitting on the end of a hollow tube. Individual flagellin
proteins travel down to the tube to be added at the tip. The flagellin
then gets placed into the space of the split washer, the washer turns,
and opens up a new space. Thus, you can envision the cap spinning
around, inserting new flagellin monomers one-at-a-time. The growth
rate of the filament to be about 50 flagellin units/sec. Since there
are 5 subunits per turn of the helical filament, this suggests that
the fliD cap rotates about ten times every second as it incorporates
about 50 flagellin subunits.

Now, the flagellin proteins, as they travel down the hollow filament
tube, are unfolded since the folded protein would have a significant
kink in its middle that would prevent transport through the tube. So,
how are these flagellin proteins folded once the get to the end of the
tube? As it turns out the fliD cap not only serves as a traffic cop,
directing the flagellin into the right spot via its rotating action,
it helps the flagellin to fold as well. In other words, the fliD is a
required chaperone as well as a directing protein. The flagellin
units do not "self-assemble," they are assembled and placed by a
chaperone at a rather impressive rate.

The other axial proteins are also fairly different in both form and
function. For example, the FliE is an unusual protein with respect to
the other proteins that form the rod and beyond. It is the only gene
in its transcriptional unit and alpha-helices run throughout the
protein (alpha-helices in the other rod proteins are restricted to the
terminal ends). Functionally, the FliE is thought to be a "structural
adapter" between the annular symmetry of the M-ring and the helical
symmetry of the rod - connecting the rod to the M-ring. Without the
FliE, the rod and all the subsequent axial structures could not attach
themselves to the M-ring. Also, although the FliE is the only protein
that sticks to the M-ring it does not form filaments either by itself
or in conjunction with any other protein(s). In fact the FliE protein
is so unique that many have suggested that it cannot be called a "rod
protein" since it differs in too many ways from the other rod and
axial proteins. And yet, the FliE is not part of the TTSS system and
has no other homolog. It is flagellum-specific.

Moving onto the "rod/driveshaft" proteins (FlgG, FlgC, and FlgB), the
FlgG is the most distal component of the rod and averages 260aa in
widely divergent species (compare this to FlgC and FlgB which average
130aa respectively). None of these proteins show significant
homology with any other secreted protein or protein used elsewhere in
the cells of eubacteria. These rod proteins are also significantly
different even when compared among themselves. They do share some
similarities, mostly conserved at the N- and C-terminal ends, which
seem to be required for them to fit together (i.e., the C-terminal end
of one protein interacts with the N-terminal end of the next protein).
However, the overall homology is rather insignificant. For example,
when FlgC and FlgB are compared, they show only a 10% overall homology
with each other. But what about FlgG? It is usually assumed that
since FlgG is about twice the size of the other two proteins that it
was formed via gene duplication the "common ancestor" of these other
proteins. But, although sequence comparisons again show fairly strong
N- and C-terminal conservation with the other two proteins, the middle
region of 160aa shows only a 32% identity - not very far about a
purely random identity of 25%.

The conservation of such differences in widely divergent bacteria over
the course of supposed billions of years suggests that they are
functionally important. Very different eubacteria all use these same
three rod proteins with relatively little variation in size or
sequence specificity. Apparently, loss of one gene product cannot be
compensated by the presence of other structurally similar gene
products. This maintenance of all three of these genes in widely
divergent bacterial species over billions of years is especially
interesting in light of the way the helical lattice is arranged.
First, six copies of FlgB form one turn, then six copies of FlgC form
the next turn, then twelve copies of FlgG form two turns . . . and so
on. This sequential order is maintained in all eubacteria. It would
seem to be selectably advantageous, but how is such an ordered process
maintained? If not specified, one would think that the individual
parts could be added at random to the forming lattice. The fact that
they aren't randomly added to the lattice suggests independent
ordering outside of these three proteins themselves. Also, if the
supposed ancestral rod was homogenous, being composed of only one type
of protein, it is difficult to imagine how it might be turned into
such a highly ordered heterogeneous structure which each step being
selectably beneficial.

So, the rod proteins are quite intriguing in the following ways: They
are secreted by the type III export machinery; they have maintained a
relatively uniform size despite long periods of different selective
pressures; they have no function apart from the flagellum; they show
an intriguingly ordered arrangement in forming the rod, and
interestingly enough, they all lack cysteine despite the fact that
many amino acid positions can mutate to cysteine with a single base
pair substitution.

> Flagellin proteins are
> still duplicating and diverging, many bacteria have more than one,
> some are known to have 6.

Given the fact that flagellin (FliC) is constrained by the recognition
of many flagellar elements, such as the cap (FliD), I doubt if these
six FliC variants you speak of are all that different in either size,
structure or sequencing.

> The L- and P-rings are derived from a
> secretin (a large, widespread family involved in many secretion
> systems) and its lipoprotein chaperone. Scratch another 2.

Ok - I'll give you the L- and P-rings for now. But, how do you
explain the interesting construction of the C-ring apparatus? The
first thing laid down in construction of this ring apparatus is the
M-ring (FliF). After the M-ring is formed in the inner membrane, FliG
is added. Finally FliM and FliN are added. The result is the
formation of the C-ring, switch complex, and rotor. What is most
interesting here is that this ring apparatus must form in order to
then form the export apparatus. This is interesting because it is
thought that the export apparatus evolved first and then additions
were made. Also, flagellar FliF, outside of a loose resemblance to
TTSS FliF, has no homologue anywhere with any other gene found in
eubacteria or archaea. So, evolution hypothesizes the evolution of a
flagellum starting with a protein ring that seems to have no other
cellular functions. Likewise, FliG, FliN and FliM have no
flagellum-independent homologues and so it seems that their only
cellular function is flagellum-specific. Also, the union of these
four proteins does not seem to have a function either. FliF, by
itself has no known function and adding the other three proteins to it
does not present an obvious function either. And yet, this four-part
C-ring must be in place before the export apparatus can be added.

> The motor
> proteins (MotA and MotB) are derived from homologs of the
> independently-functioning and widely distributed ion channel proteins
> homologous to ExbBD and TolQR. Scratch another 2.

The flagellar motor is composed of 5 proteins: MotA, MotB, FliG,
FliN, and FliM. The stator is composed of MotA and MotB. The loss of
either one destroys motility in widely different eubacteria. MotA has
four membrane-spanning regions and most of its bulk is found on the
cytoplasmic side of the membrane. MotB has only one membrane spanning
domain and most of its bulk is in the periplasm, where it is anchored
on the underside of the bacterial cell wall. Together these two
proteins form the torque generating unit as a stationary structure
against which the rotor can move and a conducting structure that
directs the flow of ions or protons from the periplasm to the
cytoplasm, which drives the spinning of the flagellum. There are 8
copies of the MotA/MotB protein group that surround the C-ring. Some
models propose that the flow of protons or other ions interact with
the very specifically positioned charged residues on the FliG
component of the C-ring, creating a dynamic electrostatic field that
moves the rotor. The torque generating function of FliG is restricted
to its C-terminal domain. Random mutations or deletions to this
domain (at positions 234, 237, 249, 252, 257, and 306) destroy
flagellar motility completely.

So, the motor apparatus depends on the simultaneous presence of MotA,
MotB and FliG. Remove or significantly alter any one of these three
proteins and the motility function simply vanishes completely. Without
the MotA and MotB, many domains of FliG have no homologous
counterparts and no known secondary function. The MotA/MotB
combination, on the other hand, could plausibly exist as some ion
channel prior to the existence of the flagella, but, despite the
common claims of evolutionists like you, there is no solid evidence
for this.

Also, correct me if I'm wrong, but it seems to me that you are
suggesting that the necessary parts of the flagellar system all came
from preformed parts found working in many different capacities
throughout the cell. Is this correct? If so, this reasoning is
seriously flawed, as I have pointed out numerous times in this forum.
Even if absolutely all of the subparts are present in an organism as
parts of several other systems of function, you still have the problem
of bringing these subparts together to form a new unified flagellar
system of motility function. You see, just because the parts are
there does not mean that they will be able to come together very
easily to form a new unified function all by themselves. In order to
do this the codes for these parts would have to get copied from their
current locations and pasted in other locations so that their
production could be regulated in such a way so that they would
assemble in a new way to form a flagellar motility system. Without
the DNA being able to tell these parts exactly when, where, and how
much to be produced, they simply will not self-assemble themselves
into a flagellar system.

So, the question is how do you copy and paste these pre-established
codes for just the right parts into just the right place so that just
the right beneficial function will be produced? For example, say I
have a book that has tens of thousands of different words in it -
already fully formed. Say that this book would be markedly improved
if it had a new paragraph in a particular place that gets across a
particularly important idea. All the words needed to form this
paragraph are already contained, fully formed, in various sentences
throughout this book. So, according to you, the problem is "simple".
All that needs to happen is that these fully formed words just get
together to form the needed paragraph and *presto* - the problem is
solved!

Now, although that is a very imaginative solution, it really isn't
that easy in real life. Just try it sometime. Try adding words
together and see how far you can go before you into non-beneficial
additions. The same problems arise when you start trying to get
various required parts of a flagellar system to come together and work
in a unique way together from various different places within a
genome.

> We're already to ~15 proteins down, more than your "gap" of 12-14.

As already pointed out, you haven't even come close to showing the
origin of the necessary parts much less how these various parts
somehow come together in some sort of sequencial order where each
addition was selectably beneficial. You still need all the extra
parts listed to be added onto the TTSS system to get the motility
function of the higher emergent function of the flagellar system.
Your argument that these parts or similar "precursor" parts already
exist in the cell does not solve your problem at all even if it were
true. Establishing the existence of the required subparts is only a
tiny part of the problem. Now, you have to explain how these
pre-formed subparts came together to form the flagellum without having
to cross vast neutral gaps. What you have to explain is how the
addition of each subpart, one at a time, was selectably beneficial to
the organism. If the minimum function of the flagellum required the
addition of 10 to 12 subparts onto the TTSS-like system, how long
would it take, on average to get all 12-subparts added if none of the
additions were more beneficial than the original TTSS system?

Now, of course, you will argue that there were in fact beneficial
advantages for the addition of each subpart to the evolving system.
You must argue this because if there were no advantages along the way,
this sort of 12-part addition would truly have required trillions up
trillions of years of average time. So, explain to me how each
addition of a subpart was in fact more beneficial than what came
before - in a stepwise fashion.

> For a kicker we can throw in the fact that numerous other parts
> included in, say, Dembski's calculation (based on 40-50 proteins) also
> have nonflagellar homologs -- FliM is homologous to FliN+CheC (one
> down), the Che proteins themselves have homologs in numerous
> nonflagellar systems (5+ proteins down), the flagellar sigma factor
> has nonflagellar homologs (another one down), MCPs (another 5+ with a
> common ancestor and nonflagellar homologs), axial protein chaperones
> (probably share a common ancestor if the axial proteins did; another
> couple reduced to one). So that's another 14+ proteins of the
> "irreducibly complex" flagellum accounted for.

Besides the fact that over a third of the proteins in the flagellar
system have no known counterparts in other systems of function, this
argument of yours is pretty much irrelevant even if all the required
parts of the flagellum had identical homologues in other systems of
function. Again, you still have the problem of how to bring these
homologues together to form a new and unique cooperative function.
Also, you mention of chaperone proteins just adds another 20 to 30
parts to the system - many of which have no known homologue outside
the flagellar system.

Just to understand what I am talking about here, consider the
formation of the P-pilus, one of the most "simple" pili around. The P
pilus is a very thin filament, whose outer diameter is only about 7 nm
with a hollow core about 2 nm in diameter. The rod is thicker near the
membrane and thins as it nears the tip. It functions as an attachment
organelle. That is, it can reach out and anchor bacterial cells to
other cells. The end of the filament has a protein that specifically
binds to certain sugar molecules found on kidney cells.

Although the P pilus is among the simplest of attachment filaments, it
is encoded by 11 genes. The filament itself is a heterogeneous
structure. The primary subunit is PapA (it forms the thicker rod near
the membrane). But as we get near the tip, we find another protein,
PapE, forms the thinner filament.. At the very end, is PapG, the
specific adhesin that binds to sugars on other cells. PaPG binds to
PaPE through an adaptor protein, PapF. And PapE binds to PapA through
another adaptor protein, PapK. Thus, the pilus itself is composed of
five different proteins that are assembled in a fixed order (PapA -
PapK - PapE-PapK-PapG, proximal to distal).

Now the question is, how is this pilus synthesized in such an orderly
fashion? Like most other pili and adhesive organelles, it starts with
the highly conserved usher/*chaperone* pathway. And here is where
things not only get interesting, but also begin to look very different
from the simplistic account of filament formation usually assumed.

It all begins with the Sec export machinery found in the cytoplasmic
membrane: Protein translocation across the bacterial cytoplasmic
membrane has been studied extensively in Escherichia coli. The
identification of the components involved and subsequent
reconstitution of the purified translocation reaction have defined the
minimal constituents that allowed extensive biochemical
characterization of the so-called translocase. This functional enzyme
complex consists of the SecYEG integral membrane protein complex and a
peripherally bound ATPase, SecA. Under translocation conditions, four
SecYEG heterotrimers assemble into one large protein complex, forming
a putative protein-conducting channel. This tetrameric arrangement of
SecYEG complexes and the highly dynamic SecA dimer together form a
proton-motive force- and ATP-driven molecular machine that drives the
stepwise translocation of targeted polypeptides across the cytoplasmic
membrane.

Also, at this point, I should mention that we're discussing
gram-negative bacteria, which have two membranes. The inner membrane
is a typical cytoplasmic membrane and the outer membrane is more
porous (due to many barrel-shaped protein pores that filter out large
material but allow smaller things like sugars and amino acids inside).
The space between the two membranes is called the periplasm. Transport
via the Sec pathway dumps material into the periplasm. The trick for
the bacteria is to grow this into a filament that penetrates the outer
membrane in a coordinated manner. So how do cells make P pili?

First, you export all the pilus subunits into the periplasm using the
sec-machinery. The proteins are threaded through the sec-machinery in
an unfolded state and most refold in the periplasm. And therein lies
the problem, as the pilus subunits easily form insoluble aggregates
(or clumps) in the periplasm through hydrophobic interactions. To
prevent this, we need to invoke another component, a special chaperone
encoded by PapD. PapD does two things - it binds to the pilus subunits
after they are pumped into the perisplasm and prevents them from
clumping with each other and also helps the pilus subunits to fold
into their proper conformation. In fact, the pilus subunits are not
stable as monomers and exist either as bound to the chaperone or as
bound to each other as part of the filament. The manner in which the
chaperones carry out their function is far more elegant than anyone
assumed, employing something that is now called "donor strand
complementation" (DSC).

The 3-D structures of PapD complexed with PapG (the adhesin on the
tip) and PapK (one of the adaptors) have been solved. PapD forms a
boomerang-shaped protein with two immunoglobulin-like (Ig-like)
domains (a structure composed of layers of antiparallel beta sheets).
The N-terminal end of PapK is also an Ig-like domain, but it lacks a
C-terminal beta sheet that normally contributes to the hydrophobic
core of the domain. This produces a cleft that exposes the hydrophobic
core, which is what makes it so sticky and prone to aggregation by
itself. The chaperone PapD masks this exposed region in a most
fascinating manner - it donates one of its beta strands to complete
the Ig-domain in PapK. But it does so in an atypical fashion, as the
beta strand it donates runs parallel, not antiparallel, with its
neighboring strand. Thus, PapD provides at least two essential
functions captured in one very elegant act - by donating one of its
beta strands, PapD simultaneously prevents aggregation of PapK while
providing the missing steric information for proper folding of PapK.
By themselves, the subunits don't fold properly and are unstable. The
steric information for proper folding is not found in a single amino
acid chain or gene, but in two distinct chains/genes. And by itself,
PapD has no function.

What happens next? The pilus subunit-chaperone complex interacts with
a protein channel on the outer membrane, PapC (also known as the
usher). The channel is large enough to accommodate the tip of the
filament, but not the rod. The actual mechanism of incorporation is
being worked out, as the chaperone somehow hands off the pilus subunit
to the usher for incorporation into the growing filament. Interaction
between the usher and chaperone-pilus subunit does not result in the
chaperone-subunit complex breaking apart, thus the mechanism of
handoff is also probably quite sophisticatedly complex as well.

But there is one more feature to the story worth mentioning. The pilus
subunits themselves are thought to form a filament through a donor
strand complementation-like mechanism. Each pilus subunit has an
N-terminal extension that does not contribute to its own folding. By
itself, it is a disordered strand. However, it has been proposed that
this N-terminal extension from one subunit (let's call it A) displaces
the displaces the donated chaperone strand associated with another
pilus subunit (B). This N-terminal strand would then form a beta
strand that runs in an antiparallel direction and complete the
Ig-domain of its neighbor in a typical fashion. Again, the steric
information for the Ig-domain of subunit B is supplied from subunit A.
This mechanism is called donor strand exchange. And the result is that
the filament is made by linking subunits, where each subunit
contributes a strand to perfectly complete the fold of its nearest
neighbor.

> Sure, we don't understand *exactly* how it evolved,

Now that's the understatement of the century!

> but the bits and
> pieces of the process are lying all over the place, just like IC says
> is impossible because any IC system lacking a part would be
> nonfunctional. It looks pretty bad for ID.

Not at all. Even if all your assumptions about homology were somehow
true, you still wouldn't even be close to explaining how the flagellum
formed via mindless processes alone.

> > Several have tried to suggest various
> > possible steppingstones, but absolutely none of these proposed steps
> > has been demonstrated to actually evolve in a living thing - not one
> > step, much less the crossing of the entire gap from TTSS to a
> > flagellum or any other equivalent level of specified complexity.
>
> Are you telling us that these homologs don't exist? The biochemists
> are lying in their peer-reviewed papers?

Many of them do not exist. But, even if they all existed and each of
the required parts had a homologue somewhere else as a part or parts
of various non-related systems of function, this still isn't even
close to explaining how these parts where joined together to make an
entirely new type of collective function.

> > Certainly it should be easy enough to set up an experiment to
> > demonstrate the crossing of one "short" step from one proposed
> > steppingstone to another? - right? Why hasn't this been done if the
> > steppingstones are really this close together?
>
> Hmm, maybe we're still trying to figure out how exactly the flagellum
> works before we figure out every last detail of how it evolved.

We aren't talking about fine details here. We are talking the
evolution of even one step from anyone proposed steppingstone function
to any other. Not even one of your proposed steps has ever been shown
to actually evolve outside of the just-so story telling of
evolutionist's imaginations.

> You
> have to walk before you can run. Unless, that is, you happen to know
> what the mechanism of the flagellar motor is, or how type III
> secretion functions (with the specific roles of each protein
> delineated). This would be a surprise since no one knows these things
> in the leading biochemistry departments of the world.

And yet you are absolutely confident that the flagellum evolved? How
so? You don't even really know how it works and yet you are sure it
evolved simply because you find a few homologous structures in other
systems of function? I find this sort of blind faith simply amazing!

Anyway, this is all I have time for right now. Your stuff about the
evolution of cascades is not relevant to my position. I have covered
my reasons for this extensively in other posts if you care to look
them up. In short, cascades are not nearly as complex as systems
where each of the parts is required to work together at the same time
- as in the flagellar system of motility.

Sean
www.naturalselection.Ocatch.com

Primary resources used for above discussion:

http://www.idthink.net/biot/flag1/
http://www.bmb.psu.edu/courses/micro401/Wk3Nts.htm
http://www.millerandlevine.com/km/evol/design2/article.html
http://www.errantskeptics.org/William_Dembski_Response.htm

[howard hershey]

Sean Pitman wrote:

How do you know this? By the amount of sequence identity, right? That
would mean that these genes represent a very ancient ancestry, right?
And any differences seen are differences due to either neutral change
*from* the ancestral sequences or due to selection *for* a specific
local reason. Ancient ancestry always means that the amount of evidence
available will be reduced.

> Despite billions of years of supposed evolutionary divergence in very
> different environments, these 21 genes have not experience significant
> permutations. Each of them still shows a fair degree of amino acid
> specificity.

Wouldn't that be expected for systems that are related by common
descent? But we are not interested in modern differences except to the
extent that those differences can show the range of possibilities that
could allow the eubacterial flagella to arise initially.

> Also, one of the subsystems of the eubacterial flagellum
> is homologous to the TTSS.

And the existence of subsystems which use some, but not all, of the
proteins that are used in flagella, or that are closely related to
flagellar proteins are exactly the types of evidence that support the
existence of subsystems with independent utility. So are the evidence
of the ways that the flagella develop ontologically.

> On the other hand, the genes that code for the archaea-type flagellum
> are very different. The resulting flagellum is somewhat smaller and
> resembles class IV pili. All pili, from the more "simple" type I pili
> to the more complex type IV pili, are all built from the bottom-up.
> So is the archaea-type flagellum. However, the eubacterial flagellum
> is built the other way around - from the top down.
>
> Just about everything about these two flagellar systems of motility is
> different. They are not related to each other in the least.

That is true. But both *are* systems that use a rotary whip as a
motility device. That is, there are at least two independent ways to
generate a motility system based on rotary movement of a whip. The
systems are analogous, not homologous. But they do show, by comparing
the systems, that some features are *unneccessarily* complex in one or
the other system to generate the same *function*. In particular, the
whip system of eubacteria are unnecessarily complex. One does not
*need* a whip composed of 12 different proteins to generate a functional
whip. That is evidence that one *can* start with a simpler system (even
one with a single protein) and, by duplication and divergence, generate
a more complex (even an irreducibly complex) whip with different
proteins that are more specialized in their function even if it is for
the final functionality of mobility. Each step can result in a
selectably *improved* whip, but not a change in function. And it is
quite possible to go from a redundantly complex state of mere
duplication to an irreducibly complex state where both components have
become necessary. For example, one can easily go from a state
where hemoglobin is a homotetramer of alpha globin (hagfish) to a state
where it is a heterotetramer of alpha and beta globins (most other
vertebrates) and loss of either one causes the collapse of its
hemoglobin function.

> Even
> from the evolutionary perspective, it seems rather hard to argue that
> they had any sort of common ancestor since they share almost no gene
> homogeny with each other. So, even from your perspective it would
> seem that the best that could be argued is that these flagellar
> systems evolved with complete independence of each other.

No. From our perspective it is absolutely certain that the systems
evolved independently. Just as, from your perspective, it must be that
they were indepedently designed. From a perspective of independent
evolution, all that is needed is that both function to improve motility
and survival in the organism. From a perspective of design by the
*same* designer, there should be a reason why one design is used for
eubacteria and a completely different one for archae. Can you come up
with such a reason if the common designer could just as easily plugged
the same sequences in either archae or eubacteria? What made it
necessary to use entirely different flagella in these two clades?

>
>>>For example, although many scientist suggest that the TTSS system
>>>evolved either independently of or directly from a fully formed
>>>flagellar system (Nguyen et. al., 2000), it is commonly suggested by
>>>evolutionists, such as yourself, that the flagellar system evolved
>>>from a TTSS-like subsystem.
>>
>>The question is open. There is one paper for (Nguyen et. al., 2000)
>>and one against (Gophna et al. 2003).
>>
>>
>>>The only problem with this is that the
>>>minimum TTSS system requirement is around 6 different types of
>>>proteins while the minimum flagellar system requires over 20 different
>>>types of fairly specified proteins. This leaves a gap of several
>>>thousand fairly specified amino acids, in the form of 12 to 14
>>>proteins in-between the TTSS function and the motility function of the
>>>flagellum. How is such a gap crossed?
>>
>>Simple.
>
>
> This should be good . . .
>
>
>>The axial proteins (5 rod, 1 hook, 1 flagellin, 2 linkers,
>>2-3 cap) all share a common ancestor as part of the axial protein
>>family. ~12 parts reduced to one right there.
>
>
> Ok, this is the problem as I see it with this little hypothesis of
> yours. Starting with a TTSS like system, what axial protein, added to
> the TTSS system, by itself, would give rise to the motility function
> of the flagellum?

Remember my little comment about duplication and divergence leading from
a redundantly complex to an irreducibly complex state, like happened
with the globin genes of hemoglobin? If you take *modern* hemoglobin in
*modern* vertebrates (except hagfish) and remove either alpha or beta
globin, you no longer have functionally useful hemoglobin activity in
these modern organisms. The same would be true for all the ancestral
organisms down to the point where the initial duplication occurred and
alpha and beta globins starting specializing. At some point in this
proceess, one would indeed have a state where one could remove the
incipient beta globin (or the ancestral alpha) and have a functional
hemoglobin because at that point the two proteins are not sufficiently
evolved away from one another. At that point they would be redundantly
complex and not irreducibly complex.

One should not confuse what happens when one mutates or loses a
functional gene in a modern organismal system with what happened during
the evolution of a particular organismal system (although Behe does this
all the time).

> Oh, but perhaps all the other required proteins are
> already there by some means or another and the only thing left to gain
> the motility function are the axial proteins. Lets say that the
> "linker" proteins happen to evolve first. Are they, by themselves,
> going to be able to gain the motility function of the flagellum? Or,
> would a few of the other types of axial proteins be needed as well? I
> mean as I try to imagine the problem, I just cannot seem to envision
> how a rod protein, without at least a few of the other types of axial
> proteins is going to get me to the motility function. How is this
> problem solved?

By starting with a single protein that can produce a whip, as happened
in the independent evolution of a rotary system in archae and as
happened (although possibly by a loss mechanism) in producing the
injectosomes of certain TTSS systems.

>
> Also, beyond this problem, you seem to significantly oversimplify the
> differences between these "12 axial proteins" Which would, I suppose,
> be represented by everything from the drive-shaft to the cap,
> including FliD, FliC, FlgE, FlgL, FlgK, FlgB, FlgC, FlgG, and FliE.
>
> Lets start with the cap protein (FliD) by first looking at some
> studies done with Shigella-type bacteria. Shigella are non-motile
> pathogens. Even though Shigella do not express flagella, they do
> possess the flagellar operons, suggesting this non-motile state was
> recently acquired. Of four strains analyzed, two strains were missing
> was FliD, the gene that codes for the protein that caps the flagellar
> filament. Obviously then, without this one protein (FliD) no filament
> forms. FliD-deficient Shigella mutants are immotile because they lack
> flagellar filaments and yet they continue to leak flagellin monomers
> out into the medium!

Remember the caution about what happens with duplication and
specialization. After specialization, the two proteins are an
interacting system, not an ancestral multi-functional protein and a
modern one. Just as losing beta globin in the evolved IC system that is
hemoglobin does not result in the *modern* alpha globin immediately
having the capacity to form homotetramers, even though the existence of
hagfish show that it is possible.

>
> FliD is a very interesting protein indeed, and obviously essential for
> filament formation.

In *modern* eubacteria.

> What happens is that the FliD gene products form
> a five-member pentagon-shaped ring that caps the hollow filament
> formed by the flagellin (FliC) subunits (which, although "axial" are
> very different in form and function). Each member of this pentamer
> has a leg-like extension that points downward and interacts snuggly
> with the filament. However, there is a symmetry mismatch between the
> cap and the filament. The cap is formed from five protein subunits,
> but the helical end of the filament itself is formed from 5.5
> flagellin subunits. This means that when one protein of the cap
> pentamer is at the dislocation point (think of a split washer), it
> will be in a very different environment from the other four members of
> the pentamer.

Something very similar happens with the self-assembly of TMV coat
proteins. Disks are formed initially, but by interacting with RNA one
starts forming split washers and then a helical tube.

> In other words, a significant crevice is associated with
> the cap and end of the filament. And it is proposed that the next
> flagellin subunit that gets added to the filament is added to this
> crevice. The addition of the new flagellin subunit is then coupled
> with the cap itself rotating along the filament axis to open up a new
> adjacent crevice. Think of the cap as a split washer (where the center
> is filled) sitting on the end of a hollow tube. Individual flagellin
> proteins travel down to the tube to be added at the tip. The flagellin
> then gets placed into the space of the split washer, the washer turns,
> and opens up a new space. Thus, you can envision the cap spinning
> around, inserting new flagellin monomers one-at-a-time. The growth
> rate of the filament to be about 50 flagellin units/sec. Since there
> are 5 subunits per turn of the helical filament, this suggests that
> the fliD cap rotates about ten times every second as it incorporates
> about 50 flagellin subunits.

And the similarity between the two proteins is good evidence that they
evolved into their specialized roles by duplication and divergence
rather than by independent evolution from scratch. The fact that the
pair is now IC is a natural *consequence* of co-evolving specialization.

> Now, the flagellin proteins, as they travel down the hollow filament
> tube, are unfolded since the folded protein would have a significant
> kink in its middle that would prevent transport through the tube. So,
> how are these flagellin proteins folded once the get to the end of the
> tube? As it turns out the fliD cap not only serves as a traffic cop,
> directing the flagellin into the right spot via its rotating action,
> it helps the flagellin to fold as well. In other words, the fliD is a
> required chaperone as well as a directing protein. The flagellin
> units do not "self-assemble," they are assembled and placed by a
> chaperone at a rather impressive rate.

Assembly in the presence of a chaparone is as much self-assembly as is
assembly of TMV in the presence of RNA or microtubules in the presence
of Ca or angry cattle in the presence of the chaparone from hell. It is
not unusual for a very similar protein to act as a kind of seed in
self-assembly. Look at the human lactase, for example, where most of
the protein (the amino end) acts as a chaparone and then gets clipped
off. That amino end looks like a duplication.

> The other axial proteins are also fairly different in both form and
> function. For example, the FliE is an unusual protein with respect to
> the other proteins that form the rod and beyond. It is the only gene
> in its transcriptional unit and alpha-helices run throughout the
> protein (alpha-helices in the other rod proteins are restricted to the
> terminal ends). Functionally, the FliE is thought to be a "structural
> adapter" between the annular symmetry of the M-ring and the helical
> symmetry of the rod - connecting the rod to the M-ring. Without the
> FliE, the rod and all the subsequent axial structures could not attach
> themselves to the M-ring. Also, although the FliE is the only protein
> that sticks to the M-ring it does not form filaments either by itself
> or in conjunction with any other protein(s). In fact the FliE protein
> is so unique that many have suggested that it cannot be called a "rod
> protein" since it differs in too many ways from the other rod and
> axial proteins. And yet, the FliE is not part of the TTSS system and
> has no other homolog. It is flagellum-specific.

It has no homolog in *modern* eubacteria. But, as we both agree that
modern flagella has a very ancient history, with 21 of the proteins
being preserved over a billion years, putting the *original* source of
the FliE in some ancient bacteria, it is perhaps more surprising that we
have been able to identify so many homologs (related families of
proteins) that are parts of flagellar subsystems that are also parts of
other independently useful systems than that there is one protein for
which we cannot identify a homolog. Is the FliE protein the best you
have for a protein that poofed into existence from some random sequence
rather than was modified from some pre-existing protein? Based, of
course, on the absence of evidence rather than evidence that it must
have arisen from a random sequence.

> Moving onto the "rod/driveshaft" proteins (FlgG, FlgC, and FlgB), the
> FlgG is the most distal component of the rod and averages 260aa in
> widely divergent species (compare this to FlgC and FlgB which average
> 130aa respectively). None of these proteins show significant
> homology with any other secreted protein or protein used elsewhere in
> the cells of eubacteria. These rod proteins are also significantly
> different even when compared among themselves. They do share some
> similarities, mostly conserved at the N- and C-terminal ends, which
> seem to be required for them to fit together (i.e., the C-terminal end
> of one protein interacts with the N-terminal end of the next protein).

Remember the discussion of the rate of neutral drift. We are talking
about a very ancient set of proteins here. And if there were a few
differences positively *selected for* in the middle of these proteins
(conservative selection would retain the ends precisely for the reasons
you point out), I would certainly expect most of the observed
differences to be in the middle of these proteins. The *similarities*
between FlgG and FlgG of different bacteria will partly be due to
selection for these different functional specializations. The
differences between FlgG and FlgC will be partly due to specialization
and partly due to neutral drift.

> However, the overall homology is rather insignificant. For example,
> when FlgC and FlgB are compared, they show only a 10% overall homology
> with each other. But what about FlgG? It is usually assumed that
> since FlgG is about twice the size of the other two proteins that it
> was formed via gene duplication the "common ancestor" of these other
> proteins. But, although sequence comparisons again show fairly strong
> N- and C-terminal conservation with the other two proteins, the middle
> region of 160aa shows only a 32% identity - not very far about a
> purely random identity of 25%.
>
> The conservation of such differences in widely divergent bacteria over
> the course of supposed billions of years suggests that they are
> functionally important.

No. The *differences* are uninformative wrt to whether they are a
consequence of selection or drift *because* we are talking about
billions of years that these proteins existed. The *similarities* of
different proteins that serve the *same* function is informative of
which amino acids serve a selectively useful function.

> Very different eubacteria all use these same
> three rod proteins with relatively little variation in size or
> sequence specificity.

Which shows selective advantage. But, the key point is how much
variation in sequence do we see for the proteins that serve the *same*
function and how deep is the split? If hagfish did not exist, one could
have made the same argument that alpha and beta globin, because they
form an IC system, must have been poofed into existence all at once.

> Apparently, loss of one gene product cannot be
> compensated by the presence of other structurally similar gene
> products. This maintenance of all three of these genes in widely
> divergent bacterial species over billions of years is especially
> interesting in light of the way the helical lattice is arranged.
> First, six copies of FlgB form one turn, then six copies of FlgC form
> the next turn, then twelve copies of FlgG form two turns . . . and so
> on. This sequential order is maintained in all eubacteria. It would
> seem to be selectably advantageous, but how is such an ordered process
> maintained?

Not by the emanations and intelligent actions-at-a-distance you seem to
think arise from DNA. Nor is it likely that these proteins are
generated one at a time by the cell and that the cell's DNA counts out 6
FlgB, then 6 FlgC then 12 copies of FlgG waiting until all the previous
copies get transported. Generating the proper ratio of proteins is a
snap, of course, even for proteins transcribed off the same operon.
This is a problem of physiology and biochemistry and how signals are
transmitted. I can think of several possible ways this sequential order
could happen. None of them involve DNA or mRNA transcription rates.
Either there is some sort of feedback system that allows discrimination
in which protein gets exported from the base or there is a self-assembly
change that occurs when a full (or 2 in the case of FlgG) helix is
completed that leads to preferentially binding the next type of protein.
The latter would be wasteful, but cells are not exactly unwasteful.

> If not specified, one would think that the individual
> parts could be added at random to the forming lattice. The fact that
> they aren't randomly added to the lattice suggests independent
> ordering outside of these three proteins themselves. Also, if the
> supposed ancestral rod was homogenous, being composed of only one type
> of protein, it is difficult to imagine how it might be turned into
> such a highly ordered heterogeneous structure which each step being
> selectably beneficial.
>
> So, the rod proteins are quite intriguing in the following ways: They
> are secreted by the type III export machinery;

Expected if they derive by duplication and divergence.

> they have maintained a
> relatively uniform size despite long periods of different selective
> pressures;

Expected by the conservative nature of selection if size is important
(such as important in forming the right type of helix).

> they have no function apart from the flagellum;

In *modern* organisms. With the exception that they can still act as
transport tubes in TTSS-like transport.

> they show
> an intriguingly ordered arrangement in forming the rod,

Which is an interesting physiological problem, but doesn't necessarily
say much about the original ancestral formation. In any case, it is
highly unlikely that some sort of supernatural intervention is necessary
for this 'intriguingly ordered arrangement' to occur. I rather doubt
that God himself is counting off the number and order of transport of
the flagellar proteins to form a rod in every single bacterial
flagellum. And DNA surely isn't.

> and
> interestingly enough, they all lack cysteine despite the fact that
> many amino acid positions can mutate to cysteine with a single base
> pair substitution.

Since cysteine is involved in forming rigid links between different
parts of a folded protein to maintain a specific shape immediately after
synthesis, it is highly likely that there is selection against it in
these proteins that have to change shape after transport and be flexible
beforehand.

See Sean propose absence of evidence as evidence. See Sean propose
personal increduility as evidence that Goddoneit. See Sean propose
ignorance as brilliant argument.

Yes. FliG is involved in linking one subsystem to another. Gee whiz.
Have you read Nic Matzke's article? So exactly which step in the
proposed evolution of the flagella do you think cannot have happened?
Clearly we are no longer talking about the brain-dead idea that all 21
proteins in the modern flagella have to poof into existence by a random
walk from 21 different random sequences at this point.

> Also, correct me if I'm wrong, but it seems to me that you are
> suggesting that the necessary parts of the flagellar system all came
> from preformed parts found working in many different capacities
> throughout the cell. Is this correct? If so, this reasoning is
> seriously flawed, as I have pointed out numerous times in this forum.

Assertion, assertion, assertion.

> Even if absolutely all of the subparts are present in an organism as
> parts of several other systems of function, you still have the problem
> of bringing these subparts together to form a new unified flagellar
> system of motility function. You see, just because the parts are
> there does not mean that they will be able to come together very
> easily to form a new unified function all by themselves. In order to
> do this the codes for these parts would have to get copied from their
> current locations and pasted in other locations so that their
> production could be regulated in such a way so that they would
> assemble in a new way to form a flagellar motility system. Without
> the DNA being able to tell these parts exactly when, where, and how
> much to be produced, they simply will not self-assemble themselves
> into a flagellar system.

DNA is a dumb, dumb, dumb molecule and a very passive one. It does not
*tell* any part of a cell anything beyond what sequence it's proteins
will have. DNA does not *tell* a protein when it will be made (the
environment tells, via a chain of proteins, DNA when it will be
transcribed). DNA does not *tell* a protein where it will be made (at
most, some mRNAs may have sequences that will direct it to different
ribosomes but more usual is that proteins have sequences that will
direct them to different locations post-translationally; neither
mRNA nor protein are DNA, AFAICT). DNA, by itself, does not *tell* how
much protein is going to be made (the interaction between regulatory
DNA-binding proteins and short regulatory sequences tells DNA how much
mRNA will be transcribed and the nature of the interaction of the mRNA
and the environment determines how much of the transcribed mRNA gets
processed and translated; the environment is what determines how much
of the protein made gets processed and used). DNA, aside from its
effect in determining a protein's sequence, plays no or very little role
in the assembly of higher order protein structures. Protein/protein
interactions (as a consequence of sequence and post-translational
processing) and protein/environment interactions (including interaction
with allosteric effectors and post-translational modifications) does.
Other than those points, you are right. [That was a joke.]

> So, the question is how do you copy and paste these pre-established
> codes for just the right parts into just the right place so that just
> the right beneficial function will be produced? For example, say I
> have a book that has tens of thousands of different words in it -
> already fully formed. Say that this book would be markedly improved
> if it had a new paragraph in a particular place that gets across a
> particularly important idea. All the words needed to form this
> paragraph are already contained, fully formed, in various sentences
> throughout this book. So, according to you, the problem is "simple".
> All that needs to happen is that these fully formed words just get
> together to form the needed paragraph and *presto* - the problem is
> solved!

Why do you think that a gene needs to be in a particular place in the
genome in order to have a beneficial effect? If a cell duplicates a
gene it is also going to be duplicating the nearby regulatory sequences.
And modifying regulatory sequences so that there is different
regulation is easy even by your random walk from a random sequence
standard. Most regulatory regions are only 6-10 nucleotides long and
their position wrt distance from the coding sequence is also relatively
flexible. To produce an operon where different genes encoded on the
mRNA are translated at widely different rates is easy. Many bacterial
operons do this.

> Now, although that is a very imaginative solution, it really isn't
> that easy in real life. Just try it sometime. Try adding words
> together and see how far you can go before you into non-beneficial
> additions. The same problems arise when you start trying to get
> various required parts of a flagellar system to come together and work
> in a unique way together from various different places within a
> genome.

At the level of genes, of course, it doesn't matter where they are in
the genome. The question of higher order structure is one that plays
out at the level of protein-protein interaction or protein-environment
interaction. Interacting between different protein subsystems is what
is involved here.

>>We're already to ~15 proteins down, more than your "gap" of 12-14.
>
>
> As already pointed out, you haven't even come close to showing the
> origin of the necessary parts much less how these various parts
> somehow come together in some sort of sequencial order where each
> addition was selectably beneficial. You still need all the extra
> parts listed to be added onto the TTSS system to get the motility
> function of the higher emergent function of the flagellar system.

Be more specific. *Which* proteins cannot be attributed to a subsystem
that has independent utility? Which steps do you think are unexplained?
How do you quantitate this "emergent function" as being too large a
change in function?

> Your argument that these parts or similar "precursor" parts already
> exist in the cell does not solve your problem at all even if it were
> true. Establishing the existence of the required subparts is only a
> tiny part of the problem. Now, you have to explain how these
> pre-formed subparts came together to form the flagellum without having
> to cross vast neutral gaps.

Have you read Nic's essay? The question is "What vast neutral gaps?"
The parts of the flagella that is attributed to a TTSS-like ancestor
does what as part of its role in the flagella? Ans: It acts as a
specialized protein export devise. What did it do before it took this
role in the flagella? Ans: It acted as a protein export device. So it
took the huge leap from being a protein export device to being a
modified, slightly different export device. What role did the motor
proteins play in cells before it became associated with the TTSS-like
system? Ans: The motor proteins of all the related non-flagellar
proteins form ion channels that energize work at a distance by a third
protein. What does the motor protein do in the flagella? Ans: The
motor proteins of flagella form ion channels that energize work at a
distance by a third protein. Now THAT is what I call a huge increase in
functional complexity. The motor proteins went from being proteins that
form ion channels that energize work at a distance by a third protein to
being proteins that form ion channels that energize work at a distance
by a third protein (FliG). That must have involved thousands of fairly
specified amino acid changes to cross such a "vast neutral gap" in
functional complexity.

> What you have to explain is how the
> addition of each subpart, one at a time, was selectably beneficial to
> the organism. If the minimum function of the flagellum required the
> addition of 10 to 12 subparts onto the TTSS-like system, how long
> would it take, on average to get all 12-subparts added if none of the
> additions were more beneficial than the original TTSS system?

Why do all the parts have to be added one at a time? The motor probably
wasn't. Nor is it likely that the TTSS-like element was. And the
changes in some of the related flagellar proteins may well have largely
arisen *after* the emergence of the motility function rather than
before. The rod protein and flagellin protein divergence seems to be
deeper than changes within rod proteins or within flagellins. But it is
quite possible to envision a non-motile system with a rod that was
functionally useful. Again, you seem to be considering motility as the
only possible function that can be selected for. That is teleological
thinking.

> Now, of course, you will argue that there were in fact beneficial
> advantages for the addition of each subpart to the evolving system.
> You must argue this because if there were no advantages along the way,
> this sort of 12-part addition would truly have required trillions up
> trillions of years of average time. So, explain to me how each
> addition of a subpart was in fact more beneficial than what came
> before - in a stepwise fashion.

Explaining it is easy. But evidence to support each of these steps is
harder and depends on what nature gives us. The available data are
certainly consistent with the step-wise (but not necessarily one protein
at a time, entire subsystems can be added at one step) explanations
given. But the flagella is a very ancient structure. So it is not
surprising that the evidence to support some steps (such as a putative
homolog from which a FliG protein might have arisen) is missing (at
least in the modern bacteria analysed; future work in different bacteria
might find one).

>>For a kicker we can throw in the fact that numerous other parts
>>included in, say, Dembski's calculation (based on 40-50 proteins) also
>>have nonflagellar homologs -- FliM is homologous to FliN+CheC (one
>>down), the Che proteins themselves have homologs in numerous
>>nonflagellar systems (5+ proteins down), the flagellar sigma factor
>>has nonflagellar homologs (another one down), MCPs (another 5+ with a
>>common ancestor and nonflagellar homologs), axial protein chaperones
>>(probably share a common ancestor if the axial proteins did; another
>>couple reduced to one). So that's another 14+ proteins of the
>>"irreducibly complex" flagellum accounted for.
>
>
> Besides the fact that over a third of the proteins in the flagellar
> system have no known counterparts in other systems of function,

Which ones? And are you counting each of the flagellin whip and rod
proteins as a different protein with no known counterpart? Despite the
homology at the amino and carboxyl ends which indicate that they are, as
a group, their own counterparts (arose by duplication and divergence)?

> this
> argument of yours is pretty much irrelevant even if all the required
> parts of the flagellum had identical homologues in other systems of
> function. Again, you still have the problem of how to bring these
> homologues together to form a new and unique cooperative function.

That's just it. A TTSS-like ancestral system is not a set of useless
junk. Adding an injectosome-like tubular structure does not suddenly
make a TTSS-system useless junk. The fact is that subsystems already
formed independently and then come together, just as happens with
metabolic pathways.

> Also, you mention of chaperone proteins just adds another 20 to 30
> parts to the system - many of which have no known homologue outside
> the flagellar system.

Which ones? Are you saying that these chaparones have no use to other
proteins or other use at all? Despite your pointing out that the
flagellar cap protein also acts like a chaparone (does it count as one
protein or two)?

[snip a long digression into another system, the P-pilus, that was equal
parts goalpost moving (if not the bacterial flagella, then the P-pilus),
hand-waving numerology (there are 11, count 'em, 11, proteins in this
pilus system, each with hundreds of "fairly specified amino acids"
resulting in a monster leap into "high functional complexity" that
couldn't happen in billyuns and billyuns of years because there must be
several hundred thousand selectively neutral steps that must occur),
personal incredulity (gee whizz, this looks complex and I can't think of
how it could evolve, therefore godidit), and the funny idea, promoted by
Behe, that the hodge-podge rube goldberg-like devices of life somehow
represents a sure sign of elegant design.]

>
>
>>Sure, we don't understand *exactly* how it evolved,
>
>
> Now that's the understatement of the century!
>
>
>>but the bits and
>>pieces of the process are lying all over the place, just like IC says
>>is impossible because any IC system lacking a part would be
>>nonfunctional. It looks pretty bad for ID.
>
>
> Not at all. Even if all your assumptions about homology were somehow
> true, you still wouldn't even be close to explaining how the flagellum
> formed via mindless processes alone.

The bits and pieces have independent utility. They are not functionless
random bits and pieces. Interestingly enough, the functions that these
proteins play *within* the flagellar system is also not randomly
different than the independent utility they are thought to have had (or
known to have) outside the flagella. Yet you call what happened a large
change in 'functional complexity'. Could you explain how you determined
that the change in 'functional complexity' was large? What was the
starting and what was the ending point in your mathematical
determination? You still haven't told us how you know this.

>>>Several have tried to suggest various
>>>possible steppingstones, but absolutely none of these proposed steps
>>>has been demonstrated to actually evolve in a living thing - not one
>>>step, much less the crossing of the entire gap from TTSS to a
>>>flagellum or any other equivalent level of specified complexity.

What gap in "level of specified complexity" do you think needs to be
crossed in one swell foop?

>>Are you telling us that these homologs don't exist? The biochemists
>>are lying in their peer-reviewed papers?
>
>
> Many of them do not exist.

'Many' meaning a third of the flagellar proteins (counting those that
are duplicates and divergences of others of that third?)? This in a
truely ancient structure in systems only a small minority of which have
been explored?

> But, even if they all existed and each of
> the required parts had a homologue somewhere else as a part or parts
> of various non-related systems of function, this still isn't even
> close to explaining how these parts where joined together to make an
> entirely new type of collective function.

Stepwise, with some of the steps involving the integration of
pre-exisiting *subsystems*.

>>>Certainly it should be easy enough to set up an experiment to
>>>demonstrate the crossing of one "short" step from one proposed
>>>steppingstone to another? - right? Why hasn't this been done if the
>>>steppingstones are really this close together?
>>
>>Hmm, maybe we're still trying to figure out how exactly the flagellum
>>works before we figure out every last detail of how it evolved.
>
>
> We aren't talking about fine details here. We are talking the
> evolution of even one step from anyone proposed steppingstone function
> to any other. Not even one of your proposed steps has ever been shown
> to actually evolve outside of the just-so story telling of
> evolutionist's imaginations.

And you have nothing but personal incredulity, hand-waving
pseudomathimatical analysis, and vague definitions used as the cross is
used against vampires (words as talismans).

>> You
>>have to walk before you can run. Unless, that is, you happen to know
>>what the mechanism of the flagellar motor is, or how type III
>>secretion functions (with the specific roles of each protein
>>delineated). This would be a surprise since no one knows these things
>>in the leading biochemistry departments of the world.
>
>
> And yet you are absolutely confident that the flagellum evolved? How
> so? You don't even really know how it works and yet you are sure it
> evolved simply because you find a few homologous structures in other
> systems of function? I find this sort of blind faith simply amazing!

What alternative testable explanation have you provided that has more
supporting evidence?

> Anyway, this is all I have time for right now. Your stuff about the
> evolution of cascades is not relevant to my position. I have covered
> my reasons for this extensively in other posts if you care to look
> them up. In short, cascades are not nearly as complex as systems
> where each of the parts is required to work together at the same time
> - as in the flagellar system of motility.

Yet the evolution of different flagellins is no different than the
evolution of different kinase kinases in a cascade. The only difference
is that one phosphylates a similar protein sequence and the other binds
to a similar protein sequence.

[zosdad]

seanpi...@naturalselection.0catch.com (Sean Pitman) wrote in message news:<80d0c26f.04011...@posting.google.com>...

C'mon Sean, think hard now, what might this indicate about the
evolution of flagella-like things?

> All pili, from the more "simple" type I pili
> to the more complex type IV pili, are all built from the bottom-up.

Except, of course, Type III pili. There are about 4 other kinds of
pili, and two of them (Type II and Type IV) are homologous to each
other. So you've really got about 3 built from the bottom and 1 from
the top. This is not exactly an impressive sample from which to get
all excited about a pilus built from the top...

> So is the archaea-type flagellum. However, the eubacterial flagellum
> is built the other way around - from the top down.

...just like type III pili, e.g. the Hrp pilus.

> Just about everything about these two flagellar systems of motility is
> different.

In other words, none of the specificity you are so required about is
specifically required for the building of prokaryote rotary swimming
structures ("flagella").

> They are not related to each other in the least.

I agree. That's my point: you can't say that flagella are a tiny,
impossible-to-hit target when the evidence indicates that there are
many different ways to build flagella.

You obviously didn't read the article proposing the model under
discussion. {erhaps you'd like another shot. It's here:

http://www.talkdesign.org/faqs/flagellum.html
Before the flagellum you have a pilus, it's the addition of Mot
proteins (cooption of a whole subsystem) that makes it a
proto-flagellum, not the addition of another axial component. The
duplication and divergence of (most) specialized axial components
takes place after the origin of motility. Specialized hooks, linkers,
etc. are helpful but not logically required for crude motility. For
some reason this simple point -- which is obvious to anyone who thinks
about the question of flagellar origins for a moment -- has been
completely missed by Behe, Dembski, your latest source Mike Gene, and
now you yourself. You, Sean, have the distinction of making the
mistake even in the face of a model that makes the point explicitly.
A singular achievement, I must say...

> Oh, but perhaps all the other required proteins are
> already there by some means or another and the only thing left to gain
> the motility function are the axial proteins.

Wow, total incomprehension of the model under discussion...

> Lets say that the
> "linker" proteins happen to evolve first.

Why? They didn't. You don't get to disprove evolution by dreaming up
a preposterous strawman and refuting that.

> Are they, by themselves,
> going to be able to gain the motility function of the flagellum?

No, silly. The reason to have linkers is to interface between
divergence axial subunits, e.g. the rod, the hook, and the filament.
You don't need linkers at all if you are starting with an immotile
Type III pilus as the model does. It could be made entirely of one
subunit (as type III pili today may to be, although they may have some
minor subunits -- at any rate they have far fewer distinct axial
proteins than flagella). But duplication and divergence of genes is a
common process, followed by specialization of function. Howard has
already given you the hemoglobin example, other references are in the
flagellum article. Since divergence can occur gradually, as two axial
proteins diverged -- specializing for their roles inside the secretion
system (pilus anchor) and outside the cell (adhesive pilus) --
mismatch would not immediately be an issue. But as specialization
continued, a minor mismatch would be created, which would create
selection pressure for another mutant duplicate to strengthen the
interface. Once the proto-linker is in place, more specialization can
occur and the process will repeat. Somewhere early in this process
the MotAB proteins are coopted to convert the pilus into a primitive
flagellum, and this creates a whole new set of selection pressures for
improved motility, and only after this would you get the development
of a hook (probably from the proto-linker between rod and pilus), and
as the hook diverges, 2 linkers between the hook and the flagellin
(one linker derived from each) and the hook and the rod (one linker,
from the hook-rod subfamily of the axial family).

Quite prosaic evolutionary processes the whole way through, and yet
you IDists have been assuming all along that all of the axial proteins
must have been independently coopted together at once in some kind of
tornado-in-a-junkyard scenario.

> Or,
> would a few of the other types of axial proteins be needed as well? I
> mean as I try to imagine the problem, I just cannot seem to envision
> how a rod protein, without at least a few of the other types of axial
> proteins is going to get me to the motility function. How is this
> problem solved?

An excellent restatement of the Argument From Personal Incredulity and
the Argument From I Haven't Read the Article That Explains This So I'm
Going To Assume a Magical Explanation Instead.

I can tell how this is going to go already...
http://www.talkdesign.org/faqs/flagellum.html

> Also, beyond this problem, you seem to significantly oversimplify the
> differences between these "12 axial proteins" Which would, I suppose,
> be represented by everything from the drive-shaft to the cap,
> including FliD, FliC, FlgE, FlgL, FlgK, FlgB, FlgC, FlgG, and FliE.

Hey, it's not I who put them in a family (and the hook-rod proteins in
a subfamily), it's the peer-reviewed lit. cited in the article.

> Lets start with the cap protein (FliD) by first looking at some
> studies done with Shigella-type bacteria. Shigella are non-motile
> pathogens. Even though Shigella do not express flagella, they do
> possess the flagellar operons, suggesting this non-motile state was
> recently acquired. Of four strains analyzed, two strains were missing
> was FliD, the gene that codes for the protein that caps the flagellar
> filament. Obviously then, without this one protein (FliD) no filament
> forms. FliD-deficient Shigella mutants are immotile because they lack
> flagellar filaments and yet they continue to leak flagellin monomers
> out into the medium!
>
> FliD is a very interesting protein indeed, and obviously essential for
> filament formation.

Funny, the non-essentialness of the cap is discussed at length right
here: http://www.talkdesign.org/faqs/flagellum.html#filament

As discussed in the article, even Robert Macnab, the world's leading
expert, didn't consider the cap an absolutely required component.
I'll quote it so that you don't have to go to the trouble to click on
the link...

"It might be objected at this point that the flagellum requires the
cap (FliD) in order to chaperone the flagellin subunits into place at
the elongating tip of the filament; without it, they diffuse away and
are lost (Blocker et al., 2003). The hook has its own temporary cap
(FlgD), and it has been suggested, but not proven (Hirano et al.,
2001; Berg, 2003; Macnab, 2003), that the rod has a cap protein as
well (FlgJ). However, the necessity of the cap for successfully
assembling subunits is ambiguous. Flagellin will self-assemble into
filaments in vitro (Hirano et al., 2001). No cap has been identified
in any type III virulence systems (Blocker et al., 2003), and although
PrgJ has been suggested as a possible cap for the Salmonella needle
(Sukhan et al., 2003), the evidence is indeterminate as Sukhan et al.
could not detect PrgJ in sheared-off needles and did not detect it at
needle tips using immunoelectron microscopy (they therefore suggest
that PrgJ may be a basal component). The polar flagellum of Vibrio
grows normally without the cap (Bardy et al., 2003), probably because
it is sheathed by an extension of the cell membrane (McCarter, 2001)
that constrains the subunits. Finally, even in the canonical E. coli
flagellum the adaptor proteins FlgK and FlgL are added without any
capping structure (Macnab, 2003), leading Macnab (2003) to argue that
"capping structures are perhaps best viewed as a means of increasing
efficiency of addition rather than as an absolute requirement." On
this view, the cap could be a relatively late evolutionary addition to
the pilus structure, originating by pentamerization of a pilus subunit
and initially improving speed and efficiency of pilus assembly. Later
co-adaptation between filament and cap subunits would make it a
more-or-less required feature."

So nice when a critic's objections have been anticipated and answered
ahead of time...just cut'n paste...

> What happens is that the FliD gene products form
> a five-member pentagon-shaped ring

<snip long lecture on how the cap works, I'm already well aware of
this...>

> The other axial proteins are also fairly different in both form and
> function. For example, the FliE is an unusual protein with respect to
> the other proteins that form the rod and beyond.

FliE is the most divergent of the lot. It's the only one that has to
bind to proteins of non-axial origin so this is not surprising. But
Type III pili don't seem to need a FliE-like protein and manage to
associate with the FliF homolog just fine. What is so unlikely about
a duplicate rod protein gradually being selected for better and better
binding to the FliF ring, reducing the chance of detachment as the
speed of motor rotation is gradually improved by optimization of the
motor proteins? It's a simple optimization problem, on the order of
complexity of Dawkins' "METHINKS" simulation.

> It is the only gene
> in its transcriptional unit and alpha-helices run throughout the
> protein (alpha-helices in the other rod proteins are restricted to the
> terminal ends). Functionally, the FliE is thought to be a "structural
> adapter" between the annular symmetry of the M-ring and the helical
> symmetry of the rod - connecting the rod to the M-ring. Without the
> FliE, the rod and all the subsequent axial structures could not attach
> themselves to the M-ring.

So, how do Type III pili do it?

> Also, although the FliE is the only protein
> that sticks to the M-ring it does not form filaments either by itself
> or in conjunction with any other protein(s). In fact the FliE protein
> is so unique that many have suggested that it cannot be called a "rod
> protein" since it differs in too many ways from the other rod and
> axial proteins. And yet, the FliE is not part of the TTSS system and
> has no other homolog. It is flagellum-specific.

All easily explained by gradual divergence. And, if you're not happy
with the explanation, we can put it to the test once we solve the
structures of the basal axial proteins. Based on the model I predict
structural homology with the rod proteins. What do you predict?

> Moving onto the "rod/driveshaft" proteins (FlgG, FlgC, and FlgB), the
> FlgG is the most distal component of the rod and averages 260aa in
> widely divergent species (compare this to FlgC and FlgB which average
> 130aa respectively). None of these proteins show significant
> homology with any other secreted protein or protein used elsewhere in
> the cells of eubacteria.

Except for structural similarity to Type III pilins (see article for
ref).

> These rod proteins are also significantly
> different even when compared among themselves. They do share some
> similarities, mostly conserved at the N- and C-terminal ends, which
> seem to be required for them to fit together (i.e., the C-terminal end
> of one protein interacts with the N-terminal end of the next protein).
> However, the overall homology is rather insignificant. For example,
> when FlgC and FlgB are compared, they show only a 10% overall homology
> with each other.

Like I said, your argument is not with me here, but with the
scientists who put all of these rod proteins not just in a family but
a subfamily.

> But what about FlgG? It is usually assumed that
> since FlgG is about twice the size of the other two proteins that it
> was formed via gene duplication the "common ancestor" of these other
> proteins. But, although sequence comparisons again show fairly strong
> N- and C-terminal conservation with the other two proteins, the middle
> region of 160aa shows only a 32% identity - not very far about a
> purely random identity of 25%.

25% is a long ways from "purely random" -- think about it, there are
20 amino acids, so 5% is what would be really "random". The real
issue is the chance of getting a "false positive" in a sequence
database, and the scientific consensus is that anything over 30% is
very, very unlikely to be a false positive. Homologies are commonly
established on 20% identity or less, however, and subsequently
verified by structural comparisons (e.g., tubulin and ftsZ).

But even if the sequence similarity were lower in the middle region (I
thought it was lower in some cases, but it's been awhile since I
looked into it), so what? What is so wildly unlikely about a gene
duplicating and then fusing, and the middle regions of the
now-twice-as-big-protein (lacking the stabilizing selective pressure
to interface with adjacent proteins) diverging more rapidly?
Increased sequence similarity decay under decreased functional
constraints is well documented.

>
> The conservation of such differences in widely divergent bacteria over
> the course of supposed billions of years suggests that they are
> functionally important. Very different eubacteria all use these same
> three rod proteins with relatively little variation in size or
> sequence specificity. Apparently, loss of one gene product cannot be
> compensated by the presence of other structurally similar gene
> products. This maintenance of all three of these genes in widely
> divergent bacterial species over billions of years is especially
> interesting in light of the way the helical lattice is arranged.
> First, six copies of FlgB form one turn, then six copies of FlgC form
> the next turn, then twelve copies of FlgG form two turns . . . and so
> on. This sequential order is maintained in all eubacteria.

Establishing this in "all eubacteria" is a far from trivial exercise,
given that barely a handful are at all well studied. But regardless,
the conservation of the proteins would be easily explained if they are
adapted to coassemble with the L- and P-rings. This is not a logical
necessity for a crude proto-flagellum, but once such an interaction is
set up to increase assembly efficiency it would be hard to undo
without messing up assembly altogether.

> It would
> seem to be selectably advantageous, but how is such an ordered process
> maintained? If not specified, one would think that the individual
> parts could be added at random to the forming lattice. The fact that
> they aren't randomly added to the lattice suggests independent
> ordering outside of these three proteins themselves.
> Also, if the
> supposed ancestral rod was homogenous, being composed of only one type
> of protein, it is difficult to imagine how it might be turned into
> such a highly ordered heterogeneous structure which each step being
> selectably beneficial.

Um, remember gene duplication? Specialize each duplicate to interact
with either FliE, the P-ring, or the L-ring, and you're there. More
of the same-ol', same-ol.

> So, the rod proteins are quite intriguing in the following ways: They
> are secreted by the type III export machinery; they have maintained a
> relatively uniform size despite long periods of different selective
> pressures;

Well, except for FliE, which is smaller, and FlgG, which is a fused
homodimer...

> they have no function apart from the flagellum;

Logically, since they arose after the origin of a proto-flagellum as
optimizations, not as in the tornado-in-a-junkyard strawman model that
you like to beat up on.

> they show
> an intriguingly ordered arrangement in forming the rod, and
> interestingly enough, they all lack cysteine despite the fact that
> many amino acid positions can mutate to cysteine with a single base
> pair substitution.

Whoa! Selection keeps out cysteine mutations so that the proteins
fold right! What awesome power it has!

>
> > Flagellin proteins are
> > still duplicating and diverging, many bacteria have more than one,
> > some are known to have 6.
>
> Given the fact that flagellin (FliC) is constrained by the recognition
> of many flagellar elements, such as the cap (FliD), I doubt if these
> six FliC variants you speak of are all that different in either size,
> structure or sequencing.

Flagellin tolerates huge variations, you can chop out whole domains
and maintain function. A lot of the variation in flagellin within and
among bacteria is used to avoid being recognized by that other
beautifully-intelligently designed, irreducibly complex biochemical
machine, the immune system. Nice of the designer to design his
flagellum with the ability to avoid the designs of his designed immune
systems, eh?

> > The L- and P-rings are derived from a
> > secretin (a large, widespread family involved in many secretion
> > systems) and its lipoprotein chaperone. Scratch another 2.
>
> Ok - I'll give you the L- and P-rings for now. But, how do you
> explain the interesting construction of the C-ring apparatus? The
> first thing laid down in construction of this ring apparatus is the
> M-ring (FliF). After the M-ring is formed in the inner membrane, FliG
> is added. Finally FliM and FliN are added. The result is the
> formation of the C-ring, switch complex, and rotor. What is most
> interesting here is that this ring apparatus must form in order to
> then form the export apparatus. This is interesting because it is
> thought that the export apparatus evolved first and then additions
> were made.

Probably it is assembled this way as an assembly checkpoint -- it is
best to make sure the chemotaxis apparatus is fully assembled before
spending a large amount of energy synthesizing flagellin. This is no
guarantee that the original protoflagellum functioned this way. Since
FlhA/FlhB appear to be the major proteins that grab the substrate,
it's not clear that there is any logical requirement for FliG or
FliM/N in secretion. The logical requirements are (1) a pore (FliF),
(2) a substrate selector (FlhA and maybe FlhB) if you want substrate
specificity, and (3) an ATPase if you want the transport to be active
rather than passive (some/most of the rest of the type III export
apparatus). The continuity between somewhat general passive pores,
more specific passive pores, and more specific active pores should be
obvious. There are a gajillion pore proteins, membrane transport
proteins, etc. exhibiting a continuum of complexity from single
proteins to type II, III and IV secretion systems. See refs in
article.

At the moment we don't know for sure that the roles of FliG and FliM/N
in secretion are basically minor and that their requirement is just an
assembly checkpoint, but the hypothesis makes several testable
predictions (such as vestigiality in type III virulence systems). We
shall see.

> Also, flagellar FliF, outside of a loose resemblance to
> TTSS FliF, has no homologue anywhere with any other gene found in
> eubacteria or archaea.

Yawn, another "Sure, you've found the homologs of A, B, C and D, but
you haven't found the homologs of Y or Z yet." This is of the same
caliber as "Well, you've found transitional fossils for the origin of
birds, mammals, whales, and humans, but not for the origin of
primitive chordates in the precambrian. We've got you now!!" Just
another argument from ignorance, and particularly weak in this case
since this protein is noncatalytic and the oldest one in the lot on
the model.

Besides, Dembski already conceded that if the flagellum can be
explained by evolution from a primitive type III secretion system,
then the IC argument has been busted and it is reasonable to think
that the type III secretion system can be similarly explained:

"Provided the bacterial flagellum can be explained in Darwinian
fashion in terms of several simpler systems, the presumption is that
these simpler systems can in turn be explained in Darwinian fashion in
terms of still simpler systems, and so on until we reach a level of
simplicity that does not require a Darwinian explanation at all. In
consequence, it is enough for Matzke to show that the origin of the
bacterial flagellum can be explained in terms of an evolving sequence
of simpler systems."

(http://www.designinference.com/documents/2003.11.Matzke_Response.htm
)

> So, evolution hypothesizes the evolution of a
> flagellum starting with a protein ring that seems to have no other
> cellular functions.

Google on "transport system database" and then tell me again about the
functions of protein rings...

> Likewise, FliG, FliN and FliM have no
> flagellum-independent homologues and so it seems that their only
> cellular function is flagellum-specific.

Careful, FliM is basically a FliN + a domain derived from CheC, which
has nonflagellar homologs. FliG and FliN are found in type III
virulence systems although to me it's debatable whether or not they
would have been in an ancestral preflagellar type III secretion
system. In the paper I make some suggestions about where to look for
nonflagellar homologs based on these proteins' interactions with
flagellar proteins with known nonflagellar homologs. Remember,
homologs based on conserved tertiary structure are discovered all the
time even though sequence similarity has decayed to nonsignificance.

> Also, the union of these
> four proteins does not seem to have a function either. FliF, by
> itself has no known function and adding the other three proteins to it
> does not present an obvious function either. And yet, this four-part
> C-ring must be in place before the export apparatus can be added.

Which gets us back to
possible-current-checkpoint-mechanism-doesn't-imply-original-requirement-for-secretion.
Or perhaps you can tell us and all the biochemists in the world --
what, exactly, do FliN/M and FliG *do* in secretion? Without knowing
this it we won't know whether "FliN/M and FliG were part of the
ancestral T3SS" or "FliN/M and FliG were added at the origin of
(respectively) chemotaxis and motility as described in Matzke's (2003)
model" is more likely.

> > The motor
> > proteins (MotA and MotB) are derived from homologs of the
> > independently-functioning and widely distributed ion channel proteins
> > homologous to ExbBD and TolQR. Scratch another 2.
>
> The flagellar motor is composed of 5 proteins: MotA, MotB, FliG,
> FliN, and FliM. The stator is composed of MotA and MotB.

IMO the "motor" is MotA & B, everything else is driveshaft. But
whatever, these are just convenient words pasted onto biological
reality.

> The loss of
> either one destroys motility in widely different eubacteria. MotA has
> four membrane-spanning regions and most of its bulk is found on the

> cytoplasmic side of the membrane. MotB has only one membrane spanning...
<snip boring lecture on how it works, we already know this>

> There are 8
> copies of the MotA/MotB protein group that surround the C-ring. Some
> models propose that the flow of protons or other ions interact with
> the very specifically positioned charged residues on the FliG
> component of the C-ring, creating a dynamic electrostatic field that
> moves the rotor.

Ah, been reading Mike Gene I see. These models are out of favor, see
Berg (2003) and Macnab (2003). The proton channel is probably
entirely *within* MotAB.

> The torque generating function of FliG is restricted
> to its C-terminal domain. Random mutations or deletions to this
> domain (at positions 234, 237, 249, 252, 257, and 306) destroy
> flagellar motility completely.

...and the usual thing with these experimentally induced point
mutations is that many of them are reparable by compensatory mutations
in FliG or other components. Protein interactions are more like a
puzzle where you are trying to evolve two text strings to match each
other, than they are like Dawkins' METHINKS simulation. There is no
"one goal" with the "correct" sequence, rather the goal is something
more general like "complimentary shapes" where both sides of the
interface can mutate. Evolving a binding site between two proteins is
mere microevolution, proteins evolve new binding sites all the time in
the wild and in the lab.

> So, the motor apparatus depends on the simultaneous presence of MotA,
> MotB and FliG. Remove or significantly alter any one of these three
> proteins and the motility function simply vanishes completely. Without
> the MotA and MotB, many domains of FliG have no homologous
> counterparts and no known secondary function. The MotA/MotB
> combination, on the other hand, could plausibly exist as some ion
> channel prior to the existence of the flagella, but, despite the
> common claims of evolutionists like you, there is no solid evidence
> for this.

Unfortunately the Mike Gene essay that you cribbed this from with
minor modifications is badly out of date. You would have known this
if you'd read this section and looked up the prominently references on
the nonflagellar homologs of MotAB. This is just one example of
Gene's essays being shown up by newly published evidence.

http://www.talkdesign.org/faqs/flagellum.html#primitive

> Also, correct me if I'm wrong, but it seems to me that you are
> suggesting that the necessary parts of the flagellar system all came
> from preformed parts found working in many different capacities
> throughout the cell. Is this correct? If so, this reasoning is
> seriously flawed, as I have pointed out numerous times in this forum.
> Even if absolutely all of the subparts are present in an organism as
> parts of several other systems of function, you still have the problem
> of bringing these subparts together to form a new unified flagellar
> system of motility function. You see, just because the parts are
> there does not mean that they will be able to come together very
> easily to form a new unified function all by themselves. In order to
> do this the codes for these parts would have to get copied from their
> current locations and pasted in other locations so that their
> production could be regulated in such a way so that they would
> assemble in a new way to form a flagellar motility system.

All of this gene duplication and regulatory evolution has happened in
short order in the evolution of 2,4-DNT degradation, which you accept,
so what's the problem?

Please, pretty please, don't fall back into the tornado-in-a-junkyard
strawman, the evolutionary model shows just how the flagellum can
evolve by the cooption of **one subsystem at a time** with
optimization in-between and afterwards.

> Without
> the DNA being able to tell these parts exactly when, where, and how
> much to be produced, they simply will not self-assemble themselves
> into a flagellar system.

Too bad you missed the part of the article that references this
article:

Kalir, S., McClure, J., Pabbaraju, K., Southward, C., Ronen, M.,
Leibler, S., Surette, M. G. and Alon, U., 2001. Ordering genes in a
flagella pathway by analysis of expression kinetics from living
bacteria. Science. 292 (5524), 2080-2083.

...look it up and see what they say about the "specificity" of the
regulation of flagellar components.

> So, the question is how do you copy and paste these pre-established
> codes for just the right parts into just the right place so that just
> the right beneficial function will be produced? For example, say I
> have a book that has tens of thousands of different words in it -
> already fully formed. Say that this book would be markedly improved
> if it had a new paragraph in a particular place that gets across a
> particularly important idea. All the words needed to form this
> paragraph are already contained, fully formed, in various sentences
> throughout this book. So, according to you, the problem is "simple".
> All that needs to happen is that these fully formed words just get
> together to form the needed paragraph and *presto* - the problem is
> solved!

Tsk, tsk, you're back to tornado-in-a-junkyard.

>
> Now, although that is a very imaginative solution, it really isn't
> that easy in real life. Just try it sometime. Try adding words
> together and see how far you can go before you into non-beneficial
> additions. The same problems arise when you start trying to get
> various required parts of a flagellar system to come together and work
> in a unique way together from various different places within a
> genome.

One subsystem at a time, Sean, one subsystem at a time.

> > We're already to ~15 proteins down, more than your "gap" of 12-14.
>
> As already pointed out, you haven't even come close to showing the
> origin of the necessary parts much less how these various parts
> somehow come together in some sort of sequencial order where each
> addition was selectably beneficial. You still need all the extra
> parts listed to be added onto the TTSS system to get the motility
> function of the higher emergent function of the flagellar system.
> Your argument that these parts or similar "precursor" parts already
> exist in the cell does not solve your problem at all even if it were
> true. Establishing the existence of the required subparts is only a
> tiny part of the problem. Now, you have to explain how these
> pre-formed subparts came together to form the flagellum without having
> to cross vast neutral gaps.

What vast gaps? The "gap" in between each major stage is bridged by
the evolution of a new binding site between two subsystems. Do you
really think that the evolution of a protein-protein binding site is
that hard?

> What you have to explain is how the
> addition of each subpart, one at a time, was selectably beneficial to
> the organism.

Tornado-in-a-junkyard strawman once again! ***EVOLUTION DOESN'T HAVE
TO ADD PARTS ONE-AT-A-TIME, IT CAN JOIN TWO PREEXISTING SUBSYSTEMS,
THEN LATER ON ADD ANOTHER SUBSYSTEM, ETC.!!!***

If you want pretty pictures of IDists' misunderstanding of this point,
versus the actual evolutionary model, see

http://www.millerandlevine.com/km/evol/design1/article.html

Please compare your strawman:
http://www.millerandlevine.com/km/evol/design1/Image1.gif

With evolutionary theory:
http://www.millerandlevine.com/km/evol/design1/Image2.gif

Here they are side-by-side for easy reference:
http://www.millerandlevine.com/km/evol/design1/two-models.jpg

> If the minimum function of the flagellum required the
> addition of 10 to 12 subparts onto the TTSS-like system,

It didn't. Who says it did? The minimum (dispersal) function of the
flagellum only logically requires a type III pilus, to which gets
attached the motor subsystem.

> how long
> would it take, on average to get all 12-subparts added if none of the
> additions were more beneficial than the original TTSS system?

A garbage-in, garbage out calculation.

> Now, of course, you will argue that there were in fact beneficial
> advantages for the addition of each subpart to the evolving system.
> You must argue this because if there were no advantages along the way,
> this sort of 12-part addition would truly have required trillions up
> trillions of years of average time. So, explain to me how each
> addition of a subpart was in fact more beneficial than what came
> before - in a stepwise fashion.

Boy, you have completely forgotten about the part where most of those
12 parts diverge from a common axial filament ancestor *after* the
origin of crude motility.

> > For a kicker we can throw in the fact that numerous other parts
> > included in, say, Dembski's calculation (based on 40-50 proteins) also
> > have nonflagellar homologs -- FliM is homologous to FliN+CheC (one
> > down), the Che proteins themselves have homologs in numerous
> > nonflagellar systems (5+ proteins down), the flagellar sigma factor
> > has nonflagellar homologs (another one down), MCPs (another 5+ with a
> > common ancestor and nonflagellar homologs), axial protein chaperones
> > (probably share a common ancestor if the axial proteins did; another
> > couple reduced to one). So that's another 14+ proteins of the
> > "irreducibly complex" flagellum accounted for.
>
> Besides the fact that over a third of the proteins in the flagellar
> system have no known counterparts in other systems of function,

Namely, the axial proteins which are all related to each other and to
the type III pilin...

> this
> argument of yours is pretty much irrelevant even if all the required
> parts of the flagellum had identical homologues in other systems of
> function. Again, you still have the problem of how to bring these
> homologues together to form a new and unique cooperative function.

Excuse me while I go bang my head against the wall...

> Also, you mention of chaperone proteins just adds another 20 to 30
> parts to the system - many of which have no known homologue outside
> the flagellar system.

Now you're just in outer space, there just a few additional chaperone
proteins. All of those MCPs and Che proteins are chemotaxis/signal
transduction, and yet are still counted by Dembski although not by
Mike Gene who is rather quicker on the uptake. Many of the regulatory
proteins also have extraflagellar homologs, e.g. one paper that I
actually missed in the original flagellum article is this one:

Cases I, Ussery DW, de Lorenzo V. The sigma54 regulon (sigmulon) of
Pseudomonas putida. Environ Microbiol. 2003 Dec; 5(12): 1281-93.

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=14641574&dopt=Abstract

...therein it is revealed that the flagellar sigma factor is not
unique either. Before this paper you could have counted it as an
component without homologs, but now you can't. That's the danger of
arguments from absence of evidence...

> Just to understand what I am talking about here, consider the
> formation of the P-pilus, one of the most "simple" pili around.

This argument didn't make any sense when Mike Gene made it, and it
doesn't make any sense when you rehash it. Just because this
particular pilus is (marginally) simpler than some others says very
little about the complexity of the first pilus based on a type III
secretion system.

<snip>

> Thus, the pilus itself is composed of
> five different proteins that are assembled in a fixed order (PapA -
> PapK - PapE-PapK-PapG, proximal to distal).

Well, in the type III virulence apparatus, the "pilus itself" (the
filament bit outside the cell) may well be made up of only 1-2
proteins. See Blocker et al. 2003.

> Now the question is, how is this pilus synthesized in such an orderly
> fashion? Like most other pili and adhesive organelles, it starts with
> the highly conserved usher/*chaperone* pathway.

Who cares? There are a lot of different secretion systems and a lot
of different pili based on them. I'm quite sure that the
proteobacterial systems reviewed in the flagellum paper only scratch
the surface of the diversity of prokaryote "sticky-outy bits".

<snippety-snip-snip, go read http://idthink.net/biot/flag2/index.html
if you want to read about the P pilus for some reason...>

> > Sure, we don't understand *exactly* how it evolved,
>
> Now that's the understatement of the century!

The 20th century, maybe, but not the 21st...

> > but the bits and
> > pieces of the process are lying all over the place, just like IC says
> > is impossible because any IC system lacking a part would be
> > nonfunctional. It looks pretty bad for ID.
>
> Not at all. Even if all your assumptions about homology were somehow
> true, you still wouldn't even be close to explaining how the flagellum
> formed via mindless processes alone.

How would you know? You're not even close to understanding (or even,
I think, reading) the model made available to you.

> > > Several have tried to suggest various
> > > possible steppingstones, but absolutely none of these proposed steps
> > > has been demonstrated to actually evolve in a living thing - not one
> > > step, much less the crossing of the entire gap from TTSS to a
> > > flagellum or any other equivalent level of specified complexity.
> >
> > Are you telling us that these homologs don't exist? The biochemists
> > are lying in their peer-reviewed papers?
>
> Many of them do not exist. But, even if they all existed and each of
> the required parts had a homologue somewhere else as a part or parts
> of various non-related systems of function, this still isn't even
> close to explaining how these parts where joined together to make an
> entirely new type of collective function.

Heck, you said 2/3s do exist, and the other 1/3 are axial proteins
that share a common ancestor. The model under discussion reduces the
origin of the flagellum to a series of subsystem-linking events that
are implemented in each case by the evolution of a binding site
between two proteins. Thus in order to maintain your much vaunted
"gaps" you (and Dembski) have to deny that a protein can evolve a
binding site.

The big gaps have been reduced to little, regular, crossed-every-day
gaps. This is pretty much the point of an evolutionary model for a
complex system.

> > > Certainly it should be easy enough to set up an experiment to
> > > demonstrate the crossing of one "short" step from one proposed
> > > steppingstone to another? - right? Why hasn't this been done if the
> > > steppingstones are really this close together?
> >
> > Hmm, maybe we're still trying to figure out how exactly the flagellum
> > works before we figure out every last detail of how it evolved.
>
> We aren't talking about fine details here. We are talking the
> evolution of even one step from anyone proposed steppingstone function
> to any other. Not even one of your proposed steps has ever been shown
> to actually evolve outside of the just-so story telling of
> evolutionist's imaginations.

Hmm, this sure means alot for a model that's about two months old, for
a system that is still incompletely understood, and even then only in
a few model organisms out of a biosphere full of uncharacterized
prokaryotes. What's important is testability. Experiments are but
one form of test.

Speaking of imaginative storytelling, give us your explanation for the
flagellum and propose some empirical tests.

> > You
> > have to walk before you can run. Unless, that is, you happen to know
> > what the mechanism of the flagellar motor is, or how type III
> > secretion functions (with the specific roles of each protein
> > delineated). This would be a surprise since no one knows these things
> > in the leading biochemistry departments of the world.
>
> And yet you are absolutely confident that the flagellum evolved? How
> so? You don't even really know how it works and yet you are sure it
> evolved simply because you find a few homologous structures in other
> systems of function? I find this sort of blind faith simply amazing!

Sean, this is science we're talking about, nothing is certain.
Certainty is for (some) religions. What I can say is that
investigation indicates that the flagellum is evolvable by natural
processes, just like wings or eyes or feathers or the 2,4 DNT
degradation pathway. There is no particular reason to attribute its
origin to intelligent design, because the ID claims (the function is
all-or-nothing, subsystems would be nonfunctional, etc.) do not hold
up.

I'm sure that further evidence will modify the model, modestly or
drastically, but this is how science proceeds in every case. What
more can we fairly ask for? You should look at how much the model for
ATP synthetases has changed over the years for example. I'd say that
the evolution of complex structures is better understood than, say,
the physicists' understanding of the phenomena that lead them to
postulate "dark matter" and "dark energy." They're ones who've really
got problems, yet it's evolution that gets all the flack from you
IDists. You'd think you be equal opportunity and be offering to help
the physicists out with your one-size-fits-all "IDdidit" explanation.
I'm sure they'd be delighted...

> Anyway, this is all I have time for right now. Your stuff about the
> evolution of cascades is not relevant to my position. I have covered
> my reasons for this extensively in other posts if you care to look
> them up. In short, cascades are not nearly as complex as systems
> where each of the parts is required to work together at the same time
> - as in the flagellar system of motility.

And as in 2,4-DNT degradation, PCP degradation, etc. You need all the
parts working at the same time, or the toxins (or toxic intermediates)
build up rapidly and the cell dies. Yet these IC systems have evolved
in recent history anyway, and achieved gene duplications and complex
regulational shifts as well. Many of these things are things that you
think evolution cannot do.

Anyhoo, sorry if I was a bit blunt in spots, it's very late. Let me
know when you come up with a model for flagellar origins superior to
mine...

nic

[howard hershey]

zosdad wrote:

> howard hershey <hers...@indiana.edu> wrote in message news:<bu9i72$rcj$1...@hood.uits.indiana.edu>...
>
>>zosdad wrote:
>>

[snip all of Sean]

>>
>>Oh, damn. I was just about to make essentially the same points and you
>>went ahead and beat me to it. [What am I complaining about? I get to
>>sit around and drink coffee and let you do the heavy lifting.] Of
>>course, I was going to use a certain article cited below by some guy
>>named Matzke as a major source in any case.
>
>
> Howdy there Howard Hershey. In case I didn't mention it before, many
> of your t.o. posts over the years were quite helpful in getting me
> started on the Big Flagellum Article.

Thanks for the compliment. But take all the credit yourself.

> BTW if Sean is feeling ambitious, after tackling the flagellum and
> immune system articles he should then rebut the literature on the
> evolutionary origins of:

I will be quite happy if he would stick to one thing rather than
changing goal post directions.

> Photosynthesis
> http://www.antievolution.org/cgi-bin/ikonboard/ikonboard.cgi?s=4009054319abffff;act=ST;f=9;t=5
>
> Blood-clotting
> http://www.antievolution.org/cgi-bin/ikonboard/ikonboard.cgi?s=4009054319abffff;act=ST;f=9;t=3
>
> Multiple-parts required toxin catabolism pathways
> http://www.antievolution.org/cgi-bin/ikonboard/ikonboard.cgi?s=4009054319abffff;act=ST;f=9;t=17
>
> And of course the many articles on the origin of the immune system(s)
> that have come out since Matt Inlay's article:
> http://www.antievolution.org/cgi-bin/ikonboard/ikonboard.cgi?s=4009054319abffff;act=ST;f=9;t=16
>
> Also acceptable would be if Sean gave us references where Behe or
> Dembski (1) acknowledged the existence of the literature cited on the
> above webpages, (2) critically reviewed it, and (3) successfully
> rebutted it. Heck, it would be a substantial improvement if they
> would just do (1). Instead, Dembski at least feels that massively
> deluded chest-thumping about the lack of literature is the way to go:
>
> =========
> Irreducible Complexity Revisited. [14Jan04] Update on where the debate
> over irreducible complexity is eight years after the publication of
> Michael Behe's book Darwin's Black Box.
> http://www.designinference.com/documents/2004.01.Irred_Compl_Revisited.pdf
> =========
>

Or Dembski's rather silly and sad review of your article (sorry, I don't
have the site right now, just a copy of the review), which is devoid of
substance.

He starts off, tellingly enough, with an ad hominem attack on your
credentials in biology (laughable coming from Dembski) and a complaint
that you cited too much literature (NWM redux--damn these scientists and
all their articles).

He then procedes to give away the ID store.

"Michael Behe, the best known proponent of intelligent design, holds to
universal common descent. He is as much an evolutionist as Matzke. Where
they differ is on how evolution brought about biological complexity. For
Matzke and the majority of biologists, the Darwinian mechanism is all
that's required. For Behe, some form of intelligent guidance is
additionally required. But clearly, if the bacterial flagellum is
evolving under intelligent guidance, then existing designed structures
are fair game for co-option into newly designed structures. Insofar as
intelligent design is a theory of evolution, it is a theory of
technological evolution, and technologies evolve by taking advantage of
existing technologies. Thus co-option will play as important a role in
any intelligent design models for the evolution of the bacterial
flagellum as in Matzke's Darwinian model."

IOW, what does ID look like? It looks just like evolution except that
Godidit. It looks like (don't peek now) that demon, satanistic ideology
called *theistic evolution* that biologists have no problem with.

His other argument is that, duh, we do not have the sequences of the
putative ancestral bacteria nor do we have that bacteria to work with.
This, of course, is just as much a problem for ID. But notice that he
did not actually say that any of Matske's suggested intermediates were
impossible or that the selective pressures and proposed mechanisms are
so far different from what bacteria do today and the wide range of
environmental constraints that today's bacteria face that using them as
a model of what is *feasible* makes no sense.

"What they don't have an empirical handle on is Matzke's "ancestral"
type III secretion system, the environment it inhabited, what sort of
environment lacking metazoans this structure might have had a selective
advantage in, the intermediates in Matzke's model leading from ancestral
type III secretion to the modern flagellum, and the environments in
which those intermediate systems (if they existed at all) might have
been functioning. Matzke's most sustained argument for selective
advantage relevant to his model concerns the minimum functional
requirements for a pilus/filament to serve as a motility structure
capable of overcoming Brownian motion (hence the title of his article
"Evolution in (Brownian) Space"). For a bacterium with an ancestral type
III secretion system that has sprouted a pilus, does motility for the
pilus (and thus for the bacterium) constitute a selective advantage that
overcomes the cost of evolving motility? Maybe. Maybe not. Who's to say?"

"Given our minimal knowledge of the ancestral environment leading up to
the bacterial flagellum (and we're talking an environment at least two
billion years old when it comes to the evolutionary origin of the
bacterial flagellum) and given only hypothesized rather than actual
evolutionary intermediates leading up to it, there is no way to decide
what does and does not constitute a selective advantage. Just about
anything could constitute a selective advantage or selective disadvantage."

"As I pointed out earlier, all we have in hand is the modern type III
secretion system, the modern bacterial flagellum, and various homologous
biochemical structures embedded in the flagellum present in extant
organisms. We don't have the intermediates that Matzke posits nor the
ancestral type III secretion system. We don't know what they look like."

"Are the transitions from one step to the next in Matzke's model
reasonably probable? Does each step in his model constitute only a
"slight modification" (sensu Darwin)? There's no way to tell because the
model is not sufficiently detailed.... We don't have the intermediates
that Matzke posits nor the ancestral type III secretion system."

Given our even more minimal knowledge about the motivations and
capacities and mechanisms of action or even the very *existence* of an
intelligent designer, of course, Dembski still thinks we are supposed to
assume that He did it *unless* we have every bit of knowledge of past
environments? Why does the absence of such direct evidence of the past
favor the HYPE explanation? No hint of an explanation is given. Nic
was not assuming selective pressures radically different from the kinds
that bacteria face today. It isn't like he was assuming that, if the
bacteria were temporarily transported to the moon, the absence of
atmosphere would favor the development of motility or a TTSS-like
ancestral transport machinery. The evidence Matzke has is the same
evidence that Dembski has to support his claim that the proposed
intermediate states couldn't exist, are infeasible, and a designer did it.

But all Dembski does is say that because we don't have all this
*detailed* information about intermediate states and selective pressures
(even reasonable ones), which would require a time machine, therefore we
cannot say *anything* at all about the *feasibility* of evolution of
this system by Matske's proposed mechanism (which he somehow avoids
descibing except in the most general terms -- as if he couldn't follow
the details). [Nic is not saying that his proposal *is* how the
flagella evolved in any case. He is saying that this is a proposed
mechanism that is both *reasonable* and *feasible*. Dembski really
needs to attack the *reasonableness* and *feasibility* to make the claim
that Matske has failed to make his case. He doesn't.] Therefore, in
the absence of all this detailed information, godidit. Ignorance means
goddidit. Sad, really, both for theological and logical reasons.
Ignorance = god to Dembski. Sean, of course, shows the same tendencies.

Then we get stuff like this toward the end, showing the "exquisite
detail and specificity" of Dembski's critique of the lack of same in
Nic's article (you really have to read both to get the point).

"But are those steps reasonably small so that they constitute what
Darwin called "successive, slight modifications." My sense is no --
getting from a type III secretion system to a bacterial flagellum in six
steps seems on its face to require at least one big leap somewhere."

Am I the only one that noticed that he didn't actually *ever* list the
six steps nor declare which one of these six steps was the "one big leap
somewhere"? Nor declare which step was *infeasible* given our knowledge
about what nature can do? Why is it only his "sense" (a highly
debatable feature of his intellect) that tells him that the answer is no?

Toss in a few "good Christian" ad hominem comparisons based on the ad
hominem attacks on 'lack of credentials. And then we have:

"Matzke's model, far from resolving the evolutionary origin of the
bacterial flagellum and despite his protestations to the contrary, is
yet another exercise in Darwinian storytelling."

As opposed to 'anti-Darwinian storytelling' which proposes a HYPE to
explain things based solely on ignorance (the more there is, the more
reasonable the HYPE becomes)? Darwinian story-telling can be judged on
the basis of reasonableness by how closely it matches the way things are
currently observed to work in nature and with certain predictions about
what should have happened if certain events happened in the past. It is
testable because particular stories have particular consequences,
especially, in this case, consequences of homology. Are the proposed
selective conditions reasonable *and* consistent with the evidence? Are
there functions that subsystems related to those in a more complex
system that *could* function as independent entities? Is there evidence
of such systems in modern organisms that have homology? That would
certainly make this a more, rather than less, feasible explanation. Are
there steps that could occur by known observed mechanisms for generating
currently observed genes (duplication and divergence, chimeric
duplication, co-option, co-evolution)? That would certainly make such
steps more rather than less *feasible*.

Remember that the entire basis of specifying ID as an alternative
explanation is by identifying those things that *cannot*, in principle,
have evolved; those things where evolutionary explantions are
*infeasible* or highly unlikely for at least *one* of the proposed
steps. In that context, Darwinian story-telling must be directly
confronted by specific evidence that specific proposed intermediate
steps are highly unlikely because they require events that are not seen
in nature today. That is, if it were impossible to generate a
two-protein system from appearing and affecting morphology or function
by generating a binding site, one could argue that such an event in the
past would be impossible. If you have *evidence* that generating a
particular system *requires* the equivalent of an n-body collision
rather than n 2-body collisions, that would make such an event less
likely. Otherwise, all you would have is unsupported assertion that
something is infeasible and Godidit means the same thing as personal
ignorance.

I haven't seen the type of argument from any creationist or IDeologist
yet that actually looks at feasibility.

[Greg G]

howard hershey <hers...@indiana.edu> wrote in message news:<bumsfm$a5h$1...@hood.uits.indiana.edu>...

It's at http://www.designinference.com/documents/2003.11.Matzke_Response.htm

> which is devoid of
> substance.
>

--
Greg

It is horrifying that we have to fight our own government to save the
environment. --Ansel Adams

[zosdad]

howard hershey <hers...@indiana.edu> wrote in message news:<bumsfm$a5h$1...@hood.uits.indiana.edu>...

> zosdad wrote:
>
> > howard hershey <hers...@indiana.edu> wrote in message news:<bu9i72$rcj$1...@hood.uits.indiana.edu>...
> >
> >>zosdad wrote:
> >>
>
> [snip all of Sean]
>
> >>
> >>Oh, damn. I was just about to make essentially the same points and you
> >>went ahead and beat me to it. [What am I complaining about? I get to
> >>sit around and drink coffee and let you do the heavy lifting.] Of
> >>course, I was going to use a certain article cited below by some guy
> >>named Matzke as a major source in any case.
> >
> >
> > Howdy there Howard Hershey. In case I didn't mention it before, many
> > of your t.o. posts over the years were quite helpful in getting me
> > started on the Big Flagellum Article.
>
> Thanks for the compliment. But take all the credit yourself.
>

I figured I should mention it somewhere...

Yeah, it was *great*:
http://www.designinference.com/documents/2003.11.Matzke_Response.htm

> He starts off, tellingly enough, with an ad hominem attack on your
> credentials in biology (laughable coming from Dembski) and a complaint
> that you cited too much literature (NWM redux--damn these scientists and
> all their articles).
>
> He then procedes to give away the ID store.
>
> "Michael Behe, the best known proponent of intelligent design, holds to
> universal common descent. He is as much an evolutionist as Matzke. Where
> they differ is on how evolution brought about biological complexity. For
> Matzke and the majority of biologists, the Darwinian mechanism is all
> that's required. For Behe, some form of intelligent guidance is
> additionally required. But clearly, if the bacterial flagellum is
> evolving under intelligent guidance, then existing designed structures
> are fair game for co-option into newly designed structures. Insofar as
> intelligent design is a theory of evolution, it is a theory of
> technological evolution, and technologies evolve by taking advantage of
> existing technologies. Thus co-option will play as important a role in
> any intelligent design models for the evolution of the bacterial
> flagellum as in Matzke's Darwinian model."
>
> IOW, what does ID look like? It looks just like evolution except that
> Godidit. It looks like (don't peek now) that demon, satanistic ideology
> called *theistic evolution* that biologists have no problem with.

Dembski has been heading that way for awhile, it's his ultimate "even
if I'm completely wrong, I'm still right in an unprovable way" backup
position.

Yeah, at various points Dembski sort-of admits that I discussed
selective pressures continually, documented analogs, etc., but he
never explains why any of them are unreasonable.

Yes, Dembski couldn't even identify the "leap", but he was *sure*
there was one there.

> Toss in a few "good Christian" ad hominem comparisons based on the ad
> hominem attacks on 'lack of credentials. And then we have:
>
> "Matzke's model, far from resolving the evolutionary origin of the
> bacterial flagellum and despite his protestations to the contrary, is
> yet another exercise in Darwinian storytelling."
>
> As opposed to 'anti-Darwinian storytelling' which proposes a HYPE to
> explain things based solely on ignorance (the more there is, the more
> reasonable the HYPE becomes)? Darwinian story-telling can be judged on
> the basis of reasonableness by how closely it matches the way things are
> currently observed to work in nature and with certain predictions about
> what should have happened if certain events happened in the past. It is
> testable because particular stories have particular consequences,
> especially, in this case, consequences of homology.

I also liked the bit where Dembski said that the evolutionary model
requires homologs, therefore they aren't predictions of the
evolutionary model.

Oh, and ID predicts homologies too, if they are found.

What gets me about the "just-so story" charge, in addition to the
usual hypocrisy and arguments from ignorance, is the assumption that
constructing models for an evolutionary process is radically different
than constructing models in any other part of science. Any in-depth
investigation of things like the flagellum, or ATP synthetase (or just
about anything incompletely understood -- that is, active science in
general) shows that models are continually proposed, critiqued, and
improved. I have a textbook only a decade old where the b subunits of
F1F0 ATP synthetase form the middle stalk of the enzyme, now those
subunits are on the outside. The delta subunit has moved from the
membrane-F1 interface to the far proximal end of the F1 subunit. The
number of c-subunits is still undetermined, ranging from 10-14. In
flagella, the action of the motor is only understood in the most
general way, and yet Dembski thinks that ID should be the conclusion
unless evolution can provide every point substitution that led to the
flagellum! Science is all about improving approximations, but for
Dembski approximations are not allowed...

Except of course for ID, which makes vagueness an art and a central
principle of the "theory".

[Gregwrld]

howard hershey <hers...@indiana.edu> wrote in message news:<bumsfm$a5h$1...@hood.uits.indiana.edu>...

> zosdad wrote:
>
> > howard hershey <hers...@indiana.edu> wrote in message news:<bu9i72$rcj$1...@hood.uits.indiana.edu>...
> >
> >>zosdad wrote:
> >>
>
> [snip all of Sean]
> >

<snip>

Funny thing about all this "Darwinian" storytelling: It doesn't
require a designer who, for some reason, can't be located or described
and whose methods remain a complete mystery even Pitman (or Dembski)
can't unravel...

Greg Czebatol
Gregwrld

I want to die peacefully, like my grandmother.
Not like the screaming passengers in the car with her...
(-corrected version) -Norm Davis

[sweetnes...@yahoo.com]

Sorry for the delay. With all due respect, posting here isn't exactly
the highest thing on my list of priorities...(a list that has been
quite long lately).

Sean: The first part of the response is just a deflection of your
latest attempt at obfuscation. The second part is the interesting one
(as in, I am actually interested to see what you will say in response
to it). So please, don't just answer to first third, and say "this is
how much I have time for today". If you don't have time today, answer
it over several days - I won't have time to write back for two weeks
at least.

First part:

>You just fail to understand why I don't think they
>support what you say they support - such as the
>notion that evolution beyond the lowest levels of
>complexity has ever occurred.

Of course I fail to do so. You read the articles in much the same way
you read my responses: not to learn something you didn't know
previously, but from a position that can be summed up as: "I already
know that this article is wrong, so let's find things I can twist
and/or misinterpret in order to 'show' it wrong to others". I am still
unsure whether this is conscious or not.

>Sure, large proteins evolve, but the examples that you've
>listed are certainly not very specified at all.

Sigh. I gave you examples of structural evolution for a reason. There
are only a few thousand structural families of proteins. Not millions,
not billions. There could be many, many more - we are now slowly
reaching the stage where we can predict structure, so people are
starting to produce folds never before seen in nature. Nature, on the
other hand, has a certain number of folds that work sufficiently well,
and is making do with them - precisely because evolution works from
what is already there, a point whose relevance seems to escape you
completely.

>Such protein systems/function require relatively little
>genetic real estate to code for them.

Nonsense. In the examples I gave the number of separate genes involved
sometimes reaches several dozen (eucariotic RNA polymerase complex,
for instance).

>In fact, it seems to me that these steppingstone systems
>are too few and are very far apart in sequence space.

Indeed. That seems to you. It also seems to you that Grand Canyon
geology is flat, and it seems to you that it is impossible to get a
specific immune response to a heme with only 3-4 amino acids attached
to it.

Let us take a look at that last problem a bit more in-depth. If those
fragments found in dinousaur bone actually had ten, fifteen or twenty
amino acids, the scientists would have said so. After all, far older
fragments consisting of much longer polypeptides have been discovered
before (and since). Preservation of proteins over spans of even
millions of years is not something that is a problem for evolution.
The implied theory is that the team of scientists that examined the
bones lied to "protect evolution" (or was "brainwashed" by evolution,
whichever) - but there was no reason whatsoever to do so. Even a bad
scientist, who is willing to tweak the data to make it fit with the
preconcieved theories had no reason to do so.

But you don't understand that. "It seems to you" that having ten or
fifteen residue polypeptide survive in an old bone would disprove age
of life on Earth. So, therefore, "it seems to you" that you know more
about the immunoassays then the people who performed them.

>Now obviously you disagree, but you seem unwilling
>to detail to me, specifically, how such a specified,
>multiprotein system as a flagellar motility system
>involving several thousand amino acids working
>together at minimum, could evolve in some sort of
>stepwise fashion.

This is a good theory:

http://www.health.adelaide.edu.au/Pharm/Musgrave/essays/flagella.htm

Which, of course, you will dismiss as a "just-so story". It is a
theory based on evolution of many other systems, for which we *have*
data (immune system, pterine biosynthesis, etc). There is also the
fact that every paper that is published on the flagellum (I refer you
to Blair, again) confirms the predictions based on the evolutionary
theory, and further weakens the case for ID.

>How are the intervening neutral gaps reduced to a
>manageable size so that they can be crossed this
>side of a zillion years?

The fact that it "seems to you" that the neutral gaps are enourmous
does not mean that it is actually so in reality.

>Come on now Sweetness. I'm not the only one who repeats
>himself over and over again.

Really? How about a statistical analysis of your posts vs. mine? We
can compare how many different examples I gave, versus how many times
you repeated nonsensical concepts like "thousands of amino acids
working together", or "specified complexity".

>You did mention other smaller molecules that could
>*potentially* perform the electron transport function,
>but none of these hypothetical molecules has ever been
>shown to work in a living creature in such a capacity.

Indeed. So, you did read the references I gave you, right? In that
case, your comprehension level is truly astonishing. Because, you see:

>Something like this might seem very good on paper, but it
>fails miserably when tried in real life.

…I have given you a list of proteins other then cytochrome-c that
perform the electron transport function. Not *potentially* perform.
Definitely perform. As can be seen by the continued survival of the
organisms they have been extracted from.

None of them is on paper. All of them come from real life. All of them
work. And your attempt to decieve people into believing that you have
a clue on the subject you are discussing is more obvious here then
ever before.

>Yes, you have done this. You have shown that evolution
>can build up very large *non-specified* structures. In
>fact, you yourself note that the large a protein gets the
>less specified it becomes on average. Isn't that so?

A statement which you don't seem to understand. The larger the protein
gets, more it can change without losing its function, yes. There are
no super-hyper large proteins in which every single base is *crucial*
for the function of the protein.

In other words, contrary to your nonsense, it isn't that much more
difficult for mutations to produce a large protein then a small one.
There are more structural elements, but they can change more without
affecting the function.

>That is my challenge. What ya got? Where is your reference
>for this one? Just one reference is all you need to shut me up.
>Go ahead then. I *repeat myself* - List your best reference that
>demonstrates the evolution of a system that requires, at minimum,
>at least a few thousand fairly specified amino acids working
>together at the same time.

Your "challenge" is truly Hovindesque. Let me give you what I believe
to be an honest analysis of your challenge: you will completely ignore
the dozens of systems that are every bit as complex as flagellum, but
have been well-researched (since their importance for medicine is far
greater), and whose evolutionary pathways have been well explained.
Then, you pick one of the poorly-researched ones. Since there is
little data on it, and since it is still poorly understood, its
evolutionary pathway is still highly hypothetical. When I give you
references to that, you point and note that it is all still highly
hypothetical. Therefore, you will conclude, you are justified in
deciding that an incredibly improbable *hypothetical* being is behind
it all. And you will not see any problem with that.

Never mind. Above, I gave you a link to a proposed pathway for the
development of flagellum. It has several references at the end, and
they are decent references. I also wrote a hypothetical evolutionary
pathway for the flagellum in one of my previous responses to you. Pick
one. Then tell me which step in the pathway is impossible.

The burden of proof is on you. Let me elucidate why:

I have the data on evolution of many complex systems (from my own
work, to the well-described systems I have mentioned above).
Therefore, I know that complex systems can and do evolve. I have
little data on the flagellum. Therefore, I cannot prove that this
particular system evolved in a particular way. However, I am sure it
did evolve, and I can propose a pathway. Until more research is done,
that pathway is hypothetical.

You, on the other hand, claim that this system could not have evolved.
Instead, you say, it is reasonable to hypothesize that there exists an
intelligent being which created it de novo. The only proof you have of
this being is this claim that the complex systems "could not evolve".

I am proposing that something that has been observed before, in other
systems, is happening in this one too. I have evidence that the
process of evolution exists. I lack the specific evidence on how that
process applied to the system in question. My hypothesis is valid,
unless it is proven that the mechanism I propose cannot produce this
system.

You propose something that has never been observed: an act of creation
de novo. You have no evidence for it, other then claiming that the
process I propose cannot produce the system. In order to prove this,
and to disprove my hypothesis, you need to prove that steps of the
process I proposed are impossible, or contrary to observable evidence.

I am waiting to see you do so.

>The reference you listed discussed large proteins with
>low specificity or small protein systems with high specificity.

Actually, it discussed some quite large proteins, with very, very high
specificity. Enzymes with size and specificities comparable to lactase
(often larger) are routinely produced by directed evolution. Again,
see the references.

>As far as I was able to ascertain, none of the listed references
>discussed the evolution of a system that was both large as well
>as fairly specified in its amino acid order.

That is because you didn't read them at all. Your ascertainment
capabilities can be seen by the fact that you thought some common
electron-transfer proteins are theoretical proposals that work on
paper, and aren't found in living systems.

>The evolution of large proteins with low specificity is not
>"improbable" at all. Why would I even try to challenge this
>notion?

Obfuscation. The large proteins evolved in the examples I gave you are
as "specific" by any measure as any other large protein. Which you
would know if you actually read the papers – or maybe not, you need to
know how *difficult* it is to get some of those activities, and how
*specific* the sequence has to be in order for those activities to be
achieved; a level of understanding of biochemistry that is far above
one you have demonstrated here.

>You've got nothing to back yourself up in your challenge
>of my actual position.

As I said, let the readers make their own minds.

Part two:

This is a continuation of the questions I asked you, and this is the
part I am really interested in (I already know what you'll answer to
the first part). If you can answer to only a part of this response,
please answer to this part, especially the highlighted questions.

>However, the evidence as I see it seems to indicate that life
>on earth and much of the geologic column is very young - less
>than 10,000 years old in fact.

Question: do you realize you will have to overturn not only evolution,
but most of biochemistry, biology, geology, geophysics, analytical
chemistry, atmospheric chemistry, etc. in order to "prove" this
hypothesis?

>I think that horses and donkeys did share a common ancestor.

I could ask many, many questions here, but I will limit myself to one:

Question: the fossil record suggests that donkeys and horses diverged,
I believe, from Pliohippus, some ten million years ago. Why is the
fossil record wrong?

Question: There is a significant number of genetic differences between
horses and donkeys. Since they had a common ancestor, we can calculate
the speed of mutations that caused them to diverge into two species.
Now, if they diverged ten million years ago, that is fine. If they
diverged ten thousand years ago, they had to both mutate very, very
quickly in order to become so different in so short a time. Why don't
we observe such fast mutation and divergence rates today?

>Although the bias is very heavy against anything that challenges
>the fundamental "truth" of evolution is popular science journals,
>I think that there might still be a chance a being accepted in such
>a journal.

Question: Scientific journals generally do not accept papers that
present no evidence for the theory they propose. What evidence for
design (other then "evolution is impossible, therefore goddidit") will
you present in your paper?

>In short, system specificity is the range to which change within a
>system can occur without a complete loss of the original type of
>beneficial function.

Really. In that case:

Question: I have provided you with references to several large
proteins evolved by random chance in laboratories. Some of them (see
the dissacharide-specific invertases, for example) exhibit very large
sequence specificities – they are at least as specific as lactase, if
not more so. Why do you discount them as proofs that large, specific
proteins can evolve, while extolling lactase as an example?

>The specificity of such systems as the flagellar motility system
>is fairly well known and established by more independent means
>(such as the variability found between the same protein doing the
>same job in the same system of function, in different creatures).

There has been far less work done on flagellar motility systems then
on other systems of equal (or greater) complexity that are more
important for the medical practice.

Question: how can complex systems (such as eucariotic RNA polymerase
complex) evolve, if evolution is impossible? What are the impossible
steps in the above-mentioned (and well documented) case?

>How would you estimate protein specificity?

Exactly in the manner you propose here. Some of your previous
statements seem to contradict this. Therefore:

Question: why do you claim that lactase is an incredibly complex
system, when the method for ascertaining complexity that you yourself
propose shows relatively low levels of sequence identity between
different lactases?

Here we run into an interesting statement:

>>Take, for example glycogen synthase (a quite distant molecule,
>>mind you), and give me a model how many mutations would be
>>required to move from it to a lactase.
>

>Since these are all relatively simple functions that require
relatively
>short proteins with fairly low specificity, why should I knock myself
>out showing how they could evolve from one place to the other? I
>don't think such evolution would be that much of a problem.

Question: are you admitting here that lactase could evolve from a
distant, unrelated enzyme, without much of a problem? How does that
reflect on all of your messages from the past several months? Why did
you, for months, claim that lactase is so impossibly complicated,
evolution could never account for it, if it is actually a "relatively
short protein with fairly low specificity"?

------------
This belongs in the first part:

>However, for systems such as the flagellar motility system,

So you are grasping for one of the few last examples that are still
sufficiently unstudied that I can't just hit you over the head with a
pile of data? No, never mind, you don't have to answer that ;) – I
just find it interesting that you switch immediately to flagellum as
soon as you run into something you can't answer directly.

Your analysis is also deeply flawed:

>The only problem with this is that the minimum TTSS system
>requirement is around 6 different types of proteins while the
>minimum flagellar system requires over 20 different types of
>fairly specified proteins. This leaves a gap of several
>thousand fairly specified amino acids, in the form of 12 to 14
>proteins in-between the TTSS function and the motility function
>of the flagellum. How is such a gap crossed? Even for a huge
>population of bacteria the random walk involved here would
>required trillions upon trillions of years of average time.

No, actually, it doesn't. Sigh. Let me give you yet another possible
line of evolution.

You start with a single pH-control pore-forming protein (a large
beta-barrel will do). Over time, add a few proteins that regulate it.
Add another protein under it, to actively pump H+ ions out. Duplicate
it, turn it into a two-step pump (more easily regulated). You have a
secretory system for pH control.

You develop passive cilia. You need a membrane-anchored protein (any
membrane-anchored protein). You need a small protein that is secreted,
which attaches to the anchored protein without significantly altering
its function (an easily achievable mutation in any small secreted
protein), which changes conformation depending on the pH. Another
mutation makes the small protein polymerize into a longish chain.

As the proton pump pumps creates a gradient, the cilia twitches; you
get uncontrolled motion. From there, it is all downhill: addition of
new segments that regulate movement and make it more efficient.

There you go. It is a hypothesis. I cannot prove that it happened this
way. It probably *didn't* happen this way, but it did happen in some
similar manner. The hypothesis uses only the systems we have already
observed in action. Each step has an intermediate function that is
important for the cell.

Now, show me which of these steps is impossible. Yes, this is a
just-so story. There are many of them, and one of them is right. But
that is not my problem. Your entire argument relies on the fact that
it is absolutely impossible for this to happen. So prove it.

>but absolutely none of these proposed steps has been demonstrated
>to actually evolve in a living thing - not one step, much less the
crossing
>of the entire gap from TTSS to a flagellum or any other equivalent
level
>of specified complexity.

Really. In the theories I saw, each of the intermediates still exists.
The original secretory system has developed to become TTSS. There are
still inactive fimbrils that are not involved in motion in many
organisms. All the components are there.

>Certainly it should be easy enough to set up an experiment
>to demonstrate the crossing of one "short" step from one
>proposed steppingstone to another? - right?

Wrong. Where do you propose finding a bacterium that does not already
have all the systems in place, and is integrated around those systems
so tightly, you would have to redesign an entire organism to get to
the point where you could perform this experiment.

>Why hasn't this been done if the steppingstones are really
>this close together?

If you had more then the very basic knowledge of biochemistry in
actual living systems, you would know. We don't have the original
organism that did it.

One would have to design a bacterium without TTSS or any other system
to control internal pH, other then a few porins. It would have to be a
very simple organism, without many of the complex functionalities that
we see today, because many of those systems have evolved to rely on
the precise pH control.

If you had such an organism, you could measure how long it took him to
make an active proton pump out of that starting porin. You could
attach some cilia to it (we already know that cilia and similar
systems readily arise through simple mutations), and see how long it
took for uncontrolled motion to arise.

And finally, the stepping stones are close together: only a few
million years apart, in populations of various bacteria evolving in
different environments across the entire planet. In lab, observing
relatively small number of colonies living in a few small, simple
environments, it could take tens of millions of years.

Which is not too slow for evolution, but who is going to sit next to
it and observe it for so long? The best we can do is observe how fast
systems arise and modify in modern systems, and how difficult it is
for a protein to achieve regulation, or for two or more proteins to
join into a system. All of those have been observed, and there is no
reason to consider any of the steps in generation of the flagellum
particularly improbable, not to mention impossible.

>bacterial flagellum, please detail your favorite one here and
>then I will respond to it. I have already done this sort of thing
>many times with others, such as Ian Musgrave and his ideas
>and several of the references that he referred to dealing with
>the supposed pathway of flagellar evolution. So, I'd be willing
>to do the same thing again for you with an equivalent system
>of your own choosing.

You can take on the flagellar evolution proposal above. It is
semi-serious, but I would like to see you prove it is impossible. Or,
you can take on the recent discoveries on the evolution of the immune
system.

-------
Now, back to the second part. These are the questions that I really
want to see the answer to:

>We also know, without a doubt, that mindful processes exist.

Ah. And there is the cincher.

We know that there is a sequence of reactions in our brains. Those
reactions happen according to the *mindless* laws of physics and
chemistry. Those processes cascade, always mindlessly, in mindless
response to impulses produced by other mindless forces around us. The
product is what we call "intelligence". However, that intelligence is
just a bunch of mindless chemical reactions happening in our brains.

Therefore, there is nothing in the world that is not a mindless
process. Our minds are just complex networks of mindless processes. A
"mindful" process that was not just a network of mindless processes
has never been observed.

>Involved does not mean entirely responsible. There is nothing
>that we see mindless processes doing now that can even remotely
>begin to explain the mindless origin of life.

Well, we are observing those processes doing exactly that. But never
mind; your point, therefore, boils down to the claim that mindless
processes cannot do it, while those same mindless processes, organized
in a network such as the one in our brains, can do it?

>It simply does not follow that since mindless processes exist
>that they must explain everything or that they are the most logical
>answer given the evidence available.

Question: What is the evidence for anything besides the mindless
processes? Is there a process in the universe that can be observed,
and which is not mindless? Name one.

>For example, lets say that we create computer robots
>to the point where they become self-aware.

That will not change the fact that that self-awareness will still be
only a function of mindless processes that push electrons through the
circuits in their "brains".

>Would it be automatically logical for them to conclude that
>their origins were the result of the mindless interactions of
>the sands, waters, electricity, etc., of the Earth? Not at all
>since no known non-directed processes ever produce anything
>beyond the lowest levels of functional/informational complexity.

Not at all, indeed, but for a slightly different reason. The nature
has never been seen to produce a microchip. There are no ultrasimple
microchips that are multiplying in the seafoam, mutating, and
developing. Indeed, even the starting materials for robots are
something that nature has never produced.

On the other hand, nature readily produces starting materials for
life. It is possible to imagine a primitive life form, and it is not
exceedingly unlikely that such a life form could further develop and
evolve. Scientists are rational people. They, in general,do not
entertain belief in highly improbable things (not for long, at least).

>This is the entire debate now isn't it? How blind
>can you get? The evidence of design is found in the
>complete inability of any known mindless process
>to even get close!

Let us, for the moment, ignore your selective blindness to the fact
that mindless processes have and do produce things that are not close,
but right there.

There is a crucial problem with your proof, and with something you
said when we initially began this discussion. You said that we cannot
infer anything about the Designer. You are wrong.

We can, with absolute certainty, from the fact that we *do* exist,
infer several things if we accept the existence of Designer. We can
know that the Designer a) had a mind complex enough to imagine and
design the universe (or at least, life), b) had the will to do it, and
c) had the means to do it (means to manipulate nonliving matter and
create life from it).

Therefore, we *know* that, if there is a Designer, he (she, it) is an
enormously complex being, with not only fantastic complexity of the
mind, but also of the enormous physical abilities. The only minds we
know of are complex networks of completely mindless processes, which
produce an overall, complex process, which we call "intelligent" (they
are not necessarily determined by this; random chance, which is also a
mindless process, could play a factor, giving us what we call "free
will").

Therefore, we are comparing two chances.

One is the chance that the mindless processes formed up the first
living organism. We are still working to see how probable is that –
but let's say that it is highly improbable.

The other is the chance that the mindless processes spontaneously
formed to immediately, in one step, produce an enormously complex
being, together with an extraordinary set of tools which is beyond
anything we have today ourselves, and that this being then proceeded
to create us.

The chances for the second option are so far less then the chances for
the first one, it is meaningless to even compare them. The chance of
things randomly happening does not, actually, support ID. It works
against it.

Oh, yes, there is the third option. That the Designer is a product of
processes we know nothing about, and that its mind operates by means
different then our minds do (i.e. that it is not just a network of
mindless processes). This is sheer imagination. There is no evidence
for either such a mind being possible, or of such processes.

If you think that imagining an unknown process to explain something is
sufficient proof for a theory, two can play that game. You propose
that there is a never-before-seen, unknowable Designer, which, by
means that are completely beyond anything that we have seen before,
thinking with a mind that is unlike any mind we know about, using
tools that are completely beyond us, created the life on the world. He
does not do it anymore for reasons we do not understand.

I have a simpler solution. There is an unknown, undetectable, mindless
quantum force which, by means that are completely beyond anything that
we have seen before, using mechanisms we know nothing about, created
life on earth. It does not do so anymore for reasons we do not
understand.

There. It is equivalent to your proposal. Chances of them being true
are the same (infinitely small, as they represent two choices out of
an infinite number of imaginary solutions we can think of).

>A priori assumptions are NOT part of the scientific method and are
>in fact contrary to it. They are nothing more than philosophical
blinders,
>not science.

Indeed. The point of discussion, however, is precisely that claim. I
look at evolution, and I see it operating. I use the predictions made
on the basis of evolutionary theory in my work, and I get results.
When I see something that is a part of the same system, I think "well,
I have seen thousands of such systems that evolved and are evolving,
so I will assume this one obeys the same laws". Much in the same way
that an astronomer assumes that the stars he is observing obey the law
of gravity; if he first attempts to verify that gravity indeed exists
and operates the same way around each star, he will never get
anywhere.

You look at it, and you don't see it operating. You think that all of
us who see it are actually blinded with our a priori assumptions (and
you wonder why we think you are insulting us all). And, at the same
time, you think that it is perfectly fine, even if you were right and
evolution does not work, to simply imagine a solution and decide that
this solution is the one that answers all the problem – without any
evidence except incredulity.

>Where do you think "modern science" started? With
>those who were fervent believers that God did a whole
>lot of things, just not everything.

It started with those who were fervent believers in God putting God
aside, and saying "well, let us examine this without getting God
involved". And it worked. Which is the reason that, as the science
progresses, God turns out to be involved in fewer and fewer things.
Quite possibly nothing.

>The best position, in my opinion, is to keep ones self open
>to both options. One should think that God may have done
>it or a mindless process may have done it - and then go from
>there to see where the evidence leads.

And how about pink unicorns did it? Or trolls? Or maybe very small,
very fast invisible people – should we devise experiments to check if
mutations are caused by invisible people who are running around
changing the genetic code? How many options should we explore? We can
imagine an infinite number of them.

No. The best position is to, FIRST, observe the evidence. See what is
there: there is this force, there is that force, there is this object,
there is that object… Then, from there, you start your theory. You do
not involve God, nor do you involve mindless processes, unless you run
into evidence for either of them.

>It is possible to support either position given the evidence.

There is evidence for existence of God? Please, share it with me.

>However, if you exclude one of these possibilities before you
>even consider the evidence, you will not be able to find the truth
>if the truth happens to be behind your pre-assumed blinders.

One has to exclude an infinite number of different possibilities, or
otherwise one cannot ever start examining the evidence. You may think
that your decision to include God is somehow more rational then
someone else's decision to include small, fast invisible men would
be…but that is just your personal perspective. Objectively, there is
no difference. You are postulating entities without having a reason to
do so.

>Mindless processes do not give rise to all that you see. Much
>of what you see can and is ONLY done by deliberate intelligent
>design. For example, look at your house or the cars on the street.
>Were they formed via mindless processes? Or, are they evidence
>of deliberate design?

They were formed by networks of mindless processes in human brains.
You call it deliberate design, which is fine with me, as long as you
understand that this deliberate design is a product of the same
mindless forces.

>When was the last time you saw a mindless process create a
>flagellum or anything even close?

I never saw a process that takes hundreds of millions of years,
because I never had hundreds of millions of years that I could spend
sitting and observing anything.

>how is it unreasonable to think that a mindful cause, at our
>own level of creativity or beyond, might actually have been
>behind such marvelous works?

Unreasonable it is, and even the odds are against it. And, crucially,
it isn't scientific, as there is zero evidence for it. Your desire to
fantasize that "intelligent" and "deliberate" processes are somehow a
different class from mindless processes that form them is just wishful
thinking (at least, according to everything we scientifically know –
what people see in visions does not really count here).

It is, actually, akin to the usual creationist belief that evolution
violates the second law of thermodynamics, because it creates more
complex entities from the less complex ones. The creationist does not
see a conflict between that statement and the fact that he himself
grew from a single cell in his mother's womb. He does not understand
that life, and intelligence, and everything else, is constrained by
the same laws, and that he is obeying them even as he is pronouncing
his ill-informed opinion.

As am I, as are you Sean. Unless, as I said, you can point to me an
example of an "intelligent" process that isn't just a complex function
of many mindless ones (which means it is a mindless process itself in
the last iteration)?

M.

[howard hershey]

sweetnes...@yahoo.com wrote:

[snip]

>
>
>>I think that horses and donkeys did share a common ancestor.
>
>
> I could ask many, many questions here, but I will limit myself to one:
>
> Question: the fossil record suggests that donkeys and horses diverged,
> I believe, from Pliohippus, some ten million years ago. Why is the
> fossil record wrong?
>
> Question: There is a significant number of genetic differences between
> horses and donkeys. Since they had a common ancestor, we can calculate
> the speed of mutations that caused them to diverge into two species.
> Now, if they diverged ten million years ago, that is fine. If they
> diverged ten thousand years ago, they had to both mutate very, very
> quickly in order to become so different in so short a time. Why don't
> we observe such fast mutation and divergence rates today?
>

[snip]

Perhaps I miscounted, but isn't that two questions? ;-)

To which I would add:

Question: If horses and asses shared a common ancestor, why are there
still horse's asses claiming they didn't?

[howard hershey]

sweetnes...@yahoo.com wrote in message news:<4d71d185.04021...@posting.google.com>...

[snip]

>
> >I think that horses and donkeys did share a common ancestor.
>
> I could ask many, many questions here, but I will limit myself to one:
>
> Question: the fossil record suggests that donkeys and horses diverged,
> I believe, from Pliohippus, some ten million years ago. Why is the
> fossil record wrong?
>
> Question: There is a significant number of genetic differences between
> horses and donkeys. Since they had a common ancestor, we can calculate
> the speed of mutations that caused them to diverge into two species.
> Now, if they diverged ten million years ago, that is fine. If they
> diverged ten thousand years ago, they had to both mutate very, very
> quickly in order to become so different in so short a time. Why don't
> we observe such fast mutation and divergence rates today?
>

[snip]

It seems that I am having problems getting posted today, so I will
repost this through Google.

Unless I miscounted, I do believe that you did not, in fact, limit
yourself to one question. ;-)

But that's O.K. I will add another.

Question: If horses and asses shared a common ancestor, why are there

still horses's asses that don't accept that fact?

[sweetnes...@yahoo.com]

> Perhaps I miscounted, but isn't that two questions? ;-)

Depends. If you use standard, godless mathematics, it is two
questions.

If you use the same ID-Mathematics that our good friend Sean uses to
calculate probabilities of evolution, it can be one. Or three.
Anything but two. ;>

Which could mean that he is actually getting to me. Damn. ;)

M.