Molecular Sequence Proof of Common Descent

1 view
Skip to first unread message

Zeus Thibault

unread,
Jul 16, 1999, 3:00:00 AM7/16/99
to
This is a reworking of an old argument in a way that I have never seen
before; I hope that it will clarify exactly why molecular sequence
similarity is such a powerful technique for demonstrating the
genealogical relatedness of different species.

I. Protein functional redundancy.

Before the advent of DNA sequencing technology, the amino acid sequences
of proteins were used to establish the phylogenetic relationships of
species. Sequence studies with functional genes have centered on genes
of proteins (or RNAs) that are ubiquitous (i.e. all organisms have
them). This is done to insure that the comparisons are independent of
the overall species phenotype.

For example, suppose we are comparing the protein sequence of a
chimpanzee and that of a human. Both of these animals have many similar
anatomical characters, so we might expect their proteins to be similar,
regardless of whether they are genealogically related or not. However,
we can compare the sequences of very basic genes that are used by all
living organisms, such as the cytochrome c gene, which have no influence
over specific chimpanzee or human characteristics.

Cytochrome c is a ubiquitous protein found in all eukaryotes, which is
absolutely required for the viability of all organisms (bacteria, or
prokaryotes, also have cytochrome c proteins, but for historical reasons
they are called c-like cytochromes). Cytochrome c is found in the
mitochondria of eukaryotic cells, where it is a key player in the
fundamental metabolic process of oxidative phosphorylation (this is
where the O2 we breathe is used to generate energy).

Using a gene like this, there is no reason to assume that the protein
sequence should be the same, unless the two organisms are genealogically
related. This is due to the functional redundancy of protein sequences
and structures. By functional redundancy I mean that many different
protein sequences form the same structure and perform the same
function. Decades of biochemical evidence have shown that most amino
acid mutations, especially of surface residues, have no effect on
protein function or on protein structure.

A striking example is that of the c-type cytochromes from various
bacteria, which have no sequence similarity. Nevertheless, they all
fold into the same three-dimensional structure, and they all perform the
same function.

Even within species, most amino acid mutations are functionally silent.
For example, there are at least 250 different amino acid mutations known
in human hemoglobin, carried by more than 3% of the world's population,
that have no clinical manifestation in either heterozygotic or
homozygotic individuals. This is a general phenomenon observed in all
proteins and genes in all species.

However, the most convincing evidence of protein functional redundancy
comes from molecular biology experiments with recombinant organisms.
For example, it is possible to delete the essential cytochrome c gene in
one species and replace it with the gene from another species. This has
been done with species as diverse as horse, wheat, pigeon, mouse, yeast,
and human. In all cases the proteins were biologically functional and
the recombinant organisms were viable.

Consider again the molecular sequences of cytochrome c. In a classic
study of ~100 cytochrome c protein sequences from organisms ranging from
yeast to human, 38 of the 104 amino acids in the protein were found to
be invariant (i.e. they were the same in all organisms studied).
However, 58 of the amino acids could be replaced by up to six different
amino acids, and eight of the amino acids were hypervariable (i.e. they
could be replaced by more than six amino acids).

Making the naive assumption that this represents the maximum number of
possible amino acid variants that allows for functional redundancy
(there are definitely more possibilities than those found in this
limited study), a conservative calculation can be made giving >10^33
different possible functionally redundant protein sequences for the 104
amino acid cytochrome c protein (1^38 x 3^58 x 6^8 > 10^33).

The proof: Man and chimpanzees have the exact same cytochrome c protein
sequence. The chance of this occurrence is less than 10^-33 (1 out of
10^33), unless one assumes genealogical relatedness. Human and
chimpanzee cytochrome c proteins differ by ~10 amino acids from all
other mammals. The chance of this occurring, without a hereditary
mechanism, is conservatively less than 10^-26. Based on this study of
protein sequences, the most distantly related organism from humans is
the yeast Candida krusei, with 51 amino acid differences. A
conservative estimate of this probability is less than 10^-7.

What makes these figures even more convincing is the fact that the
phylogenetic tree constructed from the cytochrome c data exactly
recapitulates the relationships of major classes as determined by the
completely independent morphological data.

The most likely result is that all these protein sequences would be very
different from each other. If this were the case, a phylogenetic
analysis would be impossible, and this would provide very strong
evidence for a genealogically unrelated, perhaps simultaneous, origin of
species.

II. DNA coding redundancy.

Like protein sequence similarity, the DNA sequence similarity of two
ubiquitous genes also implies common ancestry. Of course, comprehensive
DNA sequence comparisons of conserved proteins such as cytochrome c also
indirectly take into account amino acid sequences, since the DNA
sequence specifies the protein sequence.

However, with DNA sequences there is an extra level of redundancy. The
genetic code itself is redundant; on average there are 3 different
codons (a codon is a triplet of DNA bases) that can specify the exact
same amino acid.

Thus, for cytochrome c there are 3^104, or over 10^49, different DNA
sequences (and, hence, 10^49 different possible genes) that can specify
the very exact same protein sequence.

As mentioned above, the cytochrome c proteins in chimps and humans are
exactly the same. The clincher is that the two DNA sequences that code
for cytochrome c in humans and chimps differ by only one codon, even
though there are over 10^49 different sequences that could code for
these two proteins.

Thus, there are really just two choices. Either (1) from probability
considerations, we are 100% sure that humans and chimps are closely
genealogically related, or (2) a designer chose the two DNA sequences
out of the over 10^49 possibilities that make it look exactly like we
are genealogically related.

The combined effects of DNA coding redundancy and protein sequence
redundancy makes DNA sequence comparisons doubly redundant (actually
multiplicatively redundant); DNA sequences of ubiquitous proteins are
completely uncorrelated with phenotype, but are strongly correlated with
heredity. This is why DNA sequence phylogenies are considered so
robust.

The most probable result is that the DNA sequences coding for these
proteins should be radically different. This would be a resounding
falsification of macroevolution, it would be very strong proof that
chimpanzees and humans are not closely genealogically related, and would
be pretty convincing evidence of design. Since the most probable result
is not found, and what is observed is exactly what is expected due to
common descent of these two organisms, any theory of design must explain
this fact as well.

In other words, exactly why did a designer choose the two sequences that
look exactly like these organisms evolved, and not one of the other
10^82 (i.e. 10^49 x 10^33) functionally equivalent sequences?


Chris C.

unread,
Jul 17, 1999, 3:00:00 AM7/17/99
to
We really should say the multi-nested hierarchy.

Jorolat

unread,
Jul 17, 1999, 3:00:00 AM7/17/99
to
Hi,
There`s a cytochrome c percent sequence difference matrix for 21 organisms
in the book "Evolution - A Theory in Crisis".
If the average of the different groups is calculated and divided by a
common factor the reverse fibonacci series appears.
The hierarchical relationship between the groups is useful supporting
evidence of an internal evolutionary mechanism existing based on homeostasis.

http://members.aol.com/jorolat
An Internal Evolutionary Mechanism: A simple Model and Method of Testing
How Psychology colours perception of Natural Realities
Trauma/Bullying/Social Violence/Professional Abusers etc.


z@z

unread,
Jul 17, 1999, 3:00:00 AM7/17/99
to
Hello Zeus Thibault!

What you write is excellent and constitutes maybe the best refutation
of neo-Darwinism I've ever seen. For further details look at the
current thread "ENIGMAS --- Innovations vs. Mutations".
http://www.deja.com/=dnc/getdoc.xp?AN=499894064 (starting post)

| This is a reworking of an old argument in a way that I have never seen
| before; I hope that it will clarify exactly why molecular sequence
| similarity is such a powerful technique for demonstrating the
| genealogical relatedness of different species.

[snip]

| A striking example is that of the c-type cytochromes from various
| bacteria, which have no sequence similarity. Nevertheless, they all
| fold into the same three-dimensional structure, and they all perform the
| same function.
|
| Even within species, most amino acid mutations are functionally silent.
| For example, there are at least 250 different amino acid mutations known
| in human hemoglobin, carried by more than 3% of the world's population,
| that have no clinical manifestation in either heterozygotic or
| homozygotic individuals. This is a general phenomenon observed in all
| proteins and genes in all species.

[snip]

| Consider again the molecular sequences of cytochrome c. In a classic
| study of ~100 cytochrome c protein sequences from organisms ranging from
| yeast to human, 38 of the 104 amino acids in the protein were found to
| be invariant (i.e. they were the same in all organisms studied).
| However, 58 of the amino acids could be replaced by up to six different
| amino acids, and eight of the amino acids were hypervariable (i.e. they
| could be replaced by more than six amino acids).
|
| Making the naive assumption that this represents the maximum number of
| possible amino acid variants that allows for functional redundancy
| (there are definitely more possibilities than those found in this
| limited study), a conservative calculation can be made giving >10^33
| different possible functionally redundant protein sequences for the 104
| amino acid cytochrome c protein (1^38 x 3^58 x 6^8 > 10^33).
|
| The proof: Man and chimpanzees have the exact same cytochrome c protein
| sequence. The chance of this occurrence is less than 10^-33 (1 out of
| 10^33), unless one assumes genealogical relatedness. Human and
| chimpanzee cytochrome c proteins differ by ~10 amino acids from all
| other mammals. The chance of this occurring, without a hereditary
| mechanism, is conservatively less than 10^-26. Based on this study of
| protein sequences, the most distantly related organism from humans is
| the yeast Candida krusei, with 51 amino acid differences. A
| conservative estimate of this probability is less than 10^-7.

[snip]

| Like protein sequence similarity, the DNA sequence similarity of two
| ubiquitous genes also implies common ancestry. Of course, comprehensive
| DNA sequence comparisons of conserved proteins such as cytochrome c also
| indirectly take into account amino acid sequences, since the DNA
| sequence specifies the protein sequence.
|
| However, with DNA sequences there is an extra level of redundancy. The
| genetic code itself is redundant; on average there are 3 different
| codons (a codon is a triplet of DNA bases) that can specify the exact
| same amino acid.
|
| Thus, for cytochrome c there are 3^104, or over 10^49, different DNA
| sequences (and, hence, 10^49 different possible genes) that can specify
| the very exact same protein sequence.
|
| As mentioned above, the cytochrome c proteins in chimps and humans are
| exactly the same. The clincher is that the two DNA sequences that code
| for cytochrome c in humans and chimps differ by only one codon, even
| though there are over 10^49 different sequences that could code for
| these two proteins.

Because neither selection nor genetic drift can explain that
human and chimp DNA-sequences did not drift apart by random
mutations, neo-Darwinism is definitively dead !!!


Thanks
Wolfgang Gottfried G.

The best alternative after the death of neo-Darwinism:
http://members.lol.li/twostone/E/psychon.html

A Pagano

unread,
Jul 17, 1999, 3:00:00 AM7/17/99
to
Zeus Thibault wrote:
Using a gene like this, there is no reason to assume that the protein
sequence should be the same, unless the two organisms are genealogically
related.

Pagano replies:
EVOLUTIONISTS ASSUME THAT SIMILARITY IS EQUIVALENT TO COMMON DESCENT
Thibault writes as if this is some self-evident scientific truth. In
fact, modern secular theorists have taken for granted a priori that
"similarity" or "relationship" is equivalent to common ancestry.
Thibault has neither logically deduced common descent from the molecular
similarity nor has he or anyone else demonstrated that common descent is
a necessary condition to explain such similarities. The genetic
similarities confirm Linnaeus, not Darwin. The similarities tell us
that apes and humans, for example, are remarkably similar in some ways,
just as they are remarkably different in others. However, such
similarities do not tell us how either the similarities or
dissimilarities came to exist.

NO LIVING TRANSITIONAL INTERMEDIATES AND NONE FOUND IN THE FOSSIL RECORD
Every creator belongs to one and only one phylum, class, and order, and
there are no surviving intermediates. This is true of even the odd
mosaics: the lungfish is a fish, and the duckbilled platypus is a
mammal. NeoDarwinians explain this fundamental and profound
discontinuity of the living world as being due to the fact that all the
intermediate forms that once linked these discrete groups all
conveniently became extinct. Then Gould attempted to explain why the
fossil record (conveniently) never captured any of these transitional
intermediates with his conjecture of punctuated equilibrium. This
discontinuity is mirrored in the molecular evidence.
***************************************


Thibault continues:


This is due to the functional redundancy of protein sequences and
structures. By functional redundancy I mean that many different protein
sequences form the same structure and perform the same function.
Decades of biochemical evidence have shown that most amino acid
mutations, especially of surface residues, have no effect on protein
function or on protein structure.

Pagano replies:
Functional redundancy, the highly accurate error correction mechanism,
and the significantly low selective values (negative or positive) for
the point nucleotide copying errors that do occur all work against the
existence of a real neoDarwinian mechanism capable of creating NEW
information, NEW biological structures, NEW biological systems, and NEW
creatures. If the genome is the receiver of NEW information coding for
revolutionary changes then (1) functional redundancy, (2) error
correction mechanism, and (3) low selective value of mutations all
dramatically attenuate or filter out the noise of random mutations.


[more to follow if time permits]


Regards,
T Pagano


*Hemidactylus*

unread,
Jul 17, 1999, 3:00:00 AM7/17/99
to
In article <3790CE44...@fast.net>,
A Pagano <apa...@fast.net> wrote:

> Zeus Thibault wrote:
> Using a gene like this, there is no reason to assume that the protein
> sequence should be the same, unless the two organisms are genealogically
> related.
>
> Pagano replies:
> EVOLUTIONISTS ASSUME THAT SIMILARITY IS EQUIVALENT TO COMMON DESCENT
>

WRONG. EVOLUTIONARY BIOLOGISTS ARE CAUTIOUS IN THAT THEY TRY TO SEPARATE
HOMOLOGY (SIMILARITY FROM COMMON DESCENT) FROM ANALOGY (SIMILARITY FROM
CONVERGENCE).

>
>
> Thibault writes as if this is some self-evident scientific truth. In
> fact, modern secular theorists have taken for granted a priori that
> "similarity" or "relationship" is equivalent to common ancestry.
> Thibault has neither logically deduced common descent from the molecular
> similarity nor has he or anyone else demonstrated that common descent is
> a necessary condition to explain such similarities. The genetic
> similarities confirm Linnaeus, not Darwin.
>

So homologous genes having various functions across the taxa means nothing?


>
>The similarities tell us
> that apes and humans, for example, are remarkably similar in some ways,
> just as they are remarkably different in others. However, such
> similarities do not tell us how either the similarities or
> dissimilarities came to exist.
>

For morphological homologies: I guess shark jaws and human ear ossicles mean
nothing.

>
> NO LIVING TRANSITIONAL INTERMEDIATES AND NONE FOUND IN THE FOSSIL RECORD
> Every creator belongs to one and only one phylum, class, and order, and
> there are no surviving intermediates. This is true of even the odd
> mosaics: the lungfish is a fish, and the duckbilled platypus is a
> mammal.
>

The platypus still lays eggs and IIRC has some strange thing going on with
its ear ossicles.

I still haven't followed up on Woodger's paradox, IIRC dealing with
transitionals...

--
Scott Chase


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.


Felipe

unread,
Jul 18, 1999, 3:00:00 AM7/18/99
to

alan filipski <al...@netcom.com> wrote in article
<7mql74$q...@dfw-ixnews14.ix.netcom.com>...
> In article <7mqcuv$lhe$1...@pollux.ip-plus.net>, z@z <z...@z.lol.li> wrote:
> -
> -Because neither selection nor genetic drift can explain that
> -human and chimp DNA-sequences did not drift apart by random
> -mutations, neo-Darwinism is definitively dead !!!
>
> Have you ever heard the phrase "purifying selection"? Substitution
> rates are drastically slowed because the alternative genotypes have
> lowered fitness. What is so hard to understand about that?
>

I think "stabilizing selection" is the more usual term; in this case,
"strong stabilizing selection".


Dave Woetzel

unread,
Jul 18, 1999, 3:00:00 AM7/18/99
to

alan filipski wrote in message <7mql74$q...@dfw-ixnews14.ix.netcom.com>...

>In article <7mqcuv$lhe$1...@pollux.ip-plus.net>, z@z <z...@z.lol.li> wrote:
>-
>-Because neither selection nor genetic drift can explain that
>-human and chimp DNA-sequences did not drift apart by random
>-mutations, neo-Darwinism is definitively dead !!!
>
>Have you ever heard the phrase "purifying selection"? Substitution
>rates are drastically slowed because the alternative genotypes have
>lowered fitness. What is so hard to understand about that?


Alan, you snipped the web page. What did you think of...


The best alternative after the death of neo-Darwinism:
http://members.lol.li/twostone/E/psychon.html


Dave


hrgr...@my-deja.com

unread,
Jul 19, 1999, 3:00:00 AM7/19/99
to
In article <0bfdc471319...@csi.com>,
"Dave Woetzel" <dwoe...@juno.com> wrote:
>
> alan filipski wrote in message <7mql74$qt0@dfw-

Neo-Darwinism and Mark Twain have something in common: the news of
their deaths are highly exaggerated! :-)

HRG.

> Dave

hrgr...@my-deja.com

unread,
Jul 19, 1999, 3:00:00 AM7/19/99
to

Chris C.

unread,
Jul 19, 1999, 3:00:00 AM7/19/99
to
> Neo-Darwinism and Mark Twain have something in common: the news of
> their deaths are highly exaggerated! :-)

I thought I would beat the creationists to it by pointing out that Mark
Twain is pretty dead. Am I missing something obvious that makes me look like
an idiot?

hrgr...@my-deja.com

unread,
Jul 19, 1999, 3:00:00 AM7/19/99
to
In article <7mubt8$jum$1...@nntp5.atl.mindspring.net>,
No, but I did. Replace "are" with "were" .....

HRG.

acker james

unread,
Jul 19, 1999, 3:00:00 AM7/19/99
to
A Pagano (apa...@fast.net) wrote:


: NO LIVING TRANSITIONAL INTERMEDIATES AND NONE FOUND IN THE FOSSIL RECORD


: Every creator belongs to one and only one phylum, class, and order, and
: there are no surviving intermediates. This is true of even the odd
: mosaics: the lungfish is a fish, and the duckbilled platypus is a

: mammal. NeoDarwinians explain this fundamental and profound


: discontinuity of the living world as being due to the fact that all the
: intermediate forms that once linked these discrete groups all
: conveniently became extinct. Then Gould attempted to explain why the
: fossil record (conveniently) never captured any of these transitional
: intermediates with his conjecture of punctuated equilibrium. This
: discontinuity is mirrored in the molecular evidence.
: ***************************************


Tony,

This is just a followup (hopefully in a convenient thread)
about something you posted earlier this month. On July 10, you
wrote:

"Though Elsberry has run his "Transitional Fossil Challenge"
incessantly for the last 2+ years he has yet to explain stasis
or present a single unambiguous transitional fossil."

Andrew MacRae and I both wanted to know the necessary
and sufficient characteristics of an unambiguous transitional
fossil. In fact, I'd like to know your criteria so that I can decide
if there really are no "unambiguous" transitional fossils.
If you don't state any criteria, how can anyone judge the validity of
the statement that you make above?

Thanks,

Jim Acker
jac...@gl.umbc.edu


Zeus Thibault

unread,
Jul 19, 1999, 3:00:00 AM7/19/99
to
A Pagano wrote:

> EVOLUTIONISTS ASSUME THAT SIMILARITY IS EQUIVALENT TO COMMON DESCENT

> Thibault writes as if this is some self-evident scientific truth.

Actually, common descent leads to similarity, not vice versa. That is why
you look like your parents, they look like thiers, etc. Coding sequence
similarity is something that was not accepted as self-evident; it was
predicted and confirmed. My point was that the hypothesis of common
descent predicts that chimps and humans should have similar DNA sequences,
and they do despite incredible odds.

Besides common descent, there is no reason why chimps and humans should
have similar DNA sequences in genes coding for proteins like cytochrome c.
Your narrow point (i.e. that similarity does not nessesarily prove common
descent) is valid for morphological structures, since someone designing two
structures for the same function may very well use similar functions.
However, design does not predict the existence of homoplasy (i.e. different
structures performing the same function) or paralogy (different functions
performed by similar structures). But common descent does.

In addition, my whole post was devoted to protein and DNA functional
redundancy. What we find is extreme coding similarity in a context of
extreme functional redundancy. In order to make a hypothesis of
intelligient design, this must be explained. The hypothesis of common
descent has been tested against cytochrome c data, and has been found
consistent in spite of odds greater than 10^82.

Of course, you can allways claim that a designer chose the exact sequences
that mirror common descent. However, this is completely ad hoc unless you
give a reason. The evolutionary reason is that chimps and humans are
genealogically related, just like you and your parents or siblings are.
Now how about design? Do you have an explanation Mr. Pagano?

> Thibault has neither logically deduced common descent from the molecular
> similarity nor has he or anyone else demonstrated that common descent is
> a necessary condition to explain such similarities.

Science is not about deducing explanations from data. Deductions are made
from hypotheses, and these deductions are then checked to see if they
conform with the data. The hypothesis of common descent passes the test.

> NO LIVING TRANSITIONAL INTERMEDIATES AND NONE FOUND IN THE FOSSIL RECORD

Although this statement is rediculously false, it really has nothing to do
with my molecular argument.

> Every creator belongs to one and only one phylum, class, and order, and
> there are no surviving intermediates. This is true of even the odd
> mosaics: the lungfish is a fish, and the duckbilled platypus is a
> mammal. NeoDarwinians explain this fundamental and profound
> discontinuity of the living world as being due to the fact that all the
> intermediate forms that once linked these discrete groups all
> conveniently became extinct.

NeoDarwinism has nothing to do with it. Common descent predicts a family
tree of species (exactly like common descent has produced a family tree in
your genealogy). Thus, there would still be discontinuities even if no
species ever went extinct. Even so, claiming extinctions as further
explanation for discontinuities is not "convenient"; there is independent
proof of extinction today and in the past.

> Functional redundancy, the highly accurate error correction mechanism,
> and the significantly low selective values (negative or positive) for
> the point nucleotide copying errors that do occur all work against the
> existence of a real neoDarwinian mechanism capable of creating NEW
> information, NEW biological structures, NEW biological systems, and NEW
> creatures. If the genome is the receiver of NEW information coding for
> revolutionary changes then (1) functional redundancy, (2) error
> correction mechanism, and (3) low selective value of mutations all
> dramatically attenuate or filter out the noise of random mutations.

All you are saying is that beneficial mutations are rare. Most are either
detrimental or, more likely, neutral. So what? Darwinians have known
this for a long time.

Anyway, the hypothesis of common descent is independent of any explanatory
mechanistic process. Even if NeoDarwinian theory could not explain how it
happened, there is still a great amount of evidence that it DID happen.

Zeus

David E. Kahana

unread,
Jul 20, 1999, 3:00:00 AM7/20/99
to
Zeus Thibault wrote:

Actually, common descent leads to similarity, not vice versa. That is why
you look like your parents, they look like thiers, etc. Coding sequence
similarity is something that was not accepted as self-evident; it was
predicted and confirmed. My point was that the hypothesis of common
descent predicts that chimps and humans should have similar DNA sequences,
and they do despite incredible odds.

[snip]

Of course, you can allways claim that a designer chose the exact sequences
that mirror common descent. However, this is completely ad hoc unless you
give a reason. The evolutionary reason is that chimps and humans are
genealogically related, just like you and your parents or siblings are.
Now how about design? Do you have an explanation Mr. Pagano?

[snip]


I saw your original post on the subject of the identical sequence of
cytochrome c in humans and pan, and was wondering about a seemingly
trivial objection that was raised. Is there an example of a
specific protein for which the exact sequence is known in both humans
and chimpanzees, and where the sequence is different? I looked in the
FAQ, but couldn't find such a case explicitly, though various differences
in the DNA were pointed to. I also tried searching a public protein
database, but since I am not a molecular biologist, the number of proteins
I can name and spell correctly is quite limited. I came up with nothing.

cheers,

- dave k.


z@z

unread,
Jul 21, 1999, 3:00:00 AM7/21/99
to
Hello Zeus Thibault!

[snip]

| My point was that the hypothesis of common
| descent predicts that chimps and humans should have similar DNA sequences,
| and they do despite incredible odds.

But if the odds are too incredible, then some kind of order-creating
principles apart from random mutation and selection must be assumed.

David Ford has shown in http://www.deja.com/=dnc/getdoc.xp?AN=500345458
that such incredible odds are more simply explained by postulating the
existence of a designer. In this context it would be very important
to know the DNA differences of cytochrome-c genes between pigs, cows
and sheep. The cytochrome-c proteins of these three species are
identical whereas the horse protein differs in three amino acids.

Either we have a convergent evolution of cytochrome-c in pigs, cows
and sheep, or we must conclude that pigs, cows and sheep have
conserved the protein of their common ancestor. In the latter case
horses must have changed three amino acids after their separation
from cows. Anyway, both cases are evidence against random mutations.

[snip]

| However, design does not predict the existence of homoplasy (i.e. different
| structures performing the same function) or paralogy (different functions
| performed by similar structures). But common descent does.

Seems not convincing to me. Look at human designers.

| Anyway, the hypothesis of common descent is independent of any explanatory
| mechanistic process. Even if NeoDarwinian theory could not explain how it
| happened, there is still a great amount of evidence that it DID happen.

I agree with you.


Cheer, Wolfgang


The end of (purely materialistic) reductionism:
http://members.lol.li/twostone/E/reductionism.html

Loren King

unread,
Jul 21, 1999, 3:00:00 AM7/21/99
to
Wolfgang <z...@z.lol.li>:

> But if the odds are too incredible, then some kind of order-creating
> principles apart from random mutation and selection must be assumed.

No, they would be hypothesized, and then work would proceed to test
the hypothesis.


> David Ford has shown in http://www.deja.com/=dnc/getdoc.xp?AN=500345458
> that such incredible odds are more simply explained by postulating the
> existence of a designer.

He has shown no such thing. Nor, for that matter, have Behe or Dembski,
and these two scholars have provided perhaps the most serious attempts
to support the design inference as applied to biological systems. But
the inference fails, for reasons that have been detailed countless
times, but that you and a few others keep ignoring ... what gives? Are
you just not reading the stuff that gets posted in response to your
often-bizarre assertions?

L.

--------------------------------------
Loren King lk...@mit.edu
http://web.mit.edu/lking/www/home.html


Abiel...@aol.com

unread,
Jul 21, 1999, 3:00:00 AM7/21/99
to
In article <7mqcuv$lhe$1...@pollux.ip-plus.net>,
"z@z" <z...@z.lol.li> wrote:
> Hello Zeus Thibault!
>
> What you write is excellent and constitutes maybe the best refutation
> of neo-Darwinism I've ever seen. For further details look at the
> current thread "ENIGMAS --- Innovations vs. Mutations".
> http://www.deja.com/=dnc/getdoc.xp?AN=499894064 (starting post)

This is how I saw it too as I read it.

In traffic court people plead before the judge to get out of
their ticket. But most commonly they end up convicting themselves
by the very words they believe will get them out of their ticket.
This posting is a similar example.

Adam Noel Harris

unread,
Jul 21, 1999, 3:00:00 AM7/21/99
to
Abiel...@aol.com <Abiel...@aol.com> wrote:
:In article <7mqcuv$lhe$1...@pollux.ip-plus.net>,

: "z@z" <z...@z.lol.li> wrote:
:> Hello Zeus Thibault!
:>
:> What you write is excellent and constitutes maybe the best refutation
:> of neo-Darwinism I've ever seen. For further details look at the
:> current thread "ENIGMAS --- Innovations vs. Mutations".
:> http://www.deja.com/=dnc/getdoc.xp?AN=499894064 (starting post)
:
:This is how I saw it too as I read it.
:
:In traffic court people plead before the judge to get out of
:their ticket. But most commonly they end up convicting themselves
:by the very words they believe will get them out of their ticket.
:This posting is a similar example.

What are you two croning about now? Perhaps you'd like to elaborate.

-Adam
--
Opinions expressed are not necessarily those of Stanford University.
PGP Fingerprint = C0 65 A2 BD 8A 67 B3 19 F9 8B C1 4C 8E F2 EA 0E


Zeus Thibault

unread,
Jul 21, 1999, 3:00:00 AM7/21/99
to
Hello z@z,

"z@z" wrote:

> But if the odds are too incredible, then some kind of order-creating
> principles apart from random mutation and selection must be assumed.

When considering your cyt c sequence and your mother's, the exact same
incredible odds are involved - yet we don't "assume" any order-creating
principles, aside from heredity.

Cheers,

Zeus

Zeus Thibault

unread,
Jul 21, 1999, 3:00:00 AM7/21/99
to
Abiel...@aol.com wrote:

> "z@z" <z...@z.lol.li> wrote:
> > Hello Zeus Thibault!
> >
> > What you write is excellent and constitutes maybe the best refutation

> > of neo-Darwinism I've ever seen. This is how I saw it too as I read


> it.
>
> In traffic court people plead before the judge to get out of
> their ticket. But most commonly they end up convicting themselves
> by the very words they believe will get them out of their ticket.
> This posting is a similar example.

If you read my post, you will recall that I submitted a scientific proof
of common descent, not evidence for NeoDarwinism. These are two seperable
independent hypotheses.

Regardless, my post does not refute NeoDarwinism. It does, however, give
some convincing support for the neutral theory of molecular evolution.
Darwinists have long known that beneficial mutations are very rare. The
evidence which I listed, and the neutral theory in general, only show that
many of the "unbeneficial" mutations are not detrimental, but neutral.
And this is only on the molecular level anyway; by definition coding
functional redundancy does not affect phenotypes.

Cheers,

Zeus


David Johnston

unread,
Jul 21, 1999, 3:00:00 AM7/21/99
to
z@z wrote:
>
> Hello Zeus Thibault!
>
> [snip]
>
> | My point was that the hypothesis of common
> | descent predicts that chimps and humans should have similar DNA sequences,
> | and they do despite incredible odds.
>
> But if the odds are too incredible, then some kind of order-creating
> principles apart from random mutation and selection must be assumed.

Was that deliberate misinterpretation, or did you just not get his
meaning?

Abiel...@aol.com

unread,
Jul 22, 1999, 3:00:00 AM7/22/99
to
In article <379636EE...@hotmail.com>,

Zeus Thibault <zthi...@hotmail.com> wrote:
> Abiel...@aol.com wrote:
>
> > "z@z" <z...@z.lol.li> wrote:
> > > Hello Zeus Thibault!
> > >
> > > What you write is excellent and constitutes maybe the best
refutation
> > > of neo-Darwinism I've ever seen. This is how I saw it too as I
read
> > it.
> >
> > In traffic court people plead before the judge to get out of
> > their ticket. But most commonly they end up convicting themselves
> > by the very words they believe will get them out of their ticket.
> > This posting is a similar example.
>
> If you read my post, you will recall that I submitted a scientific
proof
> of common descent, not evidence for NeoDarwinism. These are two
seperable
> independent hypotheses.

This is my mistake, I did not realize you had made such
a distinction. However, I conclude from the facts you present
that there is one Creator Designer who used the same patterns
throughout creation. And also, since there are so few mutations
at the amino sites that, based on the mutation rates currently
used by evolutionists, not very much time could have passed since
the inception of life itself.

>
> Regardless, my post does not refute NeoDarwinism. It does, however,
give
> some convincing support for the neutral theory of molecular evolution.
> Darwinists have long known that beneficial mutations are very rare.
The
> evidence which I listed, and the neutral theory in general, only show
that
> many of the "unbeneficial" mutations are not detrimental, but neutral.
> And this is only on the molecular level anyway; by definition coding
> functional redundancy does not affect phenotypes.


Ah, but evolutionists right here on t.O. have pointed out that
nuetral mutatons are so numerous, and after, that is eaxactly what
you are pointing out, neutral amino acid mutations which do not
change or inhibit the function of the protein.

Anyway, nice to meet you Zeus and I do appreciate your post.
Thanks.

>
> Cheers,
>
> Zeus

Zeus Thibault

unread,
Jul 23, 1999, 3:00:00 AM7/23/99
to
"David E. Kahana" wrote:

> I saw your original post on the subject of the identical sequence of
> cytochrome c in humans and pan, and was wondering about a seemingly
> trivial objection that was raised. Is there an example of a
> specific protein for which the exact sequence is known in both humans
> and chimpanzees, and where the sequence is different? I looked in the
> FAQ, but couldn't find such a case explicitly, though various differences
> in the DNA were pointed to. I also tried searching a public protein
> database, but since I am not a molecular biologist, the number of proteins
> I can name and spell correctly is quite limited. I came up with nothing.

Dear Dave,

I haven't done a very thorough search, but from about ten random pan protiens I
selected, each was identical in amino acid sequence to the human ortholog.
What exactly was the objection you refer to? I think the fact that most pan
and human proteins are identical only strengthens my argument.

To help you in your search, you may want to try the NCBI Entrez server, which
gives you access to all known protein and DNA sequences. The protein search
engine is located here:

http://www.ncbi.nlm.nih.gov/Entrez/protein.html

At this site, you can select "organism" under "search field", search for
chimpanzee (or whatever organism you wish), and you will get a huge list of all
known chimp proteins, with links to their sequences.

You can then proceed to the BLAST server, which finds homologous proteins for
an input sequence. It is found at NCBI too:

http://www.ncbi.nlm.nih.gov/blast/blast.cgi?Jform=0

Enter the protein sequence you want to search (cut it from the files you found
above and paste it in). For proteins, select "blastp" in the "Program" field,
and hit search. Just leave the database at nr (for non-redundant). In a few
seconds you will get back a long list of homologous protiens, with thier
sequence alignments and stats on how similar they are. It will also give you
links to these sequences.

Hope this helps,

Zeus


Zeus Thibault

unread,
Jul 24, 1999, 3:00:00 AM7/24/99
to
"David E. Kahana" wrote:

> The objection seemed bizarre to me and stated the following:
>
> :Message-ID <7mqcuv$lhe$1...@pollux.ip-plus.net>
> :
> :Because neither selection nor genetic drift can explain that
> :human and chimp DNA-sequences did not drift apart by random
> :mutations, neo-Darwinism is definitively dead !!!
> :
> :Thanks
> :Wolfgang Gottfried G.
> :
> :The best alternative after the death of neo-Darwinism:
> :http://members.lol.li/twostone/E/psychon.html
> :
> [snip]

> But being obsessive, I wondered whether there was some protein that could be
> pointed to which is different. One imagines there should be some, at least.
> Some must be more variable than others.

Dave,

Here's an example - the interleukin-3 precursor (IL-3) protein. It differs between
humans and pan by 2 out of 152 amino acids. One would predict it to have more
sequence differences than most other proteins. That's how I found this example - I
know that proteins involved in the immune system are variable in large populations
since selective pressures for specific immune functions can vary wildly from region
to region (c.f. the various human immunity related factors that we must check for
before tissue transplants).

Cheers,

Zeus


z@z

unread,
Jul 25, 1999, 3:00:00 AM7/25/99
to
Hello Zeus Thibault!

| > But if the odds are too incredible, then some kind of order-creating
| > principles apart from random mutation and selection must be assumed.
|

| When considering your cyt c sequence and your mother's, the exact same
| incredible odds are involved - yet we don't "assume" any order-creating
| principles, aside from heredity.

Over one generation it is certainly not astonishing that DNA
sequences remain unchanged. But if sequences do not change over
100'000, 1'000'000, 10'000'000, 100'000'000 or even more
generations, your objection becomes unfounded.

Neither you nor Howard has yet replied to this:

"| As mentioned above, the cytochrome c proteins in chimps and humans are
"| exactly the same. The clincher is that the two DNA sequences that code
"| for cytochrome c in humans and chimps differ by only one codon, even
"| though there are over 10^49 different sequences that could code for
"| these two proteins.
"

" Because neither selection nor genetic drift can explain that
" human and chimp DNA-sequences did not drift apart by random
" mutations, neo-Darwinism is definitively dead !!!

It is a logical consequence of neo-Darwinism that all possible
DNA sequences coding for a given amino acid sequence are
equivalent. Therefore this theory predicts the existence of
lots of different DNA sequences for a given amino acid
sequence (and in addition to that also the existence of lots
of more or less functionally equivalent alleles for all gene
loci).


Cheers, Wolfgang
http://members.lol.li/twostone/links.html


PS: Cows are more related to pigs and sheep (even-toed ungulates),
and not to horses (odd-toed ungulates) as I erroneously assumed
in my previous post. Therefore the identity of the amio acid
sequence of cytochrome-c in cows, pigs and sheep can be explained
by sequence conservation over millions of generations and it is
not necessary to assume convergence.

Tim Tyler

unread,
Jul 26, 1999, 3:00:00 AM7/26/99
to
[snip the guts of the discussion]

In talk.origins z@z <z...@z.lol.li> wrote:

: It is a logical consequence of neo-Darwinism that all possible


: DNA sequences coding for a given amino acid sequence are

: equivalent. [...]

Well, non-identical sequences can behave differently from one another
under mutation and crossover - so they are not equivalent in this sense.

Different base pair sequences may even differ phenotypically - if a virus
uses something like a restriction enzyme to splice itself into the host
DNA, then the exact base pair sequence can make a difference to the host's
susceptability to disease (for example).
--
__________
|im |yler The Mandala Centre http://www.mandala.co.uk/ t...@cryogen.com

Dogs come when you call. Cats have answering machines.


howard hershey

unread,
Jul 26, 1999, 3:00:00 AM7/26/99
to
z@z wrote:
>
> Hello Zeus Thibault!
>
> What you write is excellent and constitutes maybe the best refutation
> of neo-Darwinism I've ever seen. For further details look at the
> current thread "ENIGMAS --- Innovations vs. Mutations".
> http://www.deja.com/=dnc/getdoc.xp?AN=499894064 (starting post)
>
> | This is a reworking of an old argument in a way that I have never seen
> | before; I hope that it will clarify exactly why molecular sequence
> | similarity is such a powerful technique for demonstrating the
> | genealogical relatedness of different species.
>
> [snip]
>
> | A striking example is that of the c-type cytochromes from various
> | bacteria, which have no sequence similarity. Nevertheless, they all
> | fold into the same three-dimensional structure, and they all perform the
> | same function.

You are aware that stating flatly that the cytochromeC's of various
bacteria do not *all* "have no sequence similarity" is somewhat
misleading. Rather what one observes is a nested branching series of
sequence similarities in the eubacteria (which, after all, have a deep
lineage with the last common ancestor of the currently known eubacteria
placed about 1.5 billion ybp). That is, some bacterial species have
quite similar cytochromeC sequences (and, shock of shocks, these are the
same species that also have quite similar rRNA sequences indicating
recent divergence) and other groups of bacteria have a more divergent
sequence (which, shock of shock, correlates well with the divergences in
rRNA - and other sequences). That all the cytochome C's fold into the
same 3-D structure and perform the same function is also hardly
surprising from an evolutionary perspective.

Among the 100 studied. I presume that the gorilla was not included.

> | The chance of this occurring, without a hereditary
> | mechanism, is conservatively less than 10^-26. Based on this study of
> | protein sequences, the most distantly related organism from humans is
> | the yeast Candida krusei, with 51 amino acid differences.

Among the 100 studied. Did the study include bacteria? Or did it only
include eucaryotes? Did it include trypanosomes and ciliates? Was this
a crude number count of was a path analysis done? Once you start
getting re-ratting (sequences sites that mutate back), simple number
counts are less useful.

> | A
> | conservative estimate of this probability is less than 10^-7.
>
> [snip]
>
> | Like protein sequence similarity, the DNA sequence similarity of two
> | ubiquitous genes also implies common ancestry. Of course, comprehensive
> | DNA sequence comparisons of conserved proteins such as cytochrome c also
> | indirectly take into account amino acid sequences, since the DNA
> | sequence specifies the protein sequence.
> |
> | However, with DNA sequences there is an extra level of redundancy. The
> | genetic code itself is redundant; on average there are 3 different
> | codons (a codon is a triplet of DNA bases) that can specify the exact
> | same amino acid.
> |
> | Thus, for cytochrome c there are 3^104, or over 10^49, different DNA
> | sequences (and, hence, 10^49 different possible genes) that can specify
> | the very exact same protein sequence.
> |

> | As mentioned above, the cytochrome c proteins in chimps and humans are
> | exactly the same. The clincher is that the two DNA sequences that code
> | for cytochrome c in humans and chimps differ by only one codon, even
> | though there are over 10^49 different sequences that could code for
> | these two proteins.
>
> Because neither selection nor genetic drift can explain that
> human and chimp DNA-sequences did not drift apart by random
> mutations, neo-Darwinism is definitively dead !!!

This is a really strange conclusion from the above evidence. The
conclusion is, in fact, contrary to the expectations of a randomly
chosen gene that is not crucial to the speciation event (such as
cytochrome C).

*Selection* would lead to a difference between human and chimp *only* if
there were some necessary-to-speciation requirement for a different
sequence in humans and chimps *for the cytochrome C gene product*.
Otherwise, the effect of selection would be conservative. I cannot
think of a reason why there would be *selection* for a *different*
cytochrome C in humans and chimps.

That leaves changes that occur that are related to time alone, namely
selectively neutral changes. Cytochrome C is a relatively small
molecule with strong selective constraint compared to larger molecules.
Thus the rate of neutral change in cytochrome C is slow (relative to,
say, hemoglobin). That, in fact, is why cytochrome C is so useful in
analysing neutral changes over very long periods of time (and why
fibrinogen peptide is more useful over shorter periods of time in
vertebrate lineages). Given the time since divergence of humans and
chimps (say, 10 million years), the absence of any neutral changes in
cytochrome C bween humans and chimps is, in fact, to be expected. One
*expects*, from a plot of changes in cytochrome C vrs millions of years
since divergence, that there will be, on average, about 30 changes per
100 amino acids in two organisms that diverged 600 million years ago and
cytochrome C is about 112 aa long. Thus, roughly speaking, one would
*expect*, in a 10 million year divergence, there to be about 0.5 changes
due to neutral drift. This is, need I point out, somewhat less than 1.

Zeus Thibault

unread,
Jul 26, 1999, 3:00:00 AM7/26/99
to
Dear Wolfgang,

"z@z" wrote:

> Hello Zeus Thibault!
>


> | > But if the odds are too incredible, then some kind of order-creating
> | > principles apart from random mutation and selection must be assumed.
> |
> | When considering your cyt c sequence and your mother's, the exact same
> | incredible odds are involved - yet we don't "assume" any order-creating
> | principles, aside from heredity.

> Over one generation it is certainly not astonishing that DNA
> sequences remain unchanged. But if sequences do not change over
> 100'000, 1'000'000, 10'000'000, 100'000'000 or even more
> generations, your objection becomes unfounded.

From paleontological evidence, chimps and humans diverged approx. 5 million
years ago. Given a generation time of ~15 years (probably and
underestimate), you only get about 300,000 generations.

> Neither you nor Howard has yet replied to this:

> [Zeus said -]


>
> "| As mentioned above, the cytochrome c proteins in chimps and humans are
> "| exactly the same. The clincher is that the two DNA sequences that code
> "| for cytochrome c in humans and chimps differ by only one codon, even
> "| though there are over 10^49 different sequences that could code for
> "| these two proteins.

[Wolfgang replied]

> " Because neither selection nor genetic drift can explain that
> " human and chimp DNA-sequences did not drift apart by random
> " mutations, neo-Darwinism is definitively dead !!!

> It is a logical consequence of neo-Darwinism that all possible


> DNA sequences coding for a given amino acid sequence are

> equivalent. Therefore this theory predicts the existence of
> lots of different DNA sequences for a given amino acid
> sequence (and in addition to that also the existence of lots
> of more or less functionally equivalent alleles for all gene
> loci).
>
> Cheers, Wolfgang

Both Howard Hershey and Tim Tyler now have given partial rebuttals to this
objection (please see their recent posts).

To add to what they have said:

I completely agree with Tim when he states that "non-identical sequences can


behave differently from one another under mutation and crossover - so they

are not equivalent in this sense ... Different base pair sequences may even
differ phenotypically." Both eukaryotes (including mammals) and prokaryotes
are known to display codon bias, where one or more of the redundant codons
are prefered over the others. Some of this preference is known to be due to
selection. One may think that this invalidates my argument, since it
suggests that not all equivalent DNA coding sequences are truly functionally
redundant, and that some are less fit than others.

I showed that >10^49 DNA sequences can code for the exact same cytochrome c
protein sequence in humans and chimps. If we further assume that 99.99 % of
these sequences are actually less fit, due to such factors as suboptimal
codon usage, there are still >10^45 genuinely functionally and phenotypically
equivalent DNA coding sequences that could have been used to specify this
cytochrome c protein. Obviously, my argument still holds even with this
further assumption (which is more realistic). The huge numbers make it
rather robust.

However, if 99.99 % of these DNA sequences are less fit, then they would be
selected against, and the variation after 300,000 generations could be very,
very low. This is due to the fact that the observed codon substition rate
would be four orders of magnitude less than the spontaneous mutation rate of
"conserved" codon substitutions.

Nevertheless, current estimates give about 200 base changes per generation as
the spontaneous mutation rate in humans (which is really surprisingly high).
300,000 generations of 200 base changes per generation gives a total of 60
million DNA base mutations. Spread equally over the 6 billion bases in the
human genome this gives one change per hundred bases since the chimp/human
divergence. From this naive calculation we would expect about three codon
differences in the ~300 bp cytochrome c gene, which is not far off at all
from the observed difference of one codon.

Thus, NeoDarwinism and common ancestry account rather well for the observed
DNA sequence similarity of the human and pan cytochrome c gene, both
quantitatively and qualitatively, whether you naively assume strict
functional coding redundancy (as I did for simplicity) or whether you
incorporate stringent selection on DNA sequence.

Cheers,

Zeus

Douglas Theobald

unread,
Jul 26, 1999, 3:00:00 AM7/26/99
to
Hello Howard,

howard hershey wrote:

> | A striking example is that of the c-type cytochromes from various
> | bacteria, which have no sequence similarity.  Nevertheless, they all
> | fold into the same three-dimensional structure, and they all perform the
> | same function.

You are aware that stating flatly that the cytochromeC's of various
bacteria do not *all* "have no sequence similarity" is somewhat
misleading.

[snip]

That all the cytochome C's fold into the
same 3-D structure and perform the same function is also hardly
surprising from an evolutionary perspective.

You are right here; I did not intend to state that *all* bacterial cytochrome c sequences are dissimilar.  It is true that selected ("various") ones have no detectable sequence homology with each other.  The existence of these structurally similar cyt  Cs, lacking sequence homology, still reinforces my point of functional redundancy of protein sequence.

> | The chance of this occurring, without a hereditary
> | mechanism, is conservatively less than 10^-26.  Based on this study of
> | protein sequences, the most distantly related organism from humans is
> | the yeast Candida krusei, with 51 amino acid differences.

Among the 100 studied.  Did the study include bacteria?  Or did it only
include eucaryotes?  Did it include trypanosomes and ciliates?  Was this
a crude number count of was a path analysis done?  Once you start
getting re-ratting (sequences sites that mutate back), simple number
counts are less useful.

The study I refer to did not include bacteria cyt c, but only eukaryotic cyt c from yeast, plants, and animals.  My treatment does not include any path analysis, but this does not really apply to my argument.  I gave a minimum estimate of possible functional redundancy, which is independent of evolutionary path.

[Wolfgang (z@z) wrote]

> Because neither selection nor genetic drift can explain that
> human and chimp DNA-sequences did not drift apart by random
> mutations, neo-Darwinism is definitively dead !!!

[Howard replied]

*Selection* would lead to a difference between human and chimp *only* if
there were some necessary-to-speciation requirement for a different
sequence in humans and chimps *for the cytochrome C gene product*.
Otherwise, the effect of selection would be conservative.  I cannot
think of a reason why there would be *selection* for a *different*
cytochrome C in humans and chimps.

Nevertheless, Wolfgang's objection here is somewhat valid, if mutation rates are extremely high, and if there is complete functional redundancy in *all* the possible conservative DNA coding and protein sequences that I calculated.  If these two points were valid, then phenotypically silent mutations (immune to selection) would be churning away behind the cyt c gene product.  I address this trivial complication in another post,  message ID

news://news/379CFEBC.2B08756C%40hotmail.com

Cheers,

Zeus

Brian Hartman

unread,
Jul 27, 1999, 3:00:00 AM7/27/99
to
"z@z" wrote:

> Hello Zeus Thibault!
>


> [snip]
>
> | My point was that the hypothesis of common
> | descent predicts that chimps and humans should have similar DNA sequences,
> | and they do despite incredible odds.
>

> But if the odds are too incredible, then some kind of order-creating
> principles apart from random mutation and selection must be assumed.

Maybe I'm just missing something, but why are the odds that chimps and humans
have similar DNA sequences so "incredible"? They *ought* to be similar,
according to evolution. For that matter, even *yeast* has a similar dna
structure to human DNA. (Naturally, it varies from human DNA more than chimp
DNA does.) This is precisely in keeping with evolutionary theory.

>
> David Ford has shown in http://www.deja.com/=dnc/getdoc.xp?AN=500345458
> that such incredible odds are more simply explained by postulating the

> existence of a designer. In this context it would be very important
> to know the DNA differences of cytochrome-c genes between pigs, cows
> and sheep. The cytochrome-c proteins of these three species are
> identical whereas the horse protein differs in three amino acids.
>
> Either we have a convergent evolution of cytochrome-c in pigs, cows
> and sheep, or we must conclude that pigs, cows and sheep have
> conserved the protein of their common ancestor. In the latter case
> horses must have changed three amino acids after their separation
> from cows. Anyway, both cases are evidence against random mutations.
>

Exactly how does this follow?

>
> [snip]
>
> | However, design does not predict the existence of homoplasy (i.e. different
> | structures performing the same function) or paralogy (different functions
> | performed by similar structures). But common descent does.
>
> Seems not convincing to me. Look at human designers.
>

It would make sense for different designers (not having knowledge of each
other's work) to create different structures performing the same function. Or
even different functions for the same structure. But it would not make any
sense for *one* intelligence to work that way. One would think God would find
the best structure for a function and stick with it. After all, he *knows* the
best structure, right?

stu...@my-deja.com

unread,
Jul 27, 1999, 3:00:00 AM7/27/99
to
In article <378FE47E...@hotmail.com>,

Zeus Thibault <zthi...@hotmail.com> wrote:
> This is a reworking of an old argument in a way that I have never seen
> before; I hope that it will clarify exactly why molecular sequence
> similarity is such a powerful technique for demonstrating the
> genealogical relatedness of different species.
>
> I. Protein functional redundancy.
>
> Before the advent of DNA sequencing technology, the amino acid
sequences
> of proteins were used to establish the phylogenetic relationships of
> species. Sequence studies with functional genes have centered on
genes
> of proteins (or RNAs) that are ubiquitous (i.e. all organisms have
> them). This is done to insure that the comparisons are independent of
> the overall species phenotype.
>
> For example, suppose we are comparing the protein sequence of a
> chimpanzee and that of a human. Both of these animals have many
similar
> anatomical characters, so we might expect their proteins to be
similar,
> regardless of whether they are genealogically related or not.
However,
> we can compare the sequences of very basic genes that are used by all
> living organisms, such as the cytochrome c gene, which have no
influence
> over specific chimpanzee or human characteristics.
>
> Cytochrome c is a ubiquitous protein found in all eukaryotes, which is
> absolutely required for the viability of all organisms (bacteria, or
> prokaryotes, also have cytochrome c proteins, but for historical
reasons
> they are called c-like cytochromes). Cytochrome c is found in the
> mitochondria of eukaryotic cells, where it is a key player in the
> fundamental metabolic process of oxidative phosphorylation (this is
> where the O2 we breathe is used to generate energy).
>
> Using a gene like this, there is no reason to assume that the protein
> sequence should be the same, unless the two organisms are
genealogically
> related. This is due to the functional redundancy of protein
sequences
> and structures. By functional redundancy I mean that many different
> protein sequences form the same structure and perform the same
> function. Decades of biochemical evidence have shown that most amino
> acid mutations, especially of surface residues, have no effect on
> protein function or on protein structure.

>
> A striking example is that of the c-type cytochromes from various
> bacteria, which have no sequence similarity. Nevertheless, they all
> fold into the same three-dimensional structure, and they all perform
the
> same function.
>
> Even within species, most amino acid mutations are functionally
silent.
> For example, there are at least 250 different amino acid mutations
known
> in human hemoglobin, carried by more than 3% of the world's
population,
> that have no clinical manifestation in either heterozygotic or
> homozygotic individuals. This is a general phenomenon observed in all
> proteins and genes in all species.
>
> However, the most convincing evidence of protein functional redundancy
> comes from molecular biology experiments with recombinant organisms.
> For example, it is possible to delete the essential cytochrome c gene
in
> one species and replace it with the gene from another species. This
has
> been done with species as diverse as horse, wheat, pigeon, mouse,
yeast,
> and human. In all cases the proteins were biologically functional and
> the recombinant organisms were viable.
>

Could you point me to the study done where a human had their cytochrome
c gene replaced with that of another species? Count me as a skeptic.


> Consider again the molecular sequences of cytochrome c. In a classic
> study of ~100 cytochrome c protein sequences from organisms ranging
from
> yeast to human, 38 of the 104 amino acids in the protein were found to
> be invariant (i.e. they were the same in all organisms studied).
> However, 58 of the amino acids could be replaced by up to six
different
> amino acids, and eight of the amino acids were hypervariable (i.e.
they
> could be replaced by more than six amino acids).
>
> Making the naive assumption that this represents the maximum number of
> possible amino acid variants that allows for functional redundancy
> (there are definitely more possibilities than those found in this
> limited study), a conservative calculation can be made giving >10^33
> different possible functionally redundant protein sequences for the
104
> amino acid cytochrome c protein (1^38 x 3^58 x 6^8 > 10^33).
>

> The proof: Man and chimpanzees have the exact same cytochrome c
protein


> sequence. The chance of this occurrence is less than 10^-33 (1 out of
> 10^33), unless one assumes genealogical relatedness. Human and
> chimpanzee cytochrome c proteins differ by ~10 amino acids from all

> other mammals. The chance of this occurring, without a hereditary


> mechanism, is conservatively less than 10^-26. Based on this study of
> protein sequences, the most distantly related organism from humans is

> the yeast Candida krusei, with 51 amino acid differences. A


> conservative estimate of this probability is less than 10^-7.
>

> What makes these figures even more convincing is the fact that the
> phylogenetic tree constructed from the cytochrome c data exactly
> recapitulates the relationships of major classes as determined by the
> completely independent morphological data.
>
> The most likely result is that all these protein sequences would be
very
> different from each other. If this were the case, a phylogenetic
> analysis would be impossible, and this would provide very strong
> evidence for a genealogically unrelated, perhaps simultaneous, origin
of
> species.
>
> II. DNA coding redundancy.


>
> Like protein sequence similarity, the DNA sequence similarity of two
> ubiquitous genes also implies common ancestry. Of course,
comprehensive
> DNA sequence comparisons of conserved proteins such as cytochrome c
also
> indirectly take into account amino acid sequences, since the DNA
> sequence specifies the protein sequence.
>
> However, with DNA sequences there is an extra level of redundancy.
The
> genetic code itself is redundant; on average there are 3 different
> codons (a codon is a triplet of DNA bases) that can specify the exact
> same amino acid.
>
> Thus, for cytochrome c there are 3^104, or over 10^49, different DNA
> sequences (and, hence, 10^49 different possible genes) that can
specify
> the very exact same protein sequence.
>

> As mentioned above, the cytochrome c proteins in chimps and humans are
> exactly the same. The clincher is that the two DNA sequences that
code
> for cytochrome c in humans and chimps differ by only one codon, even
> though there are over 10^49 different sequences that could code for
> these two proteins.
>

> Thus, there are really just two choices. Either (1) from probability
> considerations, we are 100% sure that humans and chimps are closely
> genealogically related, or (2) a designer chose the two DNA sequences
> out of the over 10^49 possibilities that make it look exactly like we
> are genealogically related.
>
> The combined effects of DNA coding redundancy and protein sequence
> redundancy makes DNA sequence comparisons doubly redundant (actually
> multiplicatively redundant); DNA sequences of ubiquitous proteins are
> completely uncorrelated with phenotype, but are strongly correlated
with
> heredity. This is why DNA sequence phylogenies are considered so
> robust.
>
> The most probable result is that the DNA sequences coding for these
> proteins should be radically different. This would be a resounding
> falsification of macroevolution, it would be very strong proof that
> chimpanzees and humans are not closely genealogically related, and
would
> be pretty convincing evidence of design. Since the most probable
result
> is not found, and what is observed is exactly what is expected due to
> common descent of these two organisms, any theory of design must
explain
> this fact as well.
>
> In other words, exactly why did a designer choose the two sequences
that
> look exactly like these organisms evolved, and not one of the other
> 10^82 (i.e. 10^49 x 10^33) functionally equivalent sequences?

Michael Konstantine Kalandros

unread,
Jul 27, 1999, 3:00:00 AM7/27/99
to
stu...@my-deja.com wrote:

>
> Could you point me to the study done where a human had their cytochrome
> c gene replaced with that of another species? Count me as a skeptic.

Yeah right. People get all upset about cloning, and you want to insert
other
animals' genes?

How about the other way? I found a short article where a human gene
is but in a very primitive worm:

http://news.bbc.co.uk/hi/english/sci/tech/newsid_173000/173295.stm


Mike

Jeff

unread,
Jul 27, 1999, 3:00:00 AM7/27/99
to
]>stu...@my-deja.com wrote:
>
>>
]>> Could you point me to the study done where a human had their

cytochrome
]>> c gene replaced with that of another species? Count me as a
skeptic.
>
]>Yeah right. People get all upset about cloning, and you want to

insert other animals' genes?
>
]>How about the other way? I found a short article where a human
gene is but in a very primitive worm:
>
]>http://news.bbc.co.uk/hi/english/sci/tech/newsid_173000/173295.stm
]>Mike

Quite often transgenics of mice are made with the human gene homologue
to allow detection of the transgene product.

Although not just genes there have been humans that received entire
animal organs which functioned.


Zeus Thibault

unread,
Jul 27, 1999, 3:00:00 AM7/27/99
to

stu...@my-deja.com wrote:

> In article <378FE47E...@hotmail.com>,
> Zeus Thibault <zthi...@hotmail.com> wrote:

> > However, the most convincing evidence of protein functional redundancy
> > comes from molecular biology experiments with recombinant organisms.
> > For example, it is possible to delete the essential cytochrome c gene
> in
> > one species and replace it with the gene from another species. This
> has
> > been done with species as diverse as horse, wheat, pigeon, mouse,
> yeast,
> > and human. In all cases the proteins were biologically functional and
> > the recombinant organisms were viable.
> >
>
> Could you point me to the study done where a human had their cytochrome
> c gene replaced with that of another species? Count me as a skeptic.

Come on, now, I didn't say that every possible permutation of gene
replacement had been done. Noone has replaced human genes with genes from
other animals yet, but the other way round has been done repeatedly.

Cheers,

Zeus

z@z

unread,
Jul 28, 1999, 3:00:00 AM7/28/99
to
Hello Brian Hartman!

| > | = Zeus Thibault
| > = me

| > | My point was that the hypothesis of common
| > | descent predicts that chimps and humans should have similar DNA sequences,
| > | and they do despite incredible odds.
| >
| > But if the odds are too incredible, then some kind of order-creating
| > principles apart from random mutation and selection must be assumed.
|
| Maybe I'm just missing something, but why are the odds that chimps and humans
| have similar DNA sequences so "incredible"? They *ought* to be similar,
| according to evolution. For that matter, even *yeast* has a similar dna
| structure to human DNA. (Naturally, it varies from human DNA more than chimp
| DNA does.) This is precisely in keeping with evolutionary theory.

There have been a lot of innovations in humans after their separation
from chimps several hundered thousand generations ago. If all what makes
us different from chimps depends on random mutations and selection then
mutations must occur with a certain frequency. But if not even the
many possible neutral mutations leaving amino acid chains unchanged
occur then it becomes highly improbable that mutations are the primary
cause of evolution or at least that mutations are random.

If you are really interested in this crucial question, you should
read the following discussion:

http://www.deja.com/=dnc/getdoc.xp?AN=501004397 (me)
http://www.deja.com/=dnc/getdoc.xp?AN=501873642 (Howard Hershey)
http://www.deja.com/=dnc/getdoc.xp?AN=502417358 (me)
http://www.deja.com/=dnc/getdoc.xp?AN=502836392 (Howard)
http://www.deja.com/=dnc/getdoc.xp?AN=503575173 (me)

In this context Richard Harter's arguments against the modern
synthesis are also relevant:

http://www.deja.com/=dnc/getdoc.xp?AN=505129970

| > David Ford has shown in http://www.deja.com/=dnc/getdoc.xp?AN=500345458
| > that such incredible odds are more simply explained by postulating the
| > existence of a designer. In this context it would be very important
| > to know the DNA differences of cytochrome-c genes between pigs, cows
| > and sheep. The cytochrome-c proteins of these three species are
| > identical whereas the horse protein differs in three amino acids.
| >
| > Either we have a convergent evolution of cytochrome-c in pigs, cows
| > and sheep, or we must conclude that pigs, cows and sheep have
| > conserved the protein of their common ancestor. In the latter case
| > horses must have changed three amino acids after their separation
| > from cows. Anyway, both cases are evidence against random mutations.
|
| Exactly how does this follow?

In my last post of this thread I have already admitted that in this
case the dichotomy between convergence and conservation is based on
an erroneous assumption.

"PS: Cows are more related to pigs and sheep (even-toed ungulates),
and not to horses (odd-toed ungulates) as I erroneously assumed
in my previous post. Therefore the identity of the amio acid
sequence of cytochrome-c in cows, pigs and sheep can be explained
by sequence conservation over millions of generations and it is
not necessary to assume convergence."

Nevertheless I assume that such convergent evolutions of DNA or
amino acid sequences have occured during evolution in a similar
way as convergence between marsupials and placentals has occured.

[snip]

Cheers, Wolfgang

Zeus Thibault

unread,
Jul 28, 1999, 3:00:00 AM7/28/99
to
Dear Wolfgang,

"z@z" wrote:

> Hello Zeus Thibault!
>


> | > But if the odds are too incredible, then some kind of order-creating
> | > principles apart from random mutation and selection must be assumed.
> |

> | When considering your cyt c sequence and your mother's, the exact same
> | incredible odds are involved - yet we don't "assume" any order-creating
> | principles, aside from heredity.

> Over one generation it is certainly not astonishing that DNA
> sequences remain unchanged. But if sequences do not change over
> 100'000, 1'000'000, 10'000'000, 100'000'000 or even more
> generations, your objection becomes unfounded.

>From paleontological evidence, chimps and humans diverged approx. 5 million
years ago. Given a generation time of ~15 years (probably and
underestimate), you only get about 300,000 generations.

> Neither you nor Howard has yet replied to this:

> [Zeus said -]
>


> "| As mentioned above, the cytochrome c proteins in chimps and humans are
> "| exactly the same. The clincher is that the two DNA sequences that code
> "| for cytochrome c in humans and chimps differ by only one codon, even
> "| though there are over 10^49 different sequences that could code for
> "| these two proteins.

[Wolfgang replied]

> " Because neither selection nor genetic drift can explain that
> " human and chimp DNA-sequences did not drift apart by random
> " mutations, neo-Darwinism is definitively dead !!!

> It is a logical consequence of neo-Darwinism that all possible

z@z

unread,
Jul 31, 1999, 3:00:00 AM7/31/99
to
Hello Zeus Thibault!

[snip]

| From paleontological evidence, chimps and humans diverged approx.
| 5 million years ago. Given a generation time of ~15 years (probably
| and underestimate), you only get about 300,000 generations.

To the tremendous progress over about 300,000 generations
corresponds a rather small change in the genome.

"For example, to determine the relationship between humans,
chimpanzees, and gorillas, a number of different molecules
have been sequenced and compared. It was expected that humans
would be easily shown to be more distant from the other
primates than those primates would be from each other. The
findings reveal that the evolutionary distances between the
three primates are almost identical no matter how you pair
them up. Some molecules favour a slightly closer human-chimp
relationship while others favour a closer human-gorilla
relationship. The obvious conclusion is that a human is just
another primate as far as molecular evolution goes."
(http://kat.microbio.umanitoba.ca/~gklassen/ read1.html)

We cannot explain the evolution of humans after their separation
from chimps by the same means as the evolution of bacteria. As I
have already stated in another post, it is not very likely that
natural selection could have acted on more than five births per
couple per generation, because many deaths are unrelated to
Darwinian fitness.

It would be helpful to know the number of births of chimps in the
wild and to study the effect of fitness on their death rates and
on their number of offspring.

Neo-Darwinism states that random mutations are the primary cause
of human evolution. So almost all deleterious mutations must be so
harmful that concerned gametes do not lead to pregnancy (or result
in early spontanous abortion), and the remaining (slightly)
harmful mutations in new-borns must not be substantially more
frequent than (slightly) beneficial mutations.

Two beneficial point mutations per generation affect 4 bits and
after 300'000 generations 150'000 bytes have been beneficially
changed. Can this account for all the differences between chimps
and child prodigies (e.g. chess, piano)? I don't think so.

In addition to that, the two beneficial mutations per generation
must spread over the whole population by selection whereas at
the same time deleterious mutations must disappear by selection.
Is this a realistic scenario? I don't think so.

| > It is a logical consequence of neo-Darwinism that all possible
| > DNA sequences coding for a given amino acid sequence are
| > equivalent. Therefore this theory predicts the existence of
| > lots of different DNA sequences for a given amino acid
| > sequence (and in addition to that also the existence of lots
| > of more or less functionally equivalent alleles for all gene
| > loci).

[snip]

| I completely agree with Tim when he states that "non-identical sequences can
| behave differently from one another under mutation and crossover - so they
| are not equivalent in this sense ... Different base pair sequences may even
| differ phenotypically." Both eukaryotes (including mammals) and prokaryotes
| are known to display codon bias, where one or more of the redundant codons
| are prefered over the others. Some of this preference is known to be due to
| selection.

My statement that "all possible DNA sequences coding for a given amino
acid sequence are equivalent" does not exclude the possibility that
they may behave differently in special situations (e.g. in the presence
of restriction enzymes). And I would argue that codon bias is no
prediction of neo-Darwinism. (It can be explained by the relative
abundances of the tRNA molecules assigned to each codon.)

Nevertheless, if muations are random then the disappearance of certain
codons must be explained by selection. This poses no problem in the
case of bacteria evolution because of their short replication cycles.

| One may think that this invalidates my argument, since it
| suggests that not all equivalent DNA coding sequences are truly functionally
| redundant, and that some are less fit than others.
|
| I showed that >10^49 DNA sequences can code for the exact same cytochrome c
| protein sequence in humans and chimps. If we further assume that 99.99 % of
| these sequences are actually less fit, due to such factors as suboptimal
| codon usage, there are still >10^45 genuinely functionally and phenotypically
| equivalent DNA coding sequences that could have been used to specify this
| cytochrome c protein.

I don't think that your assumption is consistent with human evolution.
There is already the problem of the missing 'functional equivalent
alleles' which must be attributed to selection. Now you assume that in
addition to that even 99.99% of silent mutations are so deleterious that
they disappear by selection. Remember, all biological innovations of
human evolution are said to derive from random mutations. The proportion
of beneficial to deleterious mutations is therefore a crucial factor.

| Obviously, my argument still holds even with this
| further assumption (which is more realistic). The huge numbers make it
| rather robust.

The huge numbers do not only make your argument rather robust but also
my derivation from your argument.

If "there are still >10^45 genuinely functionally and phenotypically
equivalent DNA coding sequences" then one must not take it for granted
that the common acestors of chimps and humans used only one single
version of more than 10^40 possiblities. There is also no obvious
reason within neo-Darwinism why most (all?) humans share one version
and most (all?) chimps share another one.

NEO-DARWINISM DOES REMAIN DEAD because it predicts either
huge numbers of functionally equivalent alleles and coding
sequences or so many constraints on non-deleterious mutations
that evolution becomes impossible (at least in the case of
species having long replication cycles).

[snip paragraphs showing that substitution rates show regularities]

Cheers, Wolfgang

Relevant extracts from my posts of June 99:
http://members.lol.li/twostone/E/deja3.html

howard hershey

unread,
Jul 31, 1999, 3:00:00 AM7/31/99
to
z@z wrote:
>
> Hello Zeus Thibault!
>
> [snip]
>
[snip]

>
> If "there are still >10^45 genuinely functionally and phenotypically
> equivalent DNA coding sequences" then one must not take it for granted
> that the common acestors of chimps and humans used only one single
> version of more than 10^40 possiblities. There is also no obvious
> reason within neo-Darwinism why most (all?) humans share one version
> and most (all?) chimps share another one.
>
> NEO-DARWINISM DOES REMAIN DEAD because it predicts either
> huge numbers of functionally equivalent alleles and coding
> sequences or so many constraints on non-deleterious mutations
> that evolution becomes impossible (at least in the case of
> species having long replication cycles).

This is a common error in understanding what happens in cases of neutral
drift, namely the erroneous idea that there will be an *accumulation* of
more and more functionally equivalent variants given enough time.

That does not happen. It does not happen because chance has no memory
and populations have real sizes.

Assume that you start with a population of some size (say 100,000
individuals) with 5 different equivalent alleles each representing 20%
of the total. When you look at that same population many generations
later, the typical result will be a population with one of these five
being present as the dominant allele (>80% of the population; the choice
of which one of the five is the dominant allele is entirely random) and
the remaining alleles being distributed largely among the other four.
One or more of the four alleles may have shrunk to the point of
extinction from the population. That is what happens with drift in real
populations. There are two 'walls' that limit the possible
frequencies: The wall of fixation and the wall of extinction. Neither
wall can be reached absolutely. There will always be new mutation
introducing variants. But in real populations (and ignoring new
mutation) that is the ultimate fate of neutral equivalent alleles. Sort
of a "ten little indians" story of chance removal until there is one.

Going the other way, if you start with a population having a single
allele being the dominant allele and all other alleles only being rare
(introduced by new mutation), the odds strongly favor the rare new
alleles going to extinction rather than increasing to eventually replace
the dominant allele with an equivalent allele purely by chance. Indeed,
the probability of a new functionally equivalent mutation in a
population becoming the replacement of the current allele is 1/N where N
is the total number of alleles in the population. The chance of a
particular mutation occurring in this population is N*u where u is the
particular mutation rate. Thus the rate of replacement of one allele by
a neutral equivalent in the population is u.

Zeus Thibault

unread,
Jul 31, 1999, 3:00:00 AM7/31/99
to
"z@z" wrote:

> | From paleontological evidence, chimps and humans diverged approx.
> | 5 million years ago. Given a generation time of ~15 years (probably
> | and underestimate), you only get about 300,000 generations.
>
> To the tremendous progress over about 300,000 generations
> corresponds a rather small change in the genome.

[snip]

> We cannot explain the evolution of humans after their separation
> from chimps by the same means as the evolution of bacteria.

If you take an objective view, there are really not that many physical differences
between humans and chimps (mostly the hair thing, which boils down to a few simple
mutations, and bidpedalism). Our brains are thrice as big, which corresponds to
~220 millidarwins for the 5 million years since divergence. We have directly
observed much greater evolutionary rates than that both in the lab and in the wild
(hundreds of darwins for recent invasions of new continents and lake habitats by
several species; thousands of darwins in Drosophila; also equine evolution has
fossil evidence of 300 millidarwins). We have no problem using Neo-Darwinism
(with neutral theory) to explain the rapid evolutionary rates observed in
Drosophila; why won't it apply to humans and chimps, too (besides appealing to
irrational incredulity)?

[snip lots of stuff about how neutral mutation theory and selection can't explain
observed polymorphisms or lack thereof]

I think you would benefit from reading Chs. 2, 3, and 9 of Li's "Molecular
Evolution."

Cheers,

Zeus


Richard Harter

unread,
Aug 1, 1999, 3:00:00 AM8/1/99
to
Marty Fouts <mathem...@usenet.nospam.fogey.com> wrote:

[snip chocolate wisdom]

>I've recently been toying with some math that suggests that if the
>rate of mutation introduction is high enough, a dominant allele would
>never arise. Now clearly, at this point it is merely twittling
>parameters in equations and bears no relationship to the actual
>situation, which leads me to wonder two things:
>
> * Are observed mutation rates high enough so that although
> particular alleles are driven to extinction the dominant allele
> never goes to being the only one?

Depends on what kind of mutations you are talking about. Over time
single point mutations keep recurring; if you are talking about point
mutations the answer is (with some qualifications) yes. OTOH there are
other kinds of mutations which, percentage wise, are a majority of the
mutations but which, individually, occur more rarely than point
mutations. The answer to your question, though, is that about 80% of
the genes are fixed in a given species, i.e., it depends on the gene.

> * Is is possible that mutation distribution is not necessarily
> random, so that certain mutations have a higher than random
> chance of arising?

The mutation rate varies with the site and the gene.

>The second case is the one that I find the more interesting to
>speculate about, and it has a correspondence in electronics in which
>certain classes of errors are more likely than others, but I have no
>idea if the chemistry of mutation has similar properties.

It does.

>The idea is interesting, because it predicts that if there is an
>instance in which a particular mutation has a high enough probability
>of arising and is neutral then it would never be driven to
>extinction.

There are genes with a negative selection coefficient (unfavorable) but
which are maintained in the population indefinitely. Hemophilia is one
of them IIANM.

>Is this a case of the math going into regions that the reality of the
>situation doesn't allow for, or has such behavior been observed?

Observed.


Richard Harter, c...@tiac.net, The Concord Research Institute
URL = http://www.tiac.net/users/cri, phone = 1-978-369-3911
The mere fact that my assumptions are wrong and my conclusions
are erroneous does not mean that I am not right in principle.


z@z

unread,
Aug 2, 1999, 3:00:00 AM8/2/99
to
Howard Hershey wrote:

[snip]

| > NEO-DARWINISM DOES REMAIN DEAD because it predicts either
| > huge numbers of functionally equivalent alleles and coding
| > sequences or so many constraints on non-deleterious mutations
| > that evolution becomes impossible (at least in the case of
| > species having long replication cycles).
|

| This is a common error in understanding what happens in cases of neutral
| drift, namely the erroneous idea that there will be an *accumulation* of
| more and more functionally equivalent variants given enough time.

Howard, the common error lies not in my understanding what happens
in cases of neutral drift, but in scientific carelessness.

For most gene loci only one or very few alleles exist. That's a
fact. Simulations of genetic drift can in principle explain this
fact under certain hypotheses. You and your fellow scientists
simply ignore the fact that these hypotheses, under which
genetic drift predicts correct allele distributions, do not
at all agree with reality.

We've already had a discussion on this in the thread "ENIGMAs
--- Innovations vs. Mutations". See my last post to you:
http://www.deja.com/=dnc/getdoc.xp?AN=503575173

And my last post to Ian Musgrave:
http://www.deja.com/=dnc/getdoc.xp?AN=501361153

The fact that substitution rates depend much stronger on time
than on the number of generations is also strong evidence
against neo-Darwinism. According to this theory it is expected
that substitution rates should rather depend on the number
of generations than on time.

[snip]


| Going the other way, if you start with a population having a single
| allele being the dominant allele and all other alleles only being rare
| (introduced by new mutation), the odds strongly favor the rare new
| alleles going to extinction rather than increasing to eventually replace
| the dominant allele with an equivalent allele purely by chance. Indeed,
| the probability of a new functionally equivalent mutation in a
| population becoming the replacement of the current allele is 1/N where N
| is the total number of alleles in the population. The chance of a
| particular mutation occurring in this population is N*u where u is the
| particular mutation rate. Thus the rate of replacement of one allele by
| a neutral equivalent in the population is u.

You have only repeated here what you have written in earlier
posts without taking into accout my counter-arguments:

"Assume that the probability for a selectively neutral random
mutation in a certain protein is 10^6. If we assume a constant
population of 10^9 with one single allele at the respective
locus in the first generation, we get 10^3 alleles differing
from the original form in the second generation. After a million
generations a huge number of different selectively neutral
allels have appeared and the frequency of the orginal allele
is reduced to around 37 percent.

That's simplest probability theory and cannot be denied in a
reasonable way."
http://www.deja.com/=dnc/getdoc.xp?AN=502417358

"In my (homozygous) example above, in the second generation 1000
new alleles have appeared. Every of these alleles would have a
theoretical probability of 10^-9 (inverse of the population size)
to reach 'fixation' by drift after a huge amount of generations,
only if no more mutations occured. But in every new generation
around 1000 alleles differing from the most frequent allele do
appear."
http://www.deja.com/=dnc/getdoc.xp?AN=503575173

My example starts with a population having only a single
allele. But if neo-Darwinism were right, it would be almost
impossible to get a population with only one allele at a
given locus because far more functionally equivalent (DNA)
sequences must exist than any species can have individuals.


Cheers, Wolfgang

howard hershey

unread,
Aug 2, 1999, 3:00:00 AM8/2/99
to
z@z wrote:
>
> Howard Hershey wrote:
>
> [snip]

>
> | > NEO-DARWINISM DOES REMAIN DEAD because it predicts either
> | > huge numbers of functionally equivalent alleles and coding
> | > sequences or so many constraints on non-deleterious mutations
> | > that evolution becomes impossible (at least in the case of
> | > species having long replication cycles).
> |

That is exactly what will not happen. An assumption I think you are
making is that all the 10^3 variant alleles will or must accumulate.
That does not happen. The probability is that all of those 10^3 variant
alleles will be lost (go to extinction) over time (I don't have my books
handy, but the average transit time of a neutral allele can be
calculated). And the distribution of the losses of these neutral
variants is not a bell-shaped curve, but a skewed curve. The largest
number of these 1000 will be lost quickly and only the occassional
variant will perform a random walk to any significant frequency in the
population. Of course, selectively non-neutral deleterious variants
will be lost even more quickly than neutral ones. And this is without
even questioning your numbers. Perhaps you can tell me how you derived
your idea that neutral variants will accumulate to 63% of the
population? My guess is that you made some mistake in the assumptions
behind your calculation. The post you refer me to makes no mention of
how this value was calculated.

I think that the point where the loss of neutral alleles by chance alone
is balanced by the input of new neutral alleles by mutation is (on
average) pretty far from your estimate and closer (on average, but not
in particular cases) to what is seen: one allele being the most common
and a few percent of different neutral variants. The frequency of such
neutral variants *will* depend, to some extent, on the history of the
species. Ancient species will have more variance. Species that have
undergone more recent population constrictions will have less variance.
The degree of heterozygosity in a species can be estimated by looking at
electrophoretic variants of different proteins. I will have to look up
that data when I get the chance. I think that its about 15% for humans
(that is about 15% of loci show heterozygosity).


>
> That's simplest probability theory and cannot be denied in a
> reasonable way."
> http://www.deja.com/=dnc/getdoc.xp?AN=502417358
>
> "In my (homozygous) example above, in the second generation 1000
> new alleles have appeared. Every of these alleles would have a
> theoretical probability of 10^-9 (inverse of the population size)
> to reach 'fixation' by drift after a huge amount of generations,
> only if no more mutations occured. But in every new generation
> around 1000 alleles differing from the most frequent allele do
> appear."
> http://www.deja.com/=dnc/getdoc.xp?AN=503575173
>
> My example starts with a population having only a single
> allele. But if neo-Darwinism were right, it would be almost
> impossible to get a population with only one allele at a
> given locus because far more functionally equivalent (DNA)
> sequences must exist than any species can have individuals.

In any species where there is a relatively small initial population
after speciation, many genes will quickly have essentially only one
allele. Again, I think the error you are making is assuming that
alleles can only be gained and not lost.
>
> Cheers, Wolfgang


z@z

unread,
Aug 2, 1999, 3:00:00 AM8/2/99
to
Hello Howard!

[snip]

| > "Assume that the probability for a selectively neutral random

| > mutation in a certain protein is 10^-6. If we assume a constant


| > population of 10^9 with one single allele at the respective
| > locus in the first generation, we get 10^3 alleles differing
| > from the original form in the second generation. After a million
| > generations a huge number of different selectively neutral

| > allels have appeared and the frequency of the original allele


| > is reduced to around 37 percent.
|
| That is exactly what will not happen. An assumption I think you are
| making is that all the 10^3 variant alleles will or must accumulate.
| That does not happen. The probability is that all of those 10^3 variant
| alleles will be lost (go to extinction) over time (I don't have my books
| handy, but the average transit time of a neutral allele can be
| calculated). And the distribution of the losses of these neutral
| variants is not a bell-shaped curve, but a skewed curve. The largest
| number of these 1000 will be lost quickly and only the occassional
| variant will perform a random walk to any significant frequency in the
| population. Of course, selectively non-neutral deleterious variants
| will be lost even more quickly than neutral ones. And this is without
| even questioning your numbers. Perhaps you can tell me how you derived
| your idea that neutral variants will accumulate to 63% of the
| population? My guess is that you made some mistake in the assumptions
| behind your calculation. The post you refer me to makes no mention of
| how this value was calculated.

My calculation assumes that every new generation reduces the
number of the original allele by the factor

1 - 10^-6

After a million generations the allele is reduced to

(1 - 10^-6) ^ 10^6 = 1/e = 37%

You probably agree with me that in the second generation we have
1000 neutral variants. The probability that any of theses variants
will reach fixation is each 10^-9. The probabilily that together
they will edge out the original allele is 10^-6. For that to
happen it would be necessary that random drift increases their
number from 1000 to 1'000'000'000.

What happens in the third generation with the variants of
the second generation? Several variants may have disappeared,
but others may have two or even more copies. On average
1000 alleles differing from the original allele give rise
to 1000 differing alleles one generation later. The
probability of less than 800 or more than 1250 is rather
low.

Because 1000 further variants appear by mutation, in the
third generation there are on average 2000 variants, in
the forth generation 3000 variants, and so on. After
thousand generations variants have reached on average
a proportion of 0.1%.

| I think that the point where the loss of neutral alleles by chance alone
| is balanced by the input of new neutral alleles by mutation is (on
| average) pretty far from your estimate and closer (on average, but not
| in particular cases) to what is seen: one allele being the most common
| and a few percent of different neutral variants.

I'd be interested in more information on that. How many different
neutral (DNA) variants have been found in the case of human or
chimp cytochrome-c or in the case of animal ubiquitin?

| The frequency of such
| neutral variants *will* depend, to some extent, on the history of the
| species. Ancient species will have more variance. Species that have
| undergone more recent population constrictions will have less variance.

Extreme bottleneck effects seem to be to the only reasonable
explanation within neo-Darwinism for the low number of
functionally equivalent alleles.

[snip]

| > My example starts with a population having only a single
| > allele. But if neo-Darwinism were right, it would be almost
| > impossible to get a population with only one allele at a
| > given locus because far more functionally equivalent (DNA)
| > sequences must exist than any species can have individuals.
|
| In any species where there is a relatively small initial population
| after speciation, many genes will quickly have essentially only one
| allele. Again, I think the error you are making is assuming that
| alleles can only be gained and not lost.

On the one hand neo-Darwinism predicts that macro-evolutionary
steps are more likely in small than in large populations. On
the other hand however, if we look at the low probability of
beneficial mutations, a rather large population is needed for
useful adaptive mutations to occur with reasonable likelihood.


Cheers, Wolfgang
http://members.lol.li/twostone/E/evidence.html

Zeus Thibault

unread,
Aug 2, 1999, 3:00:00 AM8/2/99
to
"z@z" wrote:

> The fact that substitution rates depend much stronger on time
> than on the number of generations is also strong evidence
> against neo-Darwinism. According to this theory it is expected
> that substitution rates should rather depend on the number
> of generations than on time.

Hello Wolfgang,

Substitution rates should depend on the background spontaneous mutation rate,
all else equal. The spontaneous mutation rate should depend on time, not
generations, since most mutations are of chemical origin and not due to
replication errors. Thus substitution rates should depend mostly on time, if
they are strictly neutral.

Cheers,

Zeus

Zeus Thibault

unread,
Aug 2, 1999, 3:00:00 AM8/2/99
to
"z@z" wrote:

> | > "Assume that the probability for a selectively neutral random

> | > mutation in a certain protein is 10^-6. If we assume a constant


> | > population of 10^9 with one single allele at the respective
> | > locus in the first generation, we get 10^3 alleles differing
> | > from the original form in the second generation. After a million
> | > generations a huge number of different selectively neutral

> | > allels have appeared and the frequency of the original allele


> | > is reduced to around 37 percent.
> |
> | That is exactly what will not happen. An assumption I think you are
> | making is that all the 10^3 variant alleles will or must accumulate.

Hello Wolfgang,

I agree with Howard when he states that you must be assuming incorrectly that
new alleles only accumulate and do not disappear. I also must question the
numbers you use in your assumptions. 10^-6 is an *extremely* high mutation
rate - the average rate in humans is thought to be ~30 * 10^-9. In addition,
the effective human population size has been much less that 10^9 for most of
human history. It was never over 10^8 until the last 50 generations (since
about 1000 AD). Most evolutionists use a more realistic population figure of
10,000 to 50,000 for most of human history.

> After a million generations the allele is reduced to
>
> (1 - 10^-6) ^ 10^6 = 1/e = 37%

As said previously, this assumes no allele loss. It really is not applicable
to human evolution, since one million generations = 20 million years with an
effective population size of 1 million the entire time.

[snip]

> After a thousand generations variants have reached on average
> a proportion of 0.1%.

After a population bottleneck, polymorphism is extremely reduced. This
figure of yours, using unrealistically high mutation rates, after 20,000
years, with no allele loss, only gives a polymorphism of 0.1%. Even if this
were true, we'd have to sequence hundreds of individuals just to find one
variant, and this hasn't been done.

> Extreme bottleneck effects seem to be to the only reasonable
> explanation within neo-Darwinism for the low number of
> functionally equivalent alleles.

There are actually several other explanations that are more amenable to
testing and falsification. One is pleiotropic effects; another is genetic
hitchhiking. Following I quote the abstracts from two classic papers dealing
with these two phenomena, which fall out naturally from genetic selection and
neutral theory.

D Waxman and JR Peck, Pleitropy and the Preservation of Perfection, Science,
Vol 279, 1210-1213.

"A mathematical model is presented in which a single mutation can affect
multiple phenotypic characters, each of which is subject to stabilizing
selection. A wide range of mutations is allowed, *including ones that produce
extremely small phenotypic changes.* [my emphasis] The analysis shows that,
when three or more characters are affected by each mutation, a single optimal
genetic sequence may become common. This result provides a hypothesis to
explain the low levels of variation and low rates of substitution that are
observed at some loci. "

And from the conclusion:

"For many proteins, there is very little within-population variation. Low
amounts of variation can lead to low substitution rates (3, 9, 11), and
proteins exist that have apparently not changed at all for at least 100
million years (34, 36). Lack of variation can be a consequence of small
population size and genetic drift, but drift will not stop substitutions. In
some cases, natural selection is clearly the cause of low amounts of
variation (36-38) or infrequent substitutions (34, 36). In large
populations, stabilizing selection on one or two characters can produce low
amounts of variation and substitution only if mutations that have very small
selective effects are exceedingly rare. However, our results show that, when
each mutation affects three or more phenotypic characters, variation (and
thus, substitution) can be suppressed in favor of an optimal sequence even
when mutations of very small effect are common. "

The cytochrome c gene, and ubiquitin, are definitely pleiotropic, so this
evolutionary analysis applies.

NL Kaplan, RR Hudson and CH Langley , The "hitchhiking effect" revisited,
Genetics, Vol 123, 887-899

"The number of selectively neutral polymorphic sites in a random sample of
genes can be affected by ancestral selectively favored substitutions at
linked loci. The degree to which this
happens depends on when in the history of the sample the selected
substitutions happen, the strength of selection and the amount of crossing
over between the sampled locus and the loci at which the selected
substitutions occur. This phenomenon is commonly called hitchhiking. Using
the coalescent process for a random sample of genes from a selectively
neutral locus that is linked to a locus at which selection is taking place, a
stochastic, finite population model is developed that describes the steady
state effect of hitchhiking on the distribution of the number of selectively
neutral polymorphic sites in a random sample. A prediction of the model is
that, in regions of low crossing over, strongly selected substitutions in the
history of the sample can substantially reduce the number of polymorphic
sites in a random sample of genes from that expected under a neutral model. "

I do not know whether cyt c is in a region of the genome with low
recombination. However, many parts of the Drosophila genome recombine with
low efficiency, and the genetic hitchhiking model has predicted the low
polymorphism found there.

Human mitochondrial DNA undergoes no recombination (the whole mit genome is
linked), and mit genes are notoriously known for their lack of polymorphism,
in spite of a very large mutation rate.

Hope this helps,

Zeus


howard hershey

unread,
Aug 2, 1999, 3:00:00 AM8/2/99
to
z@z wrote:
>
> Hello Howard!
>
> [snip]

>
> | > "Assume that the probability for a selectively neutral random
> | > mutation in a certain protein is 10^-6. If we assume a constant

> | > population of 10^9 with one single allele at the respective
> | > locus in the first generation, we get 10^3 alleles differing
> | > from the original form in the second generation. After a million
> | > generations a huge number of different selectively neutral
> | > allels have appeared and the frequency of the original allele

> | > is reduced to around 37 percent.
> |
> | That is exactly what will not happen. An assumption I think you are
> | making is that all the 10^3 variant alleles will or must accumulate.
> | That does not happen. The probability is that all of those 10^3 variant
> | alleles will be lost (go to extinction) over time (I don't have my books
> | handy, but the average transit time of a neutral allele can be
> | calculated). And the distribution of the losses of these neutral
> | variants is not a bell-shaped curve, but a skewed curve. The largest
> | number of these 1000 will be lost quickly and only the occassional
> | variant will perform a random walk to any significant frequency in the
> | population. Of course, selectively non-neutral deleterious variants
> | will be lost even more quickly than neutral ones. And this is without
> | even questioning your numbers. Perhaps you can tell me how you derived
> | your idea that neutral variants will accumulate to 63% of the
> | population? My guess is that you made some mistake in the assumptions
> | behind your calculation. The post you refer me to makes no mention of
> | how this value was calculated.
>
> My calculation assumes that every new generation reduces the
> number of the original allele by the factor
>
> 1 - 10^-6
>
> After a million generations the allele is reduced to
>
> (1 - 10^-6) ^ 10^6 = 1/e = 37%

In short, you are ignoring chance completely. This equation would only
hold true for an infinite population with no variation due to chance
whatsoever. My entire point deals with what happens in a real finite
population of given size where there is chance fluctuation from
generation to generation. Your equation does not have any chance
variation but, rather, is entirely deterministic. I will need to look
up that value for the average transit time of a neutral variant from
initial appearance until removal (remember that the probability of
removal, rather than fixation, of a new variant is (N-1)/N or almost,
but not quite 1). I just don't have that information with me right now.


>
> You probably agree with me that in the second generation we have
> 1000 neutral variants. The probability that any of theses variants
> will reach fixation is each 10^-9. The probabilily that together
> they will edge out the original allele is 10^-6. For that to
> happen it would be necessary that random drift increases their
> number from 1000 to 1'000'000'000.
>
> What happens in the third generation with the variants of
> the second generation? Several variants may have disappeared,
> but others may have two or even more copies. On average
> 1000 alleles differing from the original allele give rise
> to 1000 differing alleles one generation later. The
> probability of less than 800 or more than 1250 is rather
> low.
>
> Because 1000 further variants appear by mutation, in the
> third generation there are on average 2000 variants, in
> the forth generation 3000 variants, and so on. After

> thousand generations variants have reached on average
> a proportion of 0.1%.

And if the average transit time of a neutral variant is 1000
generations, this is about as high as the totality of all the variants
will get. But let me look up the value.


>
> | I think that the point where the loss of neutral alleles by chance alone
> | is balanced by the input of new neutral alleles by mutation is (on
> | average) pretty far from your estimate and closer (on average, but not
> | in particular cases) to what is seen: one allele being the most common
> | and a few percent of different neutral variants.
>

> I'd be interested in more information on that. How many different
> neutral (DNA) variants have been found in the case of human or
> chimp cytochrome-c or in the case of animal ubiquitin?

These are small proteins with fewer *possible* neutral variants (and
more deleterious ones) than larger proteins. Again, the time since the
split of humans and chimps is too short for there to have been more than
one (and more likely) no allele change by purely chance drift
replacement of neutral variants. I would suspect that the frequency of
variant alleles of *these* genes in the population will be quite low
(perhaps less than a percent or tenth of a percent). OTOH, hemoglobins
are larger molecules with more neutral sites on a percentage basis, and,
because it is easy to find neutral variants by simple blood tests and
electrophoresis, it is possible to screen (and many screenings occur as
a by-product of medical procedures) many, many people. A significant,
but small, percentage of variants - this is a guess, but I think it is
in the range of between a few percent and a few tenths) at the amino
acid level are found (and DNA variants would, of course, be higher). If
you want an estimate of the percentage of neutral allelic variants
relative to the most common allele in a case where there is no selective
constraint, the fibrinogen peptide would be a good choice, but screening
for variants would be much harder technically.


>
> | The frequency of such
> | neutral variants *will* depend, to some extent, on the history of the
> | species. Ancient species will have more variance. Species that have
> | undergone more recent population constrictions will have less variance.
>

> Extreme bottleneck effects seem to be to the only reasonable
> explanation within neo-Darwinism for the low number of
> functionally equivalent alleles.

The percentage of genes with heterozygosity (defined as electrophoretic
variance) in the typical human is not startlingly different from species
that have been around much longer (within a factor of two or so, if I
remember correctly).
>
> [snip]


>
> | > My example starts with a population having only a single
> | > allele. But if neo-Darwinism were right, it would be almost
> | > impossible to get a population with only one allele at a
> | > given locus because far more functionally equivalent (DNA)
> | > sequences must exist than any species can have individuals.
> |
> | In any species where there is a relatively small initial population
> | after speciation, many genes will quickly have essentially only one
> | allele. Again, I think the error you are making is assuming that
> | alleles can only be gained and not lost.
>

> On the one hand neo-Darwinism predicts that macro-evolutionary
> steps are more likely in small than in large populations. On
> the other hand however, if we look at the low probability of
> beneficial mutations, a rather large population is needed for
> useful adaptive mutations to occur with reasonable likelihood.

Most speciation occurs on the margins of larger populations. And
nothing about evolution is guarranteed. Selection requires both the
mutation and the environment, just as a new business idea can occur
prematurely or in the wrong time and place. The industrial melanism
mutation probably occurred time and time again throughout the moth's
history as a species, with the bearer of this mutation being little more
than feed for some other organism. But this same 'bad' idea happening
in the right time and place produced reproductive success.
>
> Cheers, Wolfgang
> http://members.lol.li/twostone/E/evidence.html


Bernd Pichulik

unread,
Aug 3, 1999, 3:00:00 AM8/3/99
to

z@z <z...@z.lol.li> wrote in message
news:7o4s73$g3a$1...@pollux.ip-plus.net...

>
> On the one hand neo-Darwinism predicts that macro-evolutionary
> steps are more likely in small than in large populations. On
> the other hand however, if we look at the low probability of
> beneficial mutations, a rather large population is needed for
> useful adaptive mutations to occur with reasonable likelihood.
>
>

Yes, That is a point that has had me worried for some time now. If
you take into account that organisms resist mutation you are between
a rock and a hard place. In large populations the probability of
beneficial mutations is larger than in a small population but then it
will take so much longer for that beneficial mutation to spread
through the population.

If indeed the neo-Darwinist view is the better one, transmission
through the small population may be rapid but what about the
probability of getting a beneficial mutation in the first place?

If we now look at the emergence of Homo sapiens sapiens from,
allegedly a group of hominids 5 million years ago, then this problem
comes onto stark focus. 5 million years give us, say, 400,000
generations of about 12 - 13 years each. I just can't see the numbers
adding up for hominids to make the transition into Homo sapiens
sapiens in this short time span, starting with, say, a troop of a few
dozen individuals.

Bernd Pichulik

> Cheers, Wolfgang
> http://members.lol.li/twostone/E/evidence.html
>
>

howard hershey

unread,
Aug 3, 1999, 3:00:00 AM8/3/99
to
Bernd Pichulik wrote:
>
> z@z <z...@z.lol.li> wrote in message
> news:7o4s73$g3a$1...@pollux.ip-plus.net...
> >
> > On the one hand neo-Darwinism predicts that macro-evolutionary
> > steps are more likely in small than in large populations. On
> > the other hand however, if we look at the low probability of
> > beneficial mutations, a rather large population is needed for
> > useful adaptive mutations to occur with reasonable likelihood.
> >
Actually, this accounts, in part for the 'constancy' of the molecular
clock in the case of neutral substitution. The probability that a new
neutral mutation will become fixed by substitution is 1/N where N is the
number of alleles in the population. Obviously, the smaller the
population, the higher the probability of substitution. OTOH, the
probability of a mutation *occurring* in the first place is u*N, where u
is the rate of mutation. Obviously, the *larger* the population, the
more frequently will that mutation occur. These opposite effects
essentially lead to a substitution rate of u regardless of population
size.

> >
>
> Yes, That is a point that has had me worried for some time now. If
> you take into account that organisms resist mutation you are between
> a rock and a hard place. In large populations the probability of
> beneficial mutations is larger than in a small population but then it
> will take so much longer for that beneficial mutation to spread
> through the population.

As you point out with the hominids below, most large populations do not
really act like a Hardy-Weinberg population, but actually like a large
series of smaller populations with some exchange of genetic information
occurring between groups. This allows for both genetic clines (that is,
if you look at blood types, you will see a cline of % B in the
population as you cross from Asia to Europe) and for small local groups
to exhibit a high frequency of alleles that are rare in the rest of the
population. Populations, looked at genetically over geographic and
social space, are more often a stew than a puree. The assumption of H-W
equilibrium for large populations is a *simplifying* assumption that
allows one to use simpler equations (often without any major effect on
the result) but it *is* a simplification.


>
> If indeed the neo-Darwinist view is the better one, transmission
> through the small population may be rapid but what about the
> probability of getting a beneficial mutation in the first place?

Depends on how many 'small groups' there are. And, of course, the time
these groups exist. A 'beneficial' mutation often already exists as a
'neutral' or even 'detrimental' variant; all it needs is the right
zeitgeist to be transformed into a 'beneficial' variant.


>
> If we now look at the emergence of Homo sapiens sapiens from,
> allegedly a group of hominids 5 million years ago, then this problem
> comes onto stark focus. 5 million years give us, say, 400,000
> generations of about 12 - 13 years each. I just can't see the numbers
> adding up for hominids to make the transition into Homo sapiens
> sapiens in this short time span, starting with, say, a troop of a few
> dozen individuals.

Depends on how many changes are necessary to 'convert' the immediate
precursor organism into Homo sapiens. It is, as physical
anthropologists will tell you, often difficult to distinguish late H.
erectus from archaic H. sapiens. And even troops have significant
variance among individuals in the types of morphological features that
differ between erectus and archaic sapiens, as any family reunion will
demonstrate.

Even larger morphological changes in a population can occur in extremely
short periods of time if the population already has a range of variation
that includes the 'extreme' position to be reached, as the Grant's
research on finch beaks and research on anole leg lengths and guppy tail
size and coloration confirm within the lifetime of humans.

howard hershey

unread,
Aug 3, 1999, 3:00:00 AM8/3/99
to

I have picked up a copy of "Genes in Populations, 2nd ed." Eliot B.
Spiess, 1989 John Wiley & Sons. The relevant chapters are Chpts. 12-13.

The average time till fixation for the rare neutral allele that becomes
fixed is about 4Ne. That is, neutral fixation rate depends on
population size with large populations being slower than small (or
temporarily constricted) populations.

There is more *variance* in the time till extinction (the fate of most)
for a neutral allele, but the average time is 2(Ne/N) log(base e)(2N).
[I think you can excuse me for not having that memorized.] Ne is the
the effective population size. Comparing the two, fixation is slower
than loss by a factor of 2N/log(base e)(2N). This will not be a small
number. But to put it into an easier perspective to comprehend, "The
vast majority of isolates lose the mutant within the first *five*
generations, as might be expected from the deterministic result of
Fisher" (p. 372).

Incidentally, the above is why the usual creationist "fix" to the Noah's
ark problem of inadequate population variance (putting four different
alleles in the two survivors) is so much bull hockey. Barring a
miracle, one would have fixation in all but a few rare cases just a few
generations out.

As pointed out above, the typical fate of a neutral mutant allele is
extinction within five generations even without the assumption of
bottlenecks.
>
[snip]


howard hershey

unread,
Aug 3, 1999, 3:00:00 AM8/3/99
to
Zeus Thibault wrote:
>
> "z@z" wrote:
>
> > | > "Assume that the probability for a selectively neutral random
> > | > mutation in a certain protein is 10^-6. If we assume a constant
> > | > population of 10^9 with one single allele at the respective
> > | > locus in the first generation, we get 10^3 alleles differing
> > | > from the original form in the second generation. After a million
> > | > generations a huge number of different selectively neutral
> > | > allels have appeared and the frequency of the original allele
> > | > is reduced to around 37 percent.
> > |
> > | That is exactly what will not happen. An assumption I think you are
> > | making is that all the 10^3 variant alleles will or must accumulate.
>
> Hello Wolfgang,
>
> I agree with Howard when he states that you must be assuming incorrectly that
> new alleles only accumulate and do not disappear. I also must question the
> numbers you use in your assumptions. 10^-6 is an *extremely* high mutation
> rate - the average rate in humans is thought to be ~30 * 10^-9.

These are different mutation rates. The 10^-6 value is not unreasonable
for change in some proteins (including all detectable change in that
protein). The 10^-9 value is not unreasonable for the rate of
nucleotide substitution.

> In addition,
> the effective human population size has been much less that 10^9 for most of
> human history. It was never over 10^8 until the last 50 generations (since
> about 1000 AD). Most evolutionists use a more realistic population figure of
> 10,000 to 50,000 for most of human history.

His value is definitely high here as an average human population size
over the history of the species.


>
> > After a million generations the allele is reduced to
> >
> > (1 - 10^-6) ^ 10^6 = 1/e = 37%
>
> As said previously, this assumes no allele loss. It really is not applicable
> to human evolution, since one million generations = 20 million years with an
> effective population size of 1 million the entire time.

And neutral allele loss is really, really fast. Most neutral mutations
that occur are lost within five generations. I rather doubt that the
amount of variation seen at given gene loci is greatly influenced by an
*average* build up of neutral alleles. Rather, those loci that show
such variance are the cases where drift has produced a significant
frequency of an alternate allele.
>
[snip]


z@z

unread,
Aug 3, 1999, 3:00:00 AM8/3/99
to
Hello Howard Hershey!

[snip]

| I have picked up a copy of "Genes in Populations, 2nd ed." Eliot B.
| Spiess, 1989 John Wiley & Sons. The relevant chapters are Chpts. 12-13.
|
| The average time till fixation for the rare neutral allele that becomes
| fixed is about 4Ne. That is, neutral fixation rate depends on
| population size with large populations being slower than small (or
| temporarily constricted) populations.

That looks quite reasonable. In the case of a population
of 10^9 (what about zygosity?) this results in, on average,
4 x 10^9 generations till fixation of a single variant
allele. Trials with the simulator posted by Ian (thanks)
http://http.bsd.uchicago.edu/hgd-sad/HWSimulator/sim.cgi
also suggest that the mean number of generations needed to
get rid of one of two equally frequent alleles is proportional
to the population size and lies in the order of magnitude of
the population size. So if a species consists of 10^9
individuals, then around 10^9 generations are necessary
for one of the two alleles to disappear.

Until now I would have rather guessed that the number of
generations needed for fixation is not

number_of_generations = constant x population_size ^ 1

but at least

number_of_generations = constant x population_size ^ 1.5

However now, the proportionality between the two quantities
seems rather obvious to me. If the mean number of generations
needed to reach fixation is inversely proportional to the
square of the 'displacement rate' and the 'displacement
rate' is inversely proportional to the square root of the
population size, then the number of generations turns out
to be proportional to the population size.

Fortunately for me and unfortunately for you, the fact that
around 10^9 generations lead to the fixation of a single allele
is irrelevant to my calculation:

"Assume that the probability for a selectively neutral random
mutation in a certain protein is 10^-6. If we assume a constant
population of 10^9 with one single allele at the respective
locus in the first generation, we get 10^3 alleles differing
from the original form in the second generation. After a million
generations a huge number of different selectively neutral
allels have appeared and the frequency of the original allele
is reduced to around 37 percent."

| There is more *variance* in the time till extinction (the fate of most)


| for a neutral allele, but the average time is 2(Ne/N) log(base e)(2N).
| [I think you can excuse me for not having that memorized.] Ne is the
| the effective population size. Comparing the two, fixation is slower
| than loss by a factor of 2N/log(base e)(2N). This will not be a small
| number. But to put it into an easier perspective to comprehend, "The
| vast majority of isolates lose the mutant within the first *five*
| generations, as might be expected from the deterministic result of
| Fisher" (p. 372).

What exactly is N?

[snip]

Wolfgang

Thread backwards:
http://www.deja.com/=dnc/getdoc.xp?AN=508432489 (Howard)
http://www.deja.com/=dnc/getdoc.xp?AN=508192334 (Howard)
http://www.deja.com/=dnc/getdoc.xp?AN=508070842 (me)
http://www.deja.com/=dnc/getdoc.xp?AN=508005177 (Howard)
http://www.deja.com/=dnc/getdoc.xp?AN=507985882 (me)

Zeus Thibault

unread,
Aug 3, 1999, 3:00:00 AM8/3/99
to
"z@z" wrote:

> Howard Hershey wrote:
>
> | The average time till fixation for the rare neutral allele that becomes
> | fixed is about 4Ne. That is, neutral fixation rate depends on
> | population size with large populations being slower than small (or
> | temporarily constricted) populations.

> | There is more *variance* in the time till extinction (the fate of most)
> | for a neutral allele, but the average time is 2(Ne/N) log(base e)(2N).
> | [I think you can excuse me for not having that memorized.] Ne is the
> | the effective population size. Comparing the two, fixation is slower
> | than loss by a factor of 2N/log(base e)(2N). This will not be a small
> | number. But to put it into an easier perspective to comprehend, "The
> | vast majority of isolates lose the mutant within the first *five*
> | generations, as might be expected from the deterministic result of
> | Fisher" (p. 372).
>

> What exactly is N?
>
> Wolfgang

Hello Wofgang,

N is the number of alleles in the population of interest. For a sexually
reproducing population, the number of alleles is twice the number of
individuals. Ne is the effective population size, which should be used in
most calculations in neutral theory. Ne is always smaller than the actual
N. There are four factors that can make Ne smaller than N:

1) Overlapping generations

Basically, this means that not all of the population is actively
reproducing. For humans, this effect gives Ne = N/3 approximately.

2) Male/female reproduction differences

In many mammalian populations, including human, fewer males contribute
gametes to the next generation than females. This is obvious in a polygamous
situation. This contributes to Ne as follows:

Ne = 4NmNf/(Nm+Nf)

where Nm is the number of reproducing males, and Nf is the number of
reproducing females. You can immendiately see that any disparity always
leads to Ne<N.

3) Population size is not constant

Many factors can lead to variable population sizes with time, including
environmental catastrophes, cyclical reproduction, local exterminations, and
colonization of new areas.

Here, Ne = n/(1/N1 + 1/N2 + 1/N3 ....)

where Ni is the size of population i, ranging from 1 to n.

Equivalently, Ne is the harmonic mean of each genera