Protein Structure Distribution in Sequence Space
Seanpit
> > On Jan 27, 12:38 pm, "Seanpit" <seanpitnos...@naturalselection.
> > Once the edge of this "magnetic field" is
> > reached, the minimum threshold is reached and further refinement
> > toward the maximum potential is no problem. It is the reaching of
> > this minimum threshold that is the problem. The island cluster of
> > potentially beneficial sequences represents the entire reach of the
> > potentially attractive "magnetic field". This entire reach is
> > completely isolated from all other island clusters and magnetic fields
> > at higher levels of functional complexity. The biosystem families you
> > speak of just are remotely clustered in one corner of sequence space
> > like you imagine. This notion of yours is way way off base.
>
> Evidence? I keep having to point out the obvious to you. The
> EVIDENCE (see the PNAS article and others that I have presented you)
> actually shows that the biologically useful proteins that *are*
> present in living organisms (and it doesn't matter whether they
> evolved or were created by the creation fairy) are NOT scattered
> throughout total structure space.
Even with regard to single protein systems, they are NOT clustered in
one tiny corner of sequence space. Look up even single protein maps
and you will see that they cover a wide range of sequence space. They
are not all clustered together like you imagine.
"Roughly speaking, however, distances are randomly distributed. This
means that, although only a small fraction of sequence space yields
uniquely folding sequences, sequence space is occupied nearly
uniformly. No "higher order" clustering (i.e., except the trivial case
of the homologous sequences) is visible.
Erich Bornberg-Bauer, "How Are Model Protein Structures Distributed in
Sequence Space?" Biophysical Journal, Volume 73, November 1997,
2393-2403
http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=1181141&blobtype=pdf
So, you see, even at very low structural threshold levels, the
distribution of potentially beneficial structures is widespread and
fairly uniform. The clustering effects that are present at lower
levels rapidly decline at higher and higher threshold levels until the
high-level islands are very remotely separated in sequence space.
Sean Pitman
www.DetectingDesign.com
Perplexed in Peoria
"Seanpit" <seanpi...@naturalselection.0catch.com> wrote in message news:1170613322.7...@v45g2000cwv.googlegroups.com...
Thanks for the link, Sean. I haven't studied it carefully yet, but my
first impression is that they are not looking at (a sample of) real
biological proteins. Instead, they are looking at computer-generated
sets of hypothetical proteins which (based on their structural model)
ought to fold stably and reliably. But I need to read more carefully.
However, I was interested in your parenthetical comment in this:
No "higher order" clustering (i.e., except the trivial case
of the homologous sequences) is visible.
I have mostly responded to your postings as if they were merely a
critique of the evolutionary model rather than positive arguments for
some other model. But I am led to wonder how your favorite model
accounts for all that 'trivial' homology. Take, for example, the
20 or so aminoacyl-tRNA-synthases. Each shows considerable homology
across the entire tree of life (with the trees constructed from any
one of them being consistent with the others, except near the roots).
Furthermore, it seems that they are all homologous with each other -
suggesting that they arose from a common protein ancestor before the
organism-level LUCA. Why would that pattern arise in some other
model besides the evolutionary one?
Furthermore, there are partial homologies between these proteins and
many, many others - especially in the 'nucleotide fold' domain. That
certainly strikes me as a kind of 'clustering'. Do you hypothesize
that this 'nucleotide fold' motif is the only good way to solve the
problem of binding ATP? Isn't it more likely that it is the first
way of binding ATP to be discovered and that it was then copied into
other proteins that needed this fragment of function?
Ron O
>
> So, you see, even at very low structural threshold levels, the
> distribution of potentially beneficial structures is widespread and
> fairly uniform. The clustering effects that are present at lower
> levels rapidly decline at higher and higher threshold levels until the
> high-level islands are very remotely separated in sequence space.
>
>
> - Show quoted text -
My first reading of this paper doesn't seem to indicate any support
for your neutral gaps model being a problem. They are talking about
small peptides that can retain a basic structure with a large number
of "neutral substitutions, but they make no conclusions about the
activity of such structures. It is known from gene families that
basic structure can be retained while function is modified.
So how would studies like this do anything to change the fact that you
don't have a model that you can support in any fashion worth talking
about. What is your alternative explanation for the evolution of
proteins and what is your evidence for it. Recent genome studies
indicate that gene duplication is a lot more frequent than we had been
able to detect before. Gene duplication seems to be a mechanism that
would explain many protein families. Do you have a better
explanation? What is your evidence for your alternative? Claiming
that there is a problem when you don't even have a clue about a
possible viable alternative is pretty bogus don't you think?
Are you ever going to get around to presenting your alternative to
common descent and your evidence for your alternative that you claimed
to have that was better than the current scientific model? You
claimed to have such a model and evidence, so why putz around with
bogus neutral gaps if you have some killer evidence for your beliefs?
Let's see it.
Ron Okimoto
_Arthur
"Empirical fitness landscapes reveal accessible evolutionary paths"
by Frank J. Poelwijk, Daniel J. Kiviet, Daniel M. Weinreich and Sander
J. Tans ?
The paper seems to address your existential fears about hypothetical
barriers.
They studied a particular enzyme which is commonplace in bacteria, and
a "mutant" version, which is 100,000 times more efficient against
antibiotics like penicillin.
The new version of the protein/enzyme happens to be 5 mutations
distant from the canonical version.
Furthermore, the searchers have demonstrated that, in that case,
theres is a mutation pathway where ALL the intermediate forms
conferred increased anti-antibiotic functionality.
So, in that case, we have a smooth pathway to a NEW FUNCTIONALITY, a
100,000-fold improvement in antibiotic resistance.
While you're at it, do comment on the paper in the January issue of
Current Biology: "Continuous molecular evolution of protein-domain
structures by single amino acid changes. " Meier S, Jensen PR, David
CN, Chapman J, Holstein TW, Grzesiek S, Ozbek S. (2007)
Bobby Bryant
"_Arthur" <Art...@sympatico.ca> writes:
> D. Pitman, are you going to comment the new paper in Nature,
> "Empirical fitness landscapes reveal accessible evolutionary paths"
> by Frank J. Poelwijk, Daniel J. Kiviet, Daniel M. Weinreich and Sander
> J. Tans ?
>
> The paper seems to address your existential fears about hypothetical
> barriers.
s/fears/hopes/
--
Bobby Bryant
Reno, Nevada
Remove your hat to reply by e-mail.
Seanpit
> D.Pitman, are you going to comment the new paper in Nature,
> "Empirical fitness landscapes reveal accessible evolutionary paths"
> by Frank J. Poelwijk, Daniel J. Kiviet, Daniel M. Weinreich and Sander
> J. Tans ?
>
> The paper seems to address your existential fears about hypothetical
> barriers.
>
> They studied a particular enzyme which is commonplace in bacteria, and
> a "mutant" version, which is 100,000 times more efficient against
> antibiotics like penicillin.
The barriers I'm talking about are minimum structural threshold
barriers between what exists in a pool of options and the minimum
structural thresholds of novel potentially beneficial functions. Once
you have any type of function to at least some useful level of
functionality, further refinements, even to the point of 100,000 times
the prior efficiency, aren't a problem. The problem is getting the
particular function, like penicillinase, to even a minimum degree of
usefulness.
The penicillinase function, in particular, has never been shown to
evolve de novo - without this function already existing in the gene
pool to at least some degree of useful activity. This is interesting
partly because the penicillinase function does not have a very high
minimum structural threshold requirement (i.e., no more than 350 or so
specified residues at minimum).
> The new version of the protein/enzyme happens to be 5 mutations
> distant from the canonical version.
> Furthermore, the searchers have demonstrated that, in that case,
> theres is a mutation pathway where ALL the intermediate forms
> conferred increased anti-antibiotic functionality.
Yes, there are a large number of these single-step pathways that have
been identified for refinement of the same type of function where the
minimum threshold requirements have already been reached. The immune
system is another fine example of this sort of thing - of improvement
of the antibody binding function over time.
> So, in that case, we have a smooth pathway to a NEW FUNCTIONALITY, a
> 100,000-fold improvement in antibiotic resistance.
That's right - and there are many more examples of this sort of
refinement activity over time. It's a bit more difficult though when
you haven't reached the minimum threshold requirements for the
function in question at all. Gaining this threshold is the issue
here.
> While you're at it, do comment on the paper in the January issue of
> Current Biology: "Continuous molecular evolution of protein-domain
> structures by single amino acid changes. " Meier S, Jensen PR, David
> CN, Chapman J, Holstein TW, Grzesiek S, Ozbek S. (2007)
Tell me, does this paper talk about the observed evolution of even one
novel function that requires a minimum of more than 1000 specifically
arranged amino acid residues? If not, I'm not interested.
Sean Pitman
www.DetectingDesign.com
Seanpit
wrote:
> >http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=1181141&blobtyp...
>
> > So, you see, even at very low structural threshold levels, the
> > distribution of potentially beneficial structures is widespread and
> > fairly uniform. The clustering effects that are present at lower
> > levels rapidly decline at higher and higher threshold levels until the
> > high-level islands are very remotely separated in sequence space.
>
> Thanks for the link,Sean. I haven't studied it carefully yet, but my
> first impression is that they are not looking at (a sample of) real
> biological proteins. Instead, they are looking at computer-generated
> sets of hypothetical proteins which (based on their structural model)
> ought to fold stably and reliably. But I need to read more carefully.
>
> However, I was interested in your parenthetical comment in this:
> No "higher order" clustering (i.e., except the trivial case
> of the homologous sequences) is visible.
> I have mostly responded to your postings as if they were merely a
> critique of the evolutionary model rather than positive arguments for
> some other model. But I am led to wonder how your favorite model
> accounts for all that 'trivial' homology. Take, for example, the
> 20 or so aminoacyl-tRNA-synthases. Each shows considerable homology
> across the entire tree of life (with the trees constructed from any
> one of them being consistent with the others, except near the roots).
> Furthermore, it seems that they are all homologous with each other -
> suggesting that they arose from a common protein ancestor before the
> organism-level LUCA. Why would that pattern arise in some other
> model besides the evolutionary one?
What about common design? - Why reinvent the wheel every time a wheel
would work just fine?
> Furthermore, there are partial homologies between these proteins and
> many, many others - especially in the 'nucleotide fold' domain. That
> certainly strikes me as a kind of 'clustering'.
I agree, but it isn't the same thing as the clustering of all types of
protein-based systems into one tiny corner of sequence space. This
limited clustering most certainly produces more closely spaced islands
of beneficial sequences. However, these islands do indeed drift
farther and farther apart as one moves up the ladder of minimum
structural threshold requirements.
> Do you hypothesize
> that this 'nucleotide fold' motif is the only good way to solve the
> problem of binding ATP? Isn't it more likely that it is the first
> way of binding ATP to be discovered and that it was then copied into
> other proteins that needed this fragment of function?
It most certainly isn't the only potentially viable way to achieve
this function. All functions have many possible structural solutions.
However, for ever one possibility that would work, there are trillions
upon trillions that won't work - and this ratio decreases,
exponentially, as one moves up the ladder of minimum structural
threshold requirements. This exponentially declining ratio rapidly
gives rise to the insurmountable non-beneficial gap problem and
evolutionary mechanisms simply stall out.
Sean Pitman
www.DetectingDesign.com
Seanpit
> >http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=1181141&blobtyp...
>
> > So, you see, even at very low structural threshold levels, the
> > distribution of potentially beneficial structures is widespread and
> > fairly uniform. The clustering effects that are present at lower
> > levels rapidly decline at higher and higher threshold levels until the
> > high-level islands are very remotely separated in sequence space.
>
> >SeanPitmanwww.DetectingDesign.com-Hide quoted text -
>
> > - Show quoted text -
>
> My first reading of this paper doesn't seem to indicate any support
> for your neutral gaps model being a problem. They are talking about
> small peptides that can retain a basic structure with a large number
> of "neutral substitutions, but they make no conclusions about the
> activity of such structures. It is known from gene families that
> basic structure can be retained while function is modified.
The question here is one of the likely distribution of potentially
beneficial sequences within sequence space. Is this distribution
tightly clustered in one tiny corner of sequence space? Or, are these
potentially beneficial sequences distributed throughout sequence
space? The author's work and evaluation of existing protein-based
systems strongly support the latter view.
Certainly a structure can be modified with a corresponding
modification in the degree of function of a pre-established system.
However, the likelihood that a novel system of function can be found
in this manner is unlikely and becomes exponentially more and more
unlikely as one considers higher and higher minimum structural
threshold requirements.
> So how would studies like this do anything to change the fact that you
> don't have a model that you can support in any fashion worth talking
> about. What is your alternative explanation for the evolution of
> proteins and what is your evidence for it. Recent genome studies
> indicate that gene duplication is a lot more frequent than we had been
> able to detect before. Gene duplication seems to be a mechanism that
> would explain many protein families. Do you have a better
> explanation? What is your evidence for your alternative? Claiming
> that there is a problem when you don't even have a clue about a
> possible viable alternative is pretty bogus don't you think?
Are you actually trying to argue that a theory cannot be questioned
until one has some other way of explaining a particular phenomenon?
Really? That's a good one!
Beyond the fact that one doesn't have to have an answer before one can
raise questions about a proposed hypothesis or theory, there is indeed
a very good alternative to assuming evolutionary mechanisms did the
job in every case. Intelligent design is quite capable and should at
least be considered in such cases - just as would be proposed if
certain radiosignals were discovered by SETI scientists.
> Are you ever going to get around to presenting your alternative to
> common descent and your evidence for your alternative that you claimed
> to have that was better than the current scientific model? You
> claimed to have such a model and evidence, so why putz around with
> bogus neutral gaps if you have some killer evidence for your beliefs?
> Let's see it.
What do you think SETI scientists are looking for? They are in fact
looking for something that has such a large gap between it and what
non-deliberate processes are capable of achieving that some smart
alien intelligence is the most viable option. A clear demonstration
of non-beneficial gaps does the same thing for demonstrating that
evolutionary mechanisms are most unlikely to have crossed the gap
leaving intelligent deliberate processes as the most likely origin.
> Ron Okimoto
Sean Pitman
www.DetectingDesign.com
Perplexed in Peoria
"Seanpit" <seanpi...@naturalselection.0catch.com> wrote in message news:1170624588.1...@j27g2000cwj.googlegroups.com...
> On Feb 4, 11:06 am, "Perplexed in Peoria" <jimmene...@sbcglobal.net>
> wrote:
> > However, I was interested in your parenthetical comment in this:
> > No "higher order" clustering (i.e., except the trivial case
> > of the homologous sequences) is visible.
> > I have mostly responded to your postings as if they were merely a
> > critique of the evolutionary model rather than positive arguments for
> > some other model. But I am led to wonder how your favorite model
> > accounts for all that 'trivial' homology. Take, for example, the
> > 20 or so aminoacyl-tRNA-synthases. Each shows considerable homology
> > across the entire tree of life (with the trees constructed from any
> > one of them being consistent with the others, except near the roots).
> > Furthermore, it seems that they are all homologous with each other -
> > suggesting that they arose from a common protein ancestor before the
> > organism-level LUCA. Why would that pattern arise in some other
> > model besides the evolutionary one?
>
> What about common design? - Why reinvent the wheel every time a wheel
> would work just fine?
Well, common design explains the existence of homology. It doesn't
explain why when you construct twenty different phylogenetic trees
for the twenty different enzymes, those trees all agree with each other
on the shape of the crown groups.
Seanpit
wrote:
> "Seanpit" <seanpitnos...@naturalselection.0catch.com> wrote in messagenews:1170624588.1...@j27g2000cwj.googlegroups.com...
It does if the different groups have different by slightly different
functional needs for their slightly different systems and
environments. Nested patterns are often present within human-designed
systems.
Sean Pitman
www.DetectingDesign.com
_Arthur
0catch.com> wrote:
> > So, in that case, we have a smooth pathway to a NEW FUNCTIONALITY, a
> > 100,000-fold improvement in antibiotic resistance.
>
> That's right - and there are many more examples of this sort of
> refinement activity over time. It's a bit more difficult though when
> you haven't reached the minimum threshold requirements for the
> function in question at all. Gaining this threshold is the issue
> here.
>
You miss the point. The enzyme had another use, unrelated to
antibiotic resistance, and offered a negigible protection against
antibiotics.
If you are unwilling to call a 100,000-fold increase in activity A NEW
FUNCTIONALITY, then nothing at all will ever qualify as such, under
your unstated criteria.
> > While you're at it, do comment on the paper in the January issue of
> > Current Biology: "Continuous molecular evolution of protein-domain
> > structures by single amino acid changes. " Meier S, Jensen PR, David
> > CN, Chapman J, Holstein TW, Grzesiek S, Ozbek S. (2007)
>
> Tell me, does this paper talk about the observed evolution of even one
> novel function that requires a minimum of more than 1000 specifically
> arranged amino acid residues? If not, I'm not interested.
>
Yes, the paper makes no mention at all of any barrier linked to 1000,
or 666, or any other arbitray number of mutations. So, of course, you
wouldn't be interrested.
After all, there is no use for you to keep abreast of current
research, since the Pitman Law isn't based on biochemical research.
Biochemist Micheal Behe was so confident in his discovery of the
unevolvability of the immune system, that he didn't bother to follow
the research in the field. That level of confidence made a big
impression on the judge.
Seanpit
> On Feb 4, 4:23 pm, "Seanpit" <seanpitnos...@naturalselection.
>
> 0catch.com> wrote:
> > > So, in that case, we have a smooth pathway to a NEW FUNCTIONALITY, a
> > > 100,000-fold improvement in antibiotic resistance.
>
> > That's right - and there are many more examples of this sort of
> > refinement activity over time. It's a bit more difficult though when
> > you haven't reached the minimum threshold requirements for the
> > function in question at all. Gaining this threshold is the issue
> > here.
>
> You miss the point. The enzyme had another use, unrelated to
> antibiotic resistance, and offered a negigible protection against
> antibiotics.
> If you are unwilling to call a 100,000-fold increase in activity A NEW
> FUNCTIONALITY, then nothing at all will ever qualify as such, under
> your unstated criteria.
And you miss the point, if each single mutation is selectably
beneficial, as you claimed, then the antibiotic activity was not
negligible along the pathway but was selectable during each step in a
positive manner until the final 100,000-fold increase was realized.
Beyond this, even if the original starting point had no selectable
activity for the resulting function, the threshold level in this case
is only a few hundred specified residues - not even close to my
challenge of 1000 specified amino acid residues.
> > > While you're at it, do comment on the paper in the January issue of
> > > Current Biology: "Continuous molecular evolution of protein-domain
> > > structures by single amino acid changes. " Meier S, Jensen PR, David
> > > CN, Chapman J, Holstein TW, Grzesiek S, Ozbek S. (2007)
>
> > Tell me, does this paper talk about the observed evolution of even one
> > novel function that requires a minimum of more than 1000 specifically
> > arranged amino acid residues? If not, I'm not interested.
>
> Yes, the paper makes no mention at all of any barrier linked to 1000,
> or 666, or any other arbitray number of mutations. So, of course, you
> wouldn't be interrested.
What is the size of the evolved sequences? Hmmmm?
> After all, there is no use for you to keep abreast of current
> research, since the Pitman Law isn't based on biochemical research.
There are literally thousands of papers like this that all say pretty
much the same thing. As far as I am aware, not one of them actually
presents the observed evolution any system that requires a minimum
structural threshold of more than 1000 specified residues.
> Biochemist Micheal Behe was so confident in his discovery of the
> unevolvability of the immune system, that he didn't bother to follow
> the research in the field. That level of confidence made a big
> impression on the judge.
Please do present any research that shows any novel function evolving
that requires a minimum structural threshold of more than 1000
specified residues. I'd be most interested. I'm very interested in
following *relevant* research. Otherwise, I read plenty of articles
beyond those that are actually relevant to this discussion. I just
don't have time to read everything that everybody thinks I should
read. So, why not present something that's actually relevant for a
change? - along with a relevant quote for once?
Sean Pitman
www.DetectingDesign.com
Seanpit
protein systems within sequence space is presented in the following
link of the work of Jim Proctor and Andrew Torda (Hamburg University,
last update May, 2004):
http://www.zbh.uni-hamburg.de/wurst/protspace/ECCB_opt.huge.pdf
http://www.zbh.uni-hamburg.de/wurst/protspace/
Sean Pitman
www.DetectingDesign.com
Seanpit
_Arthur
0catch.com> wrote:
> Please do present any research that shows any novel function evolving
> that requires a minimum structural threshold of more than 1000
> specified residues.
You're the one with the bogus theory. Name me 3 known proteins that
you know for a fact that are 1000 amino-acids away than ANY possible
precursor known or unknown, with ANY possible biologically
And, while you were shifting goalposts, you aknowledged than it
doesn't seems to matter that an enzyme gains or lose a chemical
property by a factor of 100,000 in the process, it is still selectable
-- according to you.
For example, suppose a cell makes an enzyme that has an useful -- and
measurable -- capacity at digesting lactase. And that particular
enzyme has neglible efficiency at breaking down, say, crude oil.
Myself, I see no problem of envisionning a scenario where, after a
certain number of mutations (--does the number really matter? Why --
How ?) the mutated enzyme still presents a marginal lactase activity
but also a faint crude oil degrading functionality. And according to
your newly formulated criteria, the degree of functionality is
immateral, once aquired it can be selected for and improved a
thousandfold, nay, make that 100,000-fold.
Of course in their natural environment, most bacteria have no use for
a lactase-digesting enzyme, there are very few natural sources of
lactase. Or crude oil. Or penicilin.
So name me 3 known human or bacteria proteins that are surrounded by
an impassable moat of unfruited variants utterly devoid of ANY useful
chemical property, 1000 amino-acids deep.
Name me 3 first such proteins that come to your mind.
Robin Levett
Sean, before you throw out more references to papers that you can distort to
fit your requirements, isn't it about time you addressed the Choi and Kim
paper that Howard referred you to on 16 October last year? The one which
showed that proteins *in the real world* cluster into 4 families?
www.pnas.org/cgi/content/full/103/38/14056
Oh, and even a layman like me can see that you've got your interpretation of
the Bornberg-Bauer paper wrong. He specifically states that the HP protein
sequences he is examining are not uniformly distributed across sequence
space; see the last couple of paragraphs of the paper; and the RNA
sequences, that he says do "percolate through sequence space", do so in
such a way that "Within a number of mutations small compared to the length
of the sequence, the whole shape-space can be covered".
Even in the abstract he makes clear that "In analogy to protein families,
nets are dense and well-separated in sequence space".
--
Robin Levett
rle...@rlevett.ibmuklunix.net (unmunge by removing big blue - don't yahoo)
Ron O
0catch.com> wrote:
> On Feb 4, 11:28 am, "Ron O" <rokim...@cox.net> wrote:
>
>
>
>
>
> > >http://www.pubmedcentral.nih.gov/picrender.fcgi?artid=1181141&blobtyp...
>
> > > So, you see, even at very low structural threshold levels, the
> > > distribution of potentially beneficial structures is widespread and
> > > fairly uniform. The clustering effects that are present at lower
> > > levels rapidly decline at higher and higher threshold levels until the
> > > high-level islands are very remotely separated in sequence space.
>
>
> > > - Show quoted text -
>
> > My first reading of this paper doesn't seem to indicate any support
> > for your neutral gaps model being a problem. They are talking about
> > small peptides that can retain a basic structure with a large number
> > of "neutral substitutions, but they make no conclusions about the
> > activity of such structures. It is known from gene families that
> > basic structure can be retained while function is modified.
>
> The question here is one of the likely distribution of potentially
> beneficial sequences within sequence space. Is this distribution
> tightly clustered in one tiny corner of sequence space? Or, are these
> potentially beneficial sequences distributed throughout sequence
> space? The author's work and evaluation of existing protein-based
> systems strongly support the latter view.
No the real question is why you fool around with bogus junk like this,
that even you have to admit doesn't amount to squat in the face of all
the other data that we have, and don't bother to put up or shut up.
"Put up or shut up" is just an expression that you are probably well
aware of. It denotes frustration at the fact that you can obfuscate
the issue a dozen different ways, but you never come out and
demonstrate that you really have anything to discuss. If you aren't
willing to put forward the wonderful stuff that you claim to have
supporting your idea of what happened why do you spend so much time
shooting around the edges and never hitting the target? You make a
lot of noise, but that is all it is. What is your alternative, what
is the evidence for your alternative, and do a real critical analysis
of how it stacks up to the notion that you don't like. That is how an
honest person would approach the issue. They wouldn't keep lying
about what they have and pretending it is just too good to put
forward.
Getting back to this study, why would they have found anything any
different? The distributions donot tell you anything about how much
sequence space has to be searched to do a certain function. Sequence
space is so huge that even if you ran the simulation to check 10E12
sequences where would they all fall? There is also the fact that
their simulation also supports what we see in the molecular record.
Yockey came up with a shaky estimate that based on just what was known
about the various sequences when he made the estimate in the late
1970's that there were over 10E40 different cytochrome c sequences
that would function as cyt c just limiting the sequence to 104 amino
acid residues. These guys demonstrate that in their simulations over
half the amino acids residues can be replaced by mutation and the
internal structure isn't altered that much. They also admit that they
are only using hydrophobicity and we know that disulfide bonds are
very important, and that salt bridges (attraction between acidic and
basic amino acid R groups) are important. We can see that such sites
have changed in concert.
So where does this study leave you in terms of what you think is going
on? How could it possibly help the fact that your model is hopelessly
inadequate to explain the existing data. If you don't think that it
is, put it forward for evaluation. You can't deny that you were the
one that claimed that you had such a model and evidence, so why not
put it forward? Why jerk around with junk like this?
>
> Certainly a structure can be modified with a corresponding
> modification in the degree of function of a pre-established system.
> However, the likelihood that a novel system of function can be found
> in this manner is unlikely and becomes exponentially more and more
> unlikely as one considers higher and higher minimum structural
> threshold requirements.
Where do you get that from this paper? Any limited number of
sequences is going to have large gaps between them in sequence space.
What you need to know is how many of those sequences can provide the
functions that life needs. They do not appear to provide this answer
in the paper. Where is it?
>
> > So how would studies like this do anything to change the fact that you
> > don't have a model that you can support in any fashion worth talking
> > about. What is your alternative explanation for the evolution of
> > proteins and what is your evidence for it. Recent genome studies
> > indicate that gene duplication is a lot more frequent than we had been
> > able to detect before. Gene duplication seems to be a mechanism that
> > would explain many protein families. Do you have a better
> > explanation? What is your evidence for your alternative? Claiming
> > that there is a problem when you don't even have a clue about a
> > possible viable alternative is pretty bogus don't you think?
>
> Are you actually trying to argue that a theory cannot be questioned
> until one has some other way of explaining a particular phenomenon?
> Really? That's a good one!
No, we question our theories all the time, but nothing much changes
until a better explanation comes along that is more useful and
hopefully accurate than the one before. You can't just nit pick in
science. To actually do something you have to actually accomplish
something. No scientific theory is perfect, that is a fact and isn't
any news to anyone involved in science. Obfuscating the issue so that
you can continue to lie to yourself is not only counter productive,
but isn't worth jack in science or the real world except make your
self feel better somehow or to suck money out ot the rubes or make
them bend to your political whims because they really think that you
might have something. Do you still believe that you could do better
than the Ohio rubes and teach the real science of ID? It turned out
that you had been taken in by the IDiot lies and either were clueless
about how bogus ID science was or you were lying about how clueless
you were. Face the facts, you were either scammed by ID or you were
one of the scammers. Pick one and defend it.
>
> Beyond the fact that one doesn't have to have an answer before one can
> raise questions about a proposed hypothesis or theory, there is indeed
> a very good alternative to assuming evolutionary mechanisms did the
> job in every case. Intelligent design is quite capable and should at
> least be considered in such cases - just as would be proposed if
> certain radiosignals were discovered by SETI scientists.
Raise questions and then figure out how to do the research to solve
the problem. That is how science works. If we knew everything there
would be no science just technicians and librarians. You can't just
nay say because we all know (even you in your more lucid moments) that
your alternative is so much worse than the one that you don't like
that you don't even try to put it forward. Why not put it forward and
nit pick it? It would be much more fruitful and you can get definite
answers about what is wrong with it. That is the difference between
your nit picking and what happened to your notion. This sequence
space junk is at the limits of our understanding and we are far from
understanding it. Your model failed long ago based on well known and
accepted principles. A better alternative came along and it was
demonstrated to be better in so many ways that you don't stand a
chance. All you can do is obfuscate the issue and pretend that your
nit picking really matters, when if you did the same thing to your
alternative it would be no contest on which idea came up short.
>
> > Are you ever going to get around to presenting your alternative to
> > common descent and your evidence for your alternative that you claimed
> > to have that was better than the current scientific model? You
> > claimed to have such a model and evidence, so why putz around with
> > bogus neutral gaps if you have some killer evidence for your beliefs?
> > Let's see it.
>
> What do you think SETI scientists are looking for? They are in fact
> looking for something that has such a large gap between it and what
> non-deliberate processes are capable of achieving that some smart
> alien intelligence is the most viable option. A clear demonstration
> of non-beneficial gaps does the same thing for demonstrating that
> evolutionary mechanisms are most unlikely to have crossed the gap
> leaving intelligent deliberate processes as the most likely origin.
This is a stupid analogy because SETI is looking for a signal that
they can differentiate from a natural occurring signal. If all they
were looking for is something far away from what is known, three
pulsars in close proximity would be considered a SETI signal because
if it did occur only once in the Universe that would be as far away as
they could get. Get a clue. There are certain criteria that they
have to look for, and if and when they ever detect something the
debate will be one as to what that something is. If all they had was
looking at the different proteins and claiming that there was
intelligent signal there they would be doomed because we see those
protein sequences evolving even today.
Just demonstrate that your gaps exist, and then look at your
demonstration and compare it to what is wrong with your model and tell
us which one comes up short. Why keep obfuscating when you could nail
everyone with your model and evidence? Why can you see all that is
wrong with what you don't like, but you aren't willing to do the same
thing to your own model? Your methodology not only sounds bogus it is
bogus and you know it at some level, but insanity could be a defense.
You weren't too coherant when you were blathering about how you could
do better about teaching ID than the Ohio rubes, and I've pretty much
left your alone because of that mental break, but before you put up
your wonderful science of ID I'd like to see your alternative and the
evidence to back it up. Why claim that you have it if you aren't
willing to put it forward?
Your obfuscation tactics aren't just a little problem are they?
Obfuscation is all you have for the simple reason that you have
nothing else worth putting into the argument. If this isn't true, put
your junk forward and evaluate it for everyone. Someone that can talk
about problems in sequence space should be able to tell us why your
alternative comes up short compared to common descent, right?
Ron Okimoto
>
> > Ron Okimoto
>
> Sean Pitmanwww.DetectingDesign.com- Hide quoted text -
Seanpit
> On Feb 4, 5:19 pm, "Seanpit" <seanpitnos...@naturalselection.
>
> 0catch.com> wrote:
> > Please do present any research that shows any novel function evolving
> > that requires a minimum structural threshold of more than 1000
> > specified residues.
>
> You're the one with the bogus theory. Name me 3 known proteins that
> you know for a fact that are 1000 amino-acids away than ANY possible
You don't seem to understand the challenge. I'm not asking you to
present a case of a function evolving that was 1000aa away from any
possible precursor. I'm asking you to present a case of a function
evolve that requires of minimum of 1000 specifically arranged amino
acid residues to work - regardless of how far or close it may or may
not be from what came before.
Again, the gap distance is not the same thing as the minimum
structural threshold requirement. They are related, but they are not
the same. The likely gap distance is always much smaller than the
minimum threshold size.
> And, while you were shifting goalposts, you aknowledged than it
> doesn't seems to matter that an enzyme gains or lose a chemical
> property by a factor of 100,000 in the process, it is still selectable
> -- according to you.
That's what you claimed the author's of the paper your referenced said
- that each step along the pathway was functionally selectable. If
that is true, there is really no problem at all since there is no
functionally neutral gap. You started with an intact functional
system that already had, to a selectable degree, the function in
question - or isn't that really what the author's claimed?
> For example, suppose a cell makes an enzyme that has an useful -- and
> measurable -- capacity at digesting lactase. And that particular
> enzyme has neglible efficiency at breaking down, say, crude oil.
> Myself, I see no problem of envisionning a scenario where, after a
> certain number of mutations (--does the number really matter? Why --
> How ?) the mutated enzyme still presents a marginal lactase activity
> but also a faint crude oil degrading functionality. And according to
> your newly formulated criteria, the degree of functionality is
> immateral, once aquired it can be selected for and improved a
> thousandfold, nay, make that 100,000-fold.
That's right. Once the threshold requirements of the new function are
acquired to a selectable degree, further refinements aren't a problem.
> Of course in their natural environment, most bacteria have no use for
> a lactase-digesting enzyme, there are very few natural sources of
> lactase. Or crude oil. Or penicilin.
Right - -
> So name me 3 known human or bacteria proteins that are surrounded by
> an impassable moat of unfruited variants utterly devoid of ANY useful
> chemical property, 1000 amino-acids deep.
> Name me 3 first such proteins that come to your mind.
Again, that's not my claim for my 1000aa threshold (see above
explanation). There are many types of functions that require far more
than 1000 specifically arranged residues to work at all - even a
little bit. The flagellar motility system is a classic example,
requiring well over 10,000 specified residues. Ameboid motility,
exocitosis, pinocytosis, photosynthesis, eye-spot vision systems, etc.
all require thousands of specifically arranged amino acid residues
working together at the same time.
Sean Pitman
www.DetectingDesign.com
Seanpit
< snip >
> > What do you think SETI scientists are looking for? They are in fact
> > looking for something that has such a large gap between it and what
> > non-deliberate processes are capable of achieving that some smart
> > alien intelligence is the most viable option. A clear demonstration
> > of non-beneficial gaps does the same thing for demonstrating that
> > evolutionary mechanisms are most unlikely to have crossed the gap
> > leaving intelligent deliberate processes as the most likely origin.
>
> This is a stupid analogy because SETI is looking for a signal that
> they can differentiate from a natural occurring signal. If all they
> were looking for is something far away from what is known, three
> pulsars in close proximity would be considered a SETI signal because
> if it did occur only once in the Universe that would be as far away as
> they could get. Get a clue. There are certain criteria that they
> have to look for, and if and when they ever detect something the
> debate will be one as to what that something is. If all they had was
> looking at the different proteins and claiming that there was
> intelligent signal there they would be doomed because we see those
> protein sequences evolving even today.
Not true. We do not see protien sequences evolving even today at all
beyond very low levels of functional complexity - not a single example
goes beyond the 1000 specified amino acid residue structural threshold
requirement. Therefore, finding such a functional system significantly
beyond this threshold is most certainly the same thing as "a signal
that they [SETI] can differentiate from a natural occurring signal."
Why? Because it does not occur naturally at all and statistically it
cannot occur this side of a practical eternity of time.
< snip >
Sean Pitman
www.DetectingDesign.com
Seanpit
> Seanpit wrote:
> > An interesting 3D model for the distribution of existing small single-
> > protein systems within sequence space is presented in the following
> > link of the work of Jim Proctor and Andrew Torda (Hamburg University,
> > last update May, 2004):
>
> >http://www.zbh.uni-hamburg.de/wurst/protspace/ECCB_opt.huge.pdf
> >http://www.zbh.uni-hamburg.de/wurst/protspace/
>
> Sean, before you throw out more references to papers that you can distort to
> fit your requirements, isn't it about time you addressed the Choi and Kim
> paper that Howard referred you to on 16 October last year? The one which
> showed that proteins *in the real world* cluster into 4 families?
>
> www.pnas.org/cgi/content/full/103/38/14056
If you read the paper carefully, Choi and Kim do not say that proteins
form four families. Rather, they argue that protein folds can be
roughly categorized into four families. That's a big difference.
Also, consider the actual 3D model of their sequence space:
http://www.lbl.gov/Science-Articles/Archive/PBD-Universe-map-Kim.html
Notice in this model that even when it comes to very low-level protein
folding sequence space, the viable folds do not all overlap. Sure,
they are close together, but this is only to be expected at this very
low level of sequence space. It's kinda like the sequence space of 3-
letter words. They are all very closely clustered together like
this. However, this closeness of the potential steppingstones rapidly
expands as one moves up the ladder of minimum structural threshold
requirements.
> Oh, and even a layman like me can see that you've got your interpretation of
> the Bornberg-Bauer paper wrong. He specifically states that the HP protein
> sequences he is examining are not uniformly distributed across sequence
> space; see the last couple of paragraphs of the paper; and the RNA
> sequences, that he says do "percolate through sequence space", do so in
> such a way that "Within a number of mutations small compared to the length
> of the sequence, the whole shape-space can be covered".
That's right! The *whole* shape-space can be covered via relatively
small steps. This in fact means that the steppingstones here are
indeed distributed throughout sequence space. They are not clustered
together in one tiny corner. The fact that one can get across the
whole of sequence space without having to step very far from one to
the next steppingstone simply means that the ratio of steppingstones
vs. the non-viable "gap" options is rather high at this level of
minimum structural threshold requirements - as in the case of 3-letter
words. However, as you move up the ladder of threshold requirements,
this ratio drops off exponentially. As this happens, the
steppingstones drift apart rapidly. Pretty soon, it becomes very
difficult to take the same short steps and hit any steppingstone at
all.
> Even in the abstract he makes clear that "In analogy to protein families,
> nets are dense and well-separated in sequence space".
Exactly! And they cover the entire space! The nets are dense because
of the relatively high ratio that exists at such low levels of minimum
structural threshold requirements.
> Robin Levett
Sean Pitman
www.DetectingDesign.com
_Arthur
to count them ?
Also, the flagellar motility system is not a single protein, but an
assembly of approximatively 42 unique proteins.
Are you telling me that nothing prevents any single of those proteins
to be evolved individually by the cell, but that the whole *flagella*
is surrounded by some invisible "Pitman Gap" with the magic number
"1000aa" thrown in somewhere ?
I must know where you place your goalposts, you seem to be hiding them
begind a screen of fuzzy and ever-changing non-definitions and weasel-
concepts.
richardal...@googlemail.com
WHO specified the "1000 specified amino acid residue structural
threshold" and WHY is it relevant?
You have given no reason to think that it is anything other than an
arbitrary number you have pulled out of the air to falsify an
inaccurate model of evolution you have devised for no reason other
than to "falsify" it.
Novel functions can evolve without the need to cross your "1000
specified amino acid residue structural threshold".
There is no mechanism which prevents further evolution.
You may be able to fool the creationists with this sort of drivel, but
it is very clear that you can't fool anyone who knows anything about
the subject. Evidently your motive is to impress the creationists, not
to persuade the scientists, but as your whole argument is built on a
fabric of misrepresentation, distortion and outright falsehoods it is
deeply dishonest.
Evidently you are happy to use such dishonest methods to support your
case. You present your arguments here, and on your web sites knowing
perfectly well that they are bogus, based on unfounded assertion at
best and on outright falsehoods at worst, and would be utterly
demolished by any scientist with any knowledge of the subject were
they submitted to any form of academic review.
If you had a scrap of honesty you would submit your ideas to review.
I'm totally confident that you won't.
RF
Seanpit
> Remind me again what's a "specified residue", and how one is supposed
> to count them ?
All protein-based systems of function require a certain number of
specifically arranged amino acid residues. While flexibility of
arrangement can be tolerated, the degree of flexibility is often
fairly restricted. Several papers have been published about this sort
of thing to include papers by Sauer, Olsen, and Yockey. Estimates of
flexibility produce ratios for fairly specified 100aa systems of
around 1e-35 to 1e-65 or so.
> Also, the flagellar motility system is not a single protein, but an
> assembly of approximatively 42 unique proteins.
Actually, the number is a bit less than 42, but that's not the point.
The point is that all of the proteins (which are each made up of two
or three hundred fairly specified amino acid residues) in the
flagellar motility system have to be specifically arranged for the
overall system to work.
> Are you telling me that nothing prevents any single of those proteins
> to be evolved individually by the cell, but that the whole *flagella*
> is surrounded by some invisible "PitmanGap" with the magic number
> "1000aa" thrown in somewhere ?
The relatively small single proteins aren't a problem. What is a
problem is the assembly of many single proteins to produce a system
that requires a total of many thousands of amino acid residues to be
specifically arranged with each other at the same time. You just
can't add one protein together to gain the flagellar motility system
because of a lack of beneficial steppingstones that are actually that
close together.
The gaps between steppingstones at this level may not seem very big,
only a few dozen character changes in places along the pathway, but
these seemingly small gaps would require evolutionary mechanisms
trillions upon trillions of years to cross.
> I must know where you place your goalposts, you seem to be hiding them
> begind a screen of fuzzy and ever-changing non-definitions and weasel-
> concepts.
LOL - My goalposts have been just were they are now for many years.
My definitions have not changed. You just don't seem to understand
concepts that really are quite simple and intuitively obvious.
Sean Pitman
www.DetectingDesign.com
_Arthur
0catch.com> wrote:
> The relatively small single proteins aren't a problem. What is a
> problem is the assembly of many single proteins to produce a system
> that requires a total of many thousands of amino acid residues to be
> specifically arranged with each other at the same time. You just
> can't add one protein together to gain the flagellar motility system
> because of a lack of beneficial steppingstones that are actually that
> close together.
Why the blather about *protein* structures, why all along you really
meant the total combination of all the genes necessary for a multi-
protein structure ?
Ron O
Pathetic evasion. Sean what is good enough? Show us what is good
enough for you. Put up your alternative and the evidence for it so
that we can see just how much your bullpucky of only low level changes
in functional complexity really matters. If what we have is so bad,
what does that tell you about the pathetic junk that you have to
believe means anything? Really, show us what you think is good
enough.
Ron Okimoto
>
> < snip >
Seanpit
> > Not true. We do not see protein sequences evolving even today at all
> > beyond very low levels of functional complexity - not a single example
> > goes beyond the 1000 specified amino acid residue structural threshold
> > requirement.
>
> WHO specified the "1000 specified amino acid residue structural
> threshold" and WHY is it relevant?
If you look in scientific literature you will see something very
interesting. There are many examples of novel protein-based systems
evolving. However, all of these systems require significantly less
than 1000 specified amino acid residues. There isn't a single
example, in all of literature, of any real time evolution in action
that produces any novel function that requires at least 1000
specifically arranged residues working at the same time. Beyond this,
the number of examples drops off dramatically as one approaches this
1000aa mark.
If you don't understand why that is relevant, you must be deliberately
blind.
> You have given no reason to think that it is anything other than an
> arbitrary number you have pulled out of the air to falsify an
> inaccurate model of evolution you have devised for no reason other
> than to "falsify" it.
Show me just one example then of observed evolution in action beyond
this structural threshold.
> Novel functions can evolve without the need to cross your "1000
> specified amino acid residue structural threshold".
Novel functions that require at least 1000 specifically arranged
residues have never been observed to evolve. There are no real time
examples - period. Not one. Those functions that do evolve require
only a few hundred specified residues - like your cockroach milk
evolution example (only about 200 or so loosely specified residues
needed), which isn't even an observed example.
> There is no mechanism which prevents further evolution.
Then why doesn't evolution happen beyond the 1000aa threshold? What
makes evolution so common at levels below the 1000aa threshold, but
completely absent beyond this level?
> You may be able to fool the creationists with this sort of drivel, but
> it is very clear that you can't fool anyone who knows anything about
> the subject. Evidently your motive is to impress the creationists, not
> to persuade the scientists, but as your whole argument is built on a
> fabric of misrepresentation, distortion and outright falsehoods it is
> deeply dishonest.
You don't have the first clue what you are talking about here. You
don't understand your own proposed evolutionary mechanism and are
oblivious to the fact that this mechanism only works at levels well
below the 1000aa threshold.
> Evidently you are happy to use such dishonest methods to support your
> case. You present your arguments here, and on your web sites knowing
> perfectly well that they are bogus, based on unfounded assertion at
> best and on outright falsehoods at worst, and would be utterly
> demolished by any scientist with any knowledge of the subject were
> they submitted to any form of academic review.
Have at it then. Where are your devastating arguments? Hmmmm? Just
one example is all you need. Where is it?
> If you had a scrap of honesty you would submit your ideas to review.
What do you think I'm doing here? Don't you have a real argument of
your own?
> I'm totally confident that you won't.
LOL - whatever . . .
> RF
Sean Pitman
www.DetectingDesign.com
Seanpit
> > > This is a stupid analogy because SETI is looking for a signal that
> > > they can differentiate from a natural occurring signal. If all they
> > > were looking for is something far away from what is known, three
> > > pulsars in close proximity would be considered a SETI signal because
> > > if it did occur only once in the Universe that would be as far away as
> > > they could get. Get a clue. There are certain criteria that they
> > > have to look for, and if and when they ever detect something the
> > > debate will be one as to what that something is. If all they had was
> > > looking at the different proteins and claiming that there was
> > > intelligent signal there they would be doomed because we see those
> > > protein sequences evolving even today.
>
> > Not true. We do not see protein sequences evolving even today at all
> > beyond very low levels of functional complexity - not a single example
> > goes beyond the 1000 specified amino acid residue structural threshold
> > requirement. Therefore, finding such a functional system significantly
> > beyond this threshold is most certainly the same thing as "a signal
> > that they [SETI] can differentiate from a natural occurring signal."
> > Why? Because it does not occur naturally at all and statistically it
> > cannot occur this side of a practical eternity of time.
>
> Pathetic evasion. Sean, what is good enough? Show us what is good
> enough for you.
Do you have problems with reading comprehension Ron? I just pointed
out to you that no functional system that requires a minimum of more
than 1000 specified amino acid residues evolves. Why not go about
showing some examples beyond this threshold limitation? Why not
explain why there are no current examples of evolution in action
beyond this level? Why not explain why evolutionary mechanisms stall
out very rapidly as this threshold is approached?
> Put up your alternative and the evidence for it so
> that we can see just how much your bullpucky of only low level changes
> in functional complexity really matters. If what we have is so bad,
> what does that tell you about the pathetic junk that you have to
> believe means anything? Really, show us what you think is good
> enough.
Just one example would be fine Ron.