Re: Experimental basis for the Non-Beneficial Gap Problem

On Jul 8, 5:22 pm, John Harshman <jharshman.diespam...@xxxxxxxxxxx>
wrote:
Seanpit wrote:
On Jul 7, 5:39 pm, John Harshman <jharshman.diespam...@xxxxxxxxxxx>
wrote:
Do you not remember the many times I've listed the papers by Yockey,
Sauer and Olsen in this forum dealing with this topic?
Do you know how many residues in CytoC are "invariant"? Partially
variant? Some suggest that between 27 and 29 amino acid residue
positions are invariant due to "biological requirements" (
http://www.springerlink.com/content/p840152h37w487uj/). This would
strongly suggest that the minimum size requirement cannot be less than
the invariant residue requirement for the CytoC function.
No it doesn't. It suggests that the particular protein we happen to have
for this function has 29 residues that are necessary in the context of
that protein. It doesn't mean that some other, smaller protein to
perform the same function is not possible. Still, if you have a
structural rather than functional definition of "cytochrome c", then you
are right. Try not to confuse structure and function.

On please. This is an extremely strained argument at best. Good luck
finding a functionally beneficial CytoC protein with less than 60 or
70aa, much less 29aa.

What evidence do you have to the contrary? You don't seem to have any.
Now in fact cytochrome c is fairly constrained in structure because it
has to bind to two other proteins in the electron transport chain, as
well as cofactors. So you might be right in this case; one might have to
alter relevant portions of the other two proteins in a complementary way.

That's a fundamental problem. It is the surrounding environment in
the form of other systems within the organism with which the system in
question must interact that put structural constraints on the system
in question. Not just any non-homologous arrangement of equivalent
size is going to be able to do the job. That is one of the main
reasons beyond the ratio estimates of those like Yockey, Sauer, and
Olsen, etc.

Come on John, these are published peer-reviewed estimates that are
unchallenged in literature. If you really think that these authors
were so far off base, the burden of evidence is really on you. What
evidence do you have to suggest that these estimates are likely to be
significantly off base? As far as I can tell, you've presented
nothing but incredulity without any real substance behind your
doubt.

I mean, what is your suggested ratio for CytoC functionality in
sequence space? Do you think it is likely to be as low as say, 1e-10
or maybe 1e-5? Do you really think this is at all likely given the
biological constraints of CytoC that are actually known?

Sure, it might be possible to produce a useful 20aa
CytoC protein. It is just that the odds against this notion are
enormous given the available evidence on the size and specificity of
CytoC that we have available. But, there is a tiny chance that you
could be right - - that's the nature of science. Nothing is 100%
knowable.

True, especially if you are looking in the wrong place. Would you not
agree that if cytochrome c had evolved, looking at modern examples would
be an extremely biased sample to use in determining limits on the
ancestral protein?

Not at all, since you have absolutely no idea what the "ancestral
protein" might have looked like. The best evidence one has available
to estimate the ratio for a function like CytoC must be based on what
we actually know works.

Authors like Yockey and Sauer and Olsen use this same sort of thinking
to calculate their own rough estimates of ratios of sequences that
carry certain functions.
If so, then we can throw out their results because they're based on
faulty reasoning.

Says you? ; )

If you disagree, attack my reasoning. "Says you" isn't much of an attack.

You haven't presented any reasoning at all - just bald incredulity and
the bold statement that the listed authors used "faulty reasoning".

It seems that in 1978 Yockey calculated the ratio of cytochrome c
sequences in sequence space via phylogenetic sequence comparisons and
then published an article in the Journal of Theoretical Biology
(Yockey, 1981, J Theor Biol, 91) suggesting that the ratio of
functional cytochrome c sequences in sequence space of about 100aa
would probably be around 1 in 1e65.

Faulty reasoning, since at a minimum it doesn't account for proteins
that would do the same job but were unrelated in sequence, and proteins
that would do the same job, but less well.

The number of unrelated proteins that would be able to do the same
job, even if this number were in the trillions, would not change
Yockey's proposed ratio to any significant degree.

True. But that was a nonsense objection. Let's say there were two
unrelated proteins that would do the same job. You've just halved
Yockey's number. Don't forget that the unrelated proteins would have
variants also.

If there were two unrelated non-homologous proteins of equivalent
minimum size requirements that both had at least some level of CytoC
functionality with a similar degree of specificity, the size of each
island would be equivalent in sequence space - as you note. What
would that do to Yockey's estimated ratio of 1e-65? What is 1e65
divided by 2? Is it not 1e64? Now, what does that really do to
Yockey's main point? Hmmmmm? What would a trillion (1e12) such non-
homologous CytoC islands do to Yockey's ratio? Change it to 1e-53?
Big deal - right? Do you still not see the statistical irrelevance of
your unlikely scenario? The point of the authors, that the ratio of
any sequence that has a particular function like CytoC vs. those that
have no useful level of that type of function, is very low.

That main point is overwhelmingly obvious and is unchallenged in
mainstream scientific literature.

There are at least four reasons why Yockey's number doesn't do what you
want, even for cytochrome c.

1. It doesn't account for unrelated sequences with the same function.

Statistically irrelevant - - as noted above.

2. It doesn't account for related sequences that would do the same
function, but less well than extant ones.

Yes it does. You are wrong on this point. All of the authors,
especially Yockey, take into consideration any protein that would
perform the function in question to any useful degree vs. those that
would have no useful function of the type in question at all.

3. It doesn't account for the sequences that would perform any function.
Nobody says that cytochrome c is necessary for life, or even for the
general function of getting energy out of food. If anything, you should
be asking how likely it is for a random protein to perform any useful
function whatsoever. We happened to get cytochrome c. In another world,
we might have something else entirely, functioning in some other system
entirely. You are pointing to the rarity of royal flushes in spades when
you should be calculating the chances of getting a hand better than that
of other players.

The relative rarity of particular types of functional systems in
sequence space can be used to estimate the overall rarity of all
potentially beneficial functions of all kinds at particular structural
minimum threshold requirements (i.e., minimum size and specificity).
Overall, the vast majority of potential sequences in sequence space
have no beneficial potential at all at a given level of minimum size
and specificity requirements. You yourself once recognize this as
obvious. I once wrote:

"It seems quite obvious to me that given a particular creature,
such as a bacterium, that the vast majority of possible amino acid
sequences/proteins of a given length will have no beneficial function
for that creature in its current environment."

You responded by saying, "Agreed. This is obvious."

Now I guess it is not so obvious to you any more?

4. And of course proteins aren't random sequences, and nobody says they
arose as random sequences, except in rare cases that are interesting for
that reason. So your model is pointless at its very base.

I've never argued that proteins arose from random sequences given the
existence of the "first living thing". Start with whatever you want
as a pre-established functionally beneficial starting point. My
argument is that regardless of your chosen starting genome, the odds
of adding additional beneficial functional systems to that genome is
exponentially related to the minimum structural threshold requirements
of potentially beneficial systems. The higher the minimum size and/or
sequence specificity requirements of a potentially beneficial system
in sequence space, the exponentially lower the odds that that system
will be within striking distance of your chosen starting-point genome
- - within what anyone would call a reasonable amount of time (i.e., a
few hundred million years).

Do you think Yockey, Sauer and Olsen pulled their estimates out of
their arses? Hmmmmm?

No, I just think they made estimates using faulty reasoning, which you
have extrapolated way beyond even their data.

Ok - - There's just no reasoning with you. What's the point of
continuing this discussion with someone who demands absolute proofs
before even considering the meaning of available evidence?

I don't recall asking for absolute proofs. I recall pointing out reasons
why you can't use Yockey's numbers for the purpose you intend.

Your reasons are not statistically significant relative to Yockey's
main point - - as noted above. You argue that the listed authors, to
include Yockey, used "faulty reasoning" to come up with their
published unchallenged peer-reviewed estimates. That's quite a bold
statement given the vacuity of your very strained very unlikely
counter arguments.

I will note that you have not used their methods on any other system,
only extrapolated their numbers. Though in fact the data probably don't
currently exist to do any such thing with most proteins or multi-protein
systems.

Science is all about extrapolating what is available. The
extrapolations make perfect sense and are backed up by real-time
observations (i.e., an exponential stalling out effect of observable
evolution in action as one moves up the ladder of functional
complexity toward the 1000aa threshold).

Science is about using valid reasoning. Your reasoning hits a reef;
several reefs, in fact. But what is this real-time observation you mention?

Evolution is observed all the time, in real time, when it comes to
novel functions that require no more than a few hundred fairly
specified amino acid residues working together at the same time.
There are hundreds if not thousands of such examples published in
literature. Evolution happens even more commonly and rapidly at lower
levels. However, evolution does not happen at all in observable time
beyond the level of a few hundred specified residues. It certainly
has not be observed at all beyond the 1000aa threshold level. There
is a observed exponential decay cure in evolvability that is directly
related to the likely minimum size and/or specificity requirements of
potentially beneficial systems.

So the proposed mechanism of mutation/selection may not be sufficient
to account for all of evolution? Did I read you correctly? Random
mutation and natural selection are "evolution". How can there be real
"evolution" without this mechanism?

I don't have enough information to be sure that mutation and selection
are adequate to explain all adaptive evolution, though it does look like
the way to bet at the moment. There can of course be real evolution
without either. As I've pointed out many times, common descent is
evolution, even if God personally zaps every mutation into being, or
zaps multiple mutations at one time to cross your "neutral gaps".

That's a very unusual definition of "evolution". It seems to fit more
with the definition of "slow creation" than with what any mainstream
scientist would call "evolution". Really, such slow creation is not
evolution at all in any standard Darwinian-style sense of the word.

You seem to be sort of a ToE agnostic - - at least when it comes to
the proposed mechanism of random mutation and function-based
selection. You don't really know how the mechanism works. You just
favor it because you can't think of anything else. That's also not
science.

Sure it is. Science is always provisional. We can only test the theories
we have. You, having no testable theory, present no challenge to natural
selection.

If you have no way to test the potential and limitations of NS, it
isn't a science regardless of if you know or don't know of any other
viable alternative.

But I'm certainly open to the possibility of other
mechanisms. I see no clear evidence that any other mechanisms are
needed, but there is currently room to postulate them, if you cared to
do so.

The other mechanism, of course, is ID. The same basic argument for
detecting design is used by SETI scientists. Even you use such
arguments behind your design hypotheses for various situations - like
your own scenario of finding a messy kitchen in your house. You said
yourself you would first assume ID in such a situation.

You really believe in the ToE, not because you have any scientific
evidence supporting the mechanism, but because of the your notion that
the nested pattern of life can best be explained by common descent.
But a belief in common descent isn't the same thing as a belief in the
ToE. The ToE is based on the mechanism of random mutation and
function-based selection. Since you have no scientific basis to
support the mechanism, where is your scientific basis to support your
belief in the ToE?

You certainly know how to tie yourself in knots, but forgive me if I
don't join you. As I've said before and you always deny for no clear
reason, common descent and the mechanism of adaptive change are
independent hypotheses. Evidence for or against one is not evidence for
or against the other. Common descent has very good evidence in its
support, evidence that I'm intimately familiar with, and this is the
reason I "believe" in it.

Again, a belief in evidence for common descent isn't the same thing as
a belief in evidence for Darwinian-style evolution. These are two
somewhat related but still distinctly independent concepts or
theories. Even Behe believes in common descent. Obviously, he
doesn’t believe in Darwinian-style evolution. Are you in the same
camp as Behe? I doubt you would admit it. Yet, you don't really seem
to have any good reason not to be.

There is also good evidence that natural
selection operates in the present, and some evidence that it has
operated in the past, though this is limited to fairly recent (a few
million years) cases. I don't know of any evidence that it operated in
the very distant past, say in the evolution of the bacterial flagellum.
At best, I suspect we can only come up with evidence that nothing in the
evolution of the flagellum conflicts with mutation/selection as the
mechanism. Nevertheless, we know that mutation/selection do exist, and
this gives them a leg up on other postulated mechanisms that we have no
such knowledge for. You, of course, have never actually postulated a
mechanism, certainly not a testable one.

I also believe that NS operates in the present. There are an
overwhelming number of examples of its operation. The problem is that
its creative power is demonstrably limited to very very low levels of
functional complexity - relatively speaking. It just doesn't do
anything beyond these low levels and it stalls out, observably, in an
exponential manner at low levels. Statistically these limitations can
be fairly well defined - to a very useful degree of predictive
power.

Beyond this, I have postulated the same hypothesis and basis for this
hypothesis that SETI scientists postulate - the ID-only hypothesis -
to explain functionally complex features that clearly go well beyond
the observable and statistical creative powers of the proposed
Darwinian mechanism.

This is a very interesting statement you've made here - - very
interesting indeed. I wonder of others on T.O. share your position?
I'm pretty sure those like the Hershey Collective, Ron Okimoto, Ian
Musgrave etc. do not.

I'd be interested to know. Anyone else reading this? I'm pretty sure,
given past history, that you being pretty sure of something doesn't
measurably increase the chance that it's true.

You just said that, "I can't be sure that mutation and selection are
adequate to explain all adaptive evolution. . ." Of course, you did
add the disclaimer that, "though it does look like the way to bet at
the moment". That's pretty wishy-washy sounding language if you ask
me. Much more wishy-washy compared to the bravado of those like
Howard Hershy, Okimoto, and many others in this forum and in
mainstream science to be sure. Where are the problem points, do you
think, for the proposed mechanism that aren't very clear to you? Your
position regarding the proposed Darwinian-style mechanism doesn't
really sound all that solid as far as a scientific theory with very
high predictive value if you ask me. I certainly do wonder who else
in this forum might express this same notion in the same or similar
language to that you use here? That would be interesting to know -
for sure ; )

Sean Pitman
www.DetectingDesign.com

.

Relevant Pages

  • Blind Faith in the Mechanism of the ToE
    ... The degree of sequence constraint for such named functions ... varies from protein to protein and within proteins. ... distance or total size of the system. ... named function arose by evolution. ...
    (talk.origins)
  • Re: Most valuable poster
    ... irrelevant mathematical models can't falsify the theory of evolution, ... "functional complexity" is a ratio extracted from a 15-year old paper ... to transform one sequence into another, e.t.c.), and "random walk time" ... The distance between a protein and two of its ...
    (talk.origins)
  • Re: Most valuable poster
    ... understand that denying evolution on *some* level would be lunacy. ... "neutral gap size" is the edit distance ... one sequence into another, e.t.c.), and "random walk time" is one ... The distance between a protein and two of its ...
    (talk.origins)
  • Direct Experimental Evidence for Non-Beneficial Gaps
    ... So we are all agreed that a random sequence of amino acids is very ... The initial proteins would not have such constraints. ... trying to disprove evolution using a static model. ... between the size and sequence flexibility of a protein and Dursten's ...
    (talk.origins)
  • Re: Most valuable poster
    ... nylonase or lactase evolution examples. ... residues), you have the ability, so you say, to tell us *exactly* what ... the average gap size is based on the size of the end product. ... recognizable sequence homologs or recognizable intermediate functions. ...
    (talk.origins)
Loading