Re: Direct Experimental Evidence for Non-Beneficial Gaps

On Jul 9, 5:27 pm, John Harshman <jharshman.diespam...@xxxxxxxxxxx>
wrote:
Seanpit wrote:

< snip >

Based on the number of changes tolerated, the
total ratio of all possible sequences in sequence space was
estimated. For the case of lambda repressor function, Sauer and Olsen
estimated a ratio of around 1 in 1e65. This experimental finding
supports the finding of Yockey based on sequence variability analysis
in known sequences.

Do I really have to read this? Can't Howard do it? This may take a while.

It's not that long of a paper . . . The rest of the listed papers
below are also quite relevant and very interesting.

http://www.tbiomed.com/content/pdf/1742-4682-4-47.pdf

More recently Axe (2004) has performed site directed mutagenesis
experiments on a 150-residue protein-folding domain within a B-
lactamase enzyme.

"Starting with a weakly functional sequence carrying this signature,
clusters of ten side-chains within the fold are replaced randomly,
within the boundaries of the signature, and tested for function. The
prevalence of low-level function in four such experiments indicates
that roughly one in 1e64 signature-consistent sequences forms a
working domain. Combined with the estimated prevalence of plausible
hydropathic patterns (for any fold) and of relevant folds for
particular functions, this implies the overall prevalence of sequences
performing a specific function by any domain-sized fold may be as low
as 1 in 1e77, adding to the body of evidence that functional folds
require highly extraordinary sequences."

http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6WK7-4CVV2G...

These findings are based on direct experimental evidence that tests
the boundaries of functionality John. What more do you want besides
absolute proof? - which is impossible and not a requirement of
science?

Also consider a recent 2007 paper by Dursten et. al.

http://www.tbiomed.com/content/pdf/1742-4682-4-47.pdf

Using these very same ideas, Dursten actually proposes a mathematical
formula to estimate the "Functional Sequence Complexity (FSC)" of a
protein sequence. And, surprise surprise, there is a good correlation
between the size and sequence flexibility of a protein and Dursten's
FSC number. The greater the size and/or specificity of a protein, the
greater its FSC. See Figure 2:

http://www.tbiomed.com/content/download/figures/1742-4682-4-47-2.PDF

Again, this should be intuitively obvious to you, and I think it is.
This experimental and statistical evidence only cements the obvious.

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.
Even if that estimate is logically invalid, for the reasons I have
stated? Sounds as if you're looking for your keys under the street light
because the light's better there.

You haven't stated any reasonable reasons. Several of your points are
completely wrong - to include your suggestion that the estimated ratio
do not account for weakly beneficial sequences, that they are only
based on sequence comparisons without testing for functionality, and
that there is absolutely no reasonable evidence to back up Yockey's
estimated ratios. You don't seem to be very informed on this
particular topic.

I agree that I haven't looked into it too deeply. But you have been
unable to defend your claims either. Yockey's numbers do *not* count
weakly beneficial sequences. They only count variation in existing
sequences. If you disagree, how did Yockey account for suboptimal sequences?

He made the assumption that substituting a fairly restricted residue
position with one that has significant chemical differences compared
to the original would destroy the functionality of a system beyond the
level of usefulness. This assumption was used in his estimates of the
overall ratio. His assumptions turned out to be consistent with the
experimental evidence provided by those like Sauer, Olsen, Axe, et.
al.

Not according to the above listed authors. Having 10^55 non-
homologous islands that have the same function at the level of a
minimum 100aa requirement or so is highly unlikely. Why? Because of
the previously mentioned biological constraints of an integrated
system in a particular larger biosystem or living thing. Some basic
homology at key positions for an equivalent sized protein would be
required for most functional systems in order to fit and work properly
in a particular organism or subsystem.

True, but of course this could be an entirely different homology from
system to system.

You have to explain evolution from the perspective of a particular
gene pool that has its own particular functional constraints.

Beyond this, experimental
results do not back up this suggestion of yours. All the available
experiments dealing with this issue strongly suggest far greater
island isolation for most protein-based systems of even a relatively
small protein like a 100aa protein.

At this point I have a choice either to believe your claims or look it
up myself. I'm not sure which I'll do.

Look it up. Otherwise, why waste my time?

< snip >

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.
I don't see that, since Yockey's calculations, at least, are performed
by examining only the sequences in existing organisms. Are you saying
that some of these sequences are suboptimal?

That's exactly what I'm saying. Cassette mutagenesis by direct
experimentation and delineation of the boundaries of beneficial
functionality support this conclusion.

Let's be clear. Are you saying that Yockey's assay of existing sequences
included a range of suboptimal sequences? How did he know they were
suboptimal? Why were they not eliminated by selection?

His analysis goes beyond the range of the observed sequences present
in the sequences he actually compared. It assumes a much greater
degree of functional sequence flexibility than is actually observed.

"Consider that there are usually only 20 different amino acids
possible per site for proteins, Eqn. (6) can be used to calculate a
maximum Fit value/protein amino acid site of 4.32 Fits/site. We use
the formula log (20) - H(Xf) to calculate the functional information
at a site specified by the variable Xf such that Xf corresponds to the
aligned amino acids of each sequence with the same molecular function
f. The measured FSC for the whole protein is then calculated as the
summation of that for all aligned sites. The number of Fits quantifies
the degree of algorithmic challenge, in terms of probability, in
achieving needed metabolic function. For example, if we find that the
Ribosomal S12 protein family has a Fit value of 379, we can use the
equations presented thus far to predict that there are about 1e49
different 121-residue sequences that could fall into the Ribsomal S12
family of proteins, resulting in an evolutionary search target of
approximately 1e-106 percent of 121-residue sequence space. In
general, the higher the Fit value, the more functional information is
required to encode the particular function in order to find it in
sequence space. A high Fit value for individual sites within a protein
indicates sites that require a high degree of functional information.
High Fit values may also point to the key structural or binding sites
within the overall 3-D structure. Since the functional uncertainty, as
defined by Eqn. (1) is proportional to the -log of the probability, we
can see that the cost of a linear increase in FSC is an exponential
decrease in probability."

I admit that I don't understand this paragraph. I may attempt to read
the whole thing to see if it tells me any more.

Good . . .

http://www.tbiomed.com/content/4/1/47

You have to have at least some idea on how to quantify "vast majority"
in order to make this statement at all. What is the basis behind this
statement of yours? How do you know it is the "vast majority" of
sequences in sequence space that would not produce a beneficial
function for a given organism in a given environment? What is the
basis of this statement of yours?

Just a guess, really.

Not when you say that this conclusion is "obviously true". Something
that is just a guess isn't overwhelmingly "obvious".

It has a great deal to do with this claim. If the ratio of
potentially beneficial vs. non-beneficial is extremely low even at low
levels, and becomes exponentially lower and lower at higher and higher
levels of FSC (as illustrated by Durston et. al.), you end up with
expanding minimum gap distances that require an exponential increase
in the number of random mutations to cross.

Yes, and Yockey's number has nothing to do with the claim that the ratio
of potentially beneficial vs. non-beneficial is extremely low even at
low levels. As far as I can see.

You can lead a horse to water . . .

Yockey's whole discussion is all about the ratio of potentially
beneficial vs. non-beneficial proteins with the CytoC function in
particular. The papers listed by the other authors only support
Yockey's hypothesis. Almost all of the other authors reference
Yockey's work in their own papers to boot.

< snip >

I'd hardly call the evolution of a novel 1000aa-based function a "cat
giving birth to a dog". That's a classic evolutionists over
dramatization - a red herring to misdirect; a debating tactic.

It's an analogy. What you are asking for is the protein or genome-level
equivalent.

That's odd since the level I'm talking about here is about as low-
level as it gets. It isn't remotely close to your exaggerated
analogy. The analogy is so exaggerated as to be an obvious red
herring to anyone who knows anything about the topic. Evolution
happens easily and rapidly up to the level of a few hundred specified
residues, but not at all beyond this level - not even close to the
1000aa level. That's hardly asking why a cat doesn't give birth to a
dog. That's asking why evolution stalls out so rapidly at such a low
level in an exponential manner?

Arguing that millions of years are needed is a bogus argument since
millions of years are not needed at all at the level of a few hundred
residues. Evolution happens at such levels all the time right before
our eyes - - very very rapidly; often in one or two generations!
Then, suddenly, it doesn't happen any more just a few steps farther up
the ladder of functional complexity. What happened in those few
steps to stall evolution out so dramatically?

What prevents divergence over time from finding a novel function that
wasn't every in the genome before (beyond the 1000aa threshold level)
are the non-beneficial gaps that prevent the use of natural selection
as a guiding force. Without NS, the time required to find novel
beneficial functional systems at higher and higher levels grows
exponentially with each step up the ladder.

You aren't reading. I'm not talking about any neutral gaps. I'm talking
about divergence of function, each step advantageous.

Exactly. That means you *are* talking about non-beneficial gaps. You
are basically saying that they don't exist. That each mutational step
can be functionally advantageous. That's wrong. Each mutational step
cannot be functionally advantageous since the minimum structural
threshold requirements for the systems in question are simply too high
beyond a few hundred fairly specified amino acid residue positions.
You simply cannot build up an interactive system beyond this level of
complexity because of the linearly expanding non-beneficial gap
problem.

Are you saying
such divergence isn't possible,

It is possible. It does happen - all the time. It just doesn't happen
beyond very low levels of functional complexity.

or are you saying it can't go very far?

That's right . . .

Can you, for example, name a single protein family that encompasses
different "novel" functions such that it would be vanishingly unlikely
to have evolved by known mechanisms?

I can name you several protein-based systems that are overwhelmingly
isolated since they all require well over 1000 amino acid residues all
working together at the same time in a specific orientation with each
other: DNA transcription, RNA translation, vesicle transport, ATPase,
pinocytosis, exocytosis, flagellar motility, amoeboid motility, etc.
Not to mention organoid systems like mitochondria, Golgi bodies,
centromeres, etc.

Every living system has emergent high level functions that require far
more than 1000 fairly specified residue positions to work at minimum.
All such systems are extremely isolated from all other existing or
potentially beneficial systems in sequence space.

< snip >

NS is a description of a creative force that is suppose to produce
creative acts in a predictable manner. That should be testable in a
falsifiable manner if it is really a scientific hypothesis/theory.
The same thing is true of Newton's theories of gravity. They make
falsifiable predictions that creative predictive power if they come
true and avoid falsification. You are making a false comparison here
in order to avoid having to actually subject your notions of the
creative power of NS to any sort of falsifiable testing.

I'm not sure how one would test the potential limiations of gravity in
your sense. You certainly can't distinguish it from the angel theory.
But you seem to be claiming that unless I can re-evolve a bacterial
flagellum before your very eyes, the power of natural selection is
falsified.

I'm not asking for the re-evolution of the bacterial flagellum from
scratch. I'm asking for a demonstration of just one of the proposed
steps in the flagellar evolutionary model - or any other demonstration
of the evolution of any other novel functional system with a minimum
structural threshold of more than the relatively paltry 1000aa. This
shouldn't be so hard. After all, such demonstrations at the level of
a few hundred residues are very easy to come up with. Why is the
1000aa level so difficult to demonstrate? Hmmmm?

< snip repetitive claims that SETI scientists are somehow claiming
something fundamentally different than I'm claiming for their ID
hypothesis; whatever that might be is still very mysterious ; ) >

Behe doesn't propose any particular entity as a viable alternative.

Yes he does. He just doesn't want to admit it in some contexts. Behe
thinks that god did it, and has admitted this in other contexts.

Oh come on. The term "god" is just a word to mean some very
intelligent being. It doesn't have to be the Christian God or any
other particular God. Behe is a Catholic so he obviously favors the
idea that a Christian-style God did the job. However, the basic
argument does not require that the identity of the intelligent agent
or "god" be known - - exactly the same goes for SETI scientists. You
could call their alien intelligence that they are looking for a Little
Green God (LGG) for all they care. It really doesn't matter one lick
to the basic theory.

He, like other IDists, proposes a non-specific intelligent entity.
All he says is that some form of high-level directed deliberate
intelligence had to have been involved in the creation of certain
aspects of living things.

And has presented no evidence for the existence for this phenomenon,
whatever that would be.

He has presented just as much evidence as SETI scientists are looking
for with their peculiar radiosignals - evidence of something within
the realm of human-level creativity but beyond the realm of non-
deliberate natural production.

< snip >

Please do tell me the basic difference between my basis for suggesting
ID and a SETI scientist's basis for suggesting ID. What's the
fundamental difference as far as you can tell? How can you say that
the ID-only hypothesis is not testable in a falsifiable manner? That
is exactly the same hypothesis that SETI scientists are proposing.
Exactly the same.

We've been over this, and it bores me. Argue with someone else.

You are the one who keeps bringing it up - saying that I have no
evidence at all when I have exactly the same evidence that SETI
scientists are looking for.

You call such scenarios "plausible"? - when every single one of their
proposed steps involve vast gaps in non-beneficial genetic changes?

How can you say that? It involves only random phenotypic variance of
exactly the sort that regularly exists in living populations. It assumes
only that this random variance will be recharged by mutation, which
seems a reasonable assumption not requiring descent into any genetic
details.

We aren't talking about the evolution of eye color here. We are
talking about unique functional aspects between different kinds of
eyes. These differences are not simple phenotypic variances. Getting
such a functional variance when it did not already exist requires the
crossing of a huge non-beneficial genetic gap that cannot be traversed
by any series of shortly spaced sequentially beneficial mutational
steps.

Sure, the morphology doesn't seem like much between these proposed
evolutionary steps, but the underlying genetic changes needed are
absolutely enormous.

What is your evidence or argument for this claim?

This is true for just-so stories of eye
evolution as they are for flagellar evolution - like the best one I
could find proposed by Matzke. How are such scenarios statistically
"plausible"? They simply aren't. Not remotely so.

Why?

I ask you the same question. Please list out just one of the steps
that you think is phenotypically so simple and then demonstrate that
the underlying genetic differences that would be required are small
enough to cross in a few hundred million years via random mutation
alone.

As far my part, see my essays on flagellar and eye evolution:

http://www.detectingdesign.com/flagellum.html

http://www.detectingdesign.com/humaneye.html


How did I misconstrue what you said? Do you see good evidence for the
mechanism of random mutation and function-based selection or do you
just accept it beyond very low levels of functional complexity because
you don't see a decent viable alternative?

I reject all your personal jargon as not meaningful. I see evidence that
selection exists.

At what level? Do you really think it is reasonable to extrapolate
very low level examples that require minimums of no more than a few
hundred residues to higher level systems like you do? Upon what
basis do you do this? Have you actually done any statistical
calculations to see if you extrapolations are at all reasonable? I
don't think you have.

I see evidence that it can move a character
indefinitely far from the original population mean, given the right set
of changes in environment.

The odds of success by random mutations, small or large, are exactly
the same. The odds that a change in environment will cause the
realization of a beneficial target are also exactly the same as a
random genetic mutation. There is no statistical advantage either
way. Moving the population mean around on a beneficial island with a
particular type of function is no problem with a change of
environment. What is a problem is getting off that beneficial
starting point and actually finding a new island with a uniquely
beneficial type of function - such as a lactase enzyme in a genome
that never had the lactase ability before - or the nylonase function.
Those are real examples of the evolution of novel protein-based
beneficial functional proteins. Problem is, they are very low-level
examples require far fewer than 1000aa at minimum.

I don't know that selection is responsible
for huge historical changes (which you deny are changes, but that's your
personal problem), but I see nothing inconsistent with that view, and no
evidence of an alternative mechanism.

I don't deny the functional differences. What I deny is the notion
that the proposed mechanism could have done the job because of the
evidence for the expanding gap problem. Evidence which you have
admittedly not really investigated at all.

You yourself propose no evidence
for such a mechanism; at most you show that natural selection is
inadequate. The step from that to the idea that intelligence is the only
explanation is a large neutral gap indeed.

Uh huh - - not when we know that even human-level intelligence can and
has made this leap. Given that knowledge, the conclusion of ID behind
such a leap is no different than the conclusion SETI scientists
hypothesize for their artifactual radiosignals.

Accepting something
because you don't see a viable alternative is a much different
position from suggesting that a particular position actually has good
predictive value as a scientific hypothesis much less theory.

Considering that these major changes you don't like happen over immense
time scales, what exactly are we going to be predicting?

The changes I'm talking about are not major and at slightly lower
levels do not require immense time scales at all - only a handful of
generations (often just one or two generations). What is being
predicted here is the exponential decline of evolvability with
linearly increasing functional complexity (FSC as note earlier).

I'm also
currently unaware of any way to test for selection farther in the past
than comparison with neutral markers can reach. And come to think of it,
that's not a test for selection either, because god could have pushed an
allele and its hitchhikers to fixation personally, without selection.
God makes any test of anything pretty much impossible if you introduce
him into science.

Not if you use the ID-only hypothesis . . . like SETI scientists do.

Sean Pitman
www.DetectingDesign.com

.

Relevant Pages

  • Re: Howard Hersheys Challenge of Sean Pitmans Assumptions
    ... misrepresenting what he says this series of calculations does. ... This creates a ratio for a 1,000aa system of about ... potentially beneficial sequences at this level would be about 308 ... you ACTUALLY use to calculate "average gap size" represents "potential ...
    (talk.origins)
  • Re: Howard Hersheys Challenge of Sean Pitmans Assumptions
    ... This creates a ratio for a 1,000aa system of about ... you might be fooled into thinking that Sean actually is ... "potential beneficial vs. non-benecial" sequences for 100aa systems. ... your claim in the quoted mathematics. ...
    (talk.origins)
  • Re: Howard Hersheys Challenge of Sean Pitmans Assumptions
    ... ratio of sequences in sequence space that could produce some useful ... did not say that this ratio represents all 100 aa sequences that have ... in minimum structural threshold requirements. ...
    (talk.origins)
  • Re: Non-beneficial Gaps
    ... relative to the number of potential character arrangements drops off ... ratio of potentially meaningful vs. meaningless 2-character sequences ...
    (talk.origins)
  • Re: Most valuable poster
    ... Howard argues that a step of only one single residue change ... Howard argues that a single residue change could do what job? ... exponential decrease in the ratio of potentially beneficial vs. ... non-beneficial sequences within sequence space. ...
    (talk.origins)