Re: Of Mice and Straw Men

Seanpit wrote:
> hersheyhv wrote:
> > Seanpit wrote:
> > > hersheyhv wrote:
> > > > Seanpit wrote:
> > > > > Howard,
> > > > >
> > > > > You obviously don't get my point about cytochrome C at all. I'm not
> > > > > saying that CC cannot evolve or that it is not similar to other
> > > > > functional sequences within the electron transport chain.
> > > >
> > > > No. But you claim that it evolved from a random sequence by a random
> > > > walk with no possibility of intermediate utility.
> > >
> > > No, I never did say this.
> >
> > Your math says it. Your mathematical calculation of probabilities
> > explicitly implies that cytorchrome c and *every other system you
> > mention* starts out as a random sequence and proceeds in an aimless
> > random walk and only reaches *the predetermined* utility by chance
> > alone.
>
> Not true.

Of course its true. It is true no matter how many times you claim it
is not what you are doing. It is why you keep claiming that because
modern cytochrome c has 80 (or whatever) fairly specified residues that
you need to generate 80 different mutations *before* there is *any*
selectable function. That is, you *assume* that the way that
cytochrome c arose is by independently inventing, from a sequence that
had none of these 80 residues, cytochrome c *without* having any
intermediate of selectable utility (that is, it has to wander in the
ocean of non-utile functionality).

> My position is that with increasing minimum size and/or
> specificity requirements the average distance to the next closest
> beneficial island cluster of beneficial sequences grows in absolute
> distance in a rather linear fashion - creating a multicharacter neutral
> gap. This increase in absolute distance translates into an exponential
> increase in random walk distance/time.

And how does that differ from starting out from a sequence position
with *none* of the fairly specified residues that you observe in the
final "modern" protein and proceding through non-functional (*any*
function) sequence space until it hits your teleologically defined
beneficial island cluster (specifically any sequence that has modern
cytochrome c function)?

> As I've said before, you can start wherever you want and the average
> time it takes to find new islands of beneficial sequences will not
> change. You might happen to be really lucky and start really close to
> a particular beneficial island cluster at a high level, but odds are
> evolution will never make it off this island to any other at the same
> level or higher.

What you call "really lucky" is precisely how evolution works. It
works by swimming the small gaps between existing islands in an island
archipelligo. That is why you are arguing against a strawman when you
assert even the 80 fairly specified residues of cytochrome c. The
specific *function* of cytochrome c is to shuttle electrons from one
complex that acts as a proton pump to another that also acts as a
proton pump. This function evolved *after* the complexes existed and
has clear selective value when they do co-exist in a cell in an
O2-containing environment.

Unlike rotating bacterial flagella, cytochrome c only evolved once.
But that doesn't mean that it would necessarily be impossible to evolve
that function in some other way. The specific *function* of cytochrome
c requires a 1) heme-binding moiety, 2) interaction with the heme of
cytochrome c1, 3) interaction with the heme of cytochrome a, 4)
motility on the surface of the mitochondrial membrane, 5) a redox
potential between that of cytochrome c and cytochrome a. Cytochrome c
obviously did not *independently* evolve the heme binding function,
which is the same as that of cytochrome c1 and even more ancestral
cytochome c's involved in anaerobic metabolism. The covalent linkage
of the protein to heme in the c cytochromes is not necessary for
heme-binding and evolved well before cytochrome c arose. In short,
cytochrome c did not have to re-invent all 80 sites before one got the
specific *function* of cytochrome c (unlike you, I actually named the
*function*, because it is relevant to the evolution).

I have proposed a model whereby one can go from the island of
selectable cytochrome c1 *function* to the different island of
selectable cytochrome c *function* by a minimum of a single mutation (a
partial duplication). Once one has a selectable *function*, evolution
is very good at optimizing that function very rapidly. Most of your 80
"fairly specified" sites are undoubtedly primarily involved in
*optimizing* function.

> Functions like lactase, nylonase, cytochrome c, and other such
> relatively short single protein functions are not at very high levels
> of functional complexity. Their minimum requirements are no more than
> a few hundred fairly specified residues at most. The neutral gaps
> between what is already in the genome and such functions are on
> occasion fairly small and evolution can be and has been successful in
> real time. Go beyond this though to those functions requiring a few
> thousand fairly specified residues at minimum and evolution just
> doesn't happen.

And that is a strawman argument because there are no cases of evolution
requiring crossing such large gaps. You certainly have not presented
any. All you have done is present this bogus argument about a
*hypothetical* "gap" of 80 'fairly specified' residues that must be
crossed with no possibility of intermediate utility. I have explicitly
pointed out that there *are* intermediates and the gap from no specific
cytochrome c *function* (shuttling the electrons between complexes) to
selectable cytochrome c *function* may involve no more than a single
mutational event (a partial duplication) of a pre-existing functional
intermediate (cytochrome c1). One is a far cry from 80 in terms of the
probability of its occurring.

I have done the same for bacterial flagellar *function*. From a
precursor state of no rotary motility to rotary motility may take no
more than a single mutation (a chimeric protein formation) that links a
pre-existing motor to a pre-existing protein exporting pore. The "gap"
of thousands of "fairly specified residues" that, according to the
math you present, must wander through entirely non-functional sequence
space simply does not exist in the *real* model of evolution. In that
model both the motor and the pore existed in *functional* sequence
space, but did not have the motility *function* that is the specific
*function* of flagella. The motility *function* arose, in my model, in
a single step, not thousands of random, functionally useless steps.

> > So, yes. I will keep repeating that you either do not
> > understand the meaning of your calculations or you really believe that
> > evolution works the way that you propose, despite my pointing out
> > repeatedly that it does not. You have not presented a single case
> > where evolution ever *had to* produce a protein by a random walk that
> > required even 10 selectively neutral non-functional intermediates
> > before reaching the 'function'. You have merely *hypothesized* that
> > they *might* exist and then acted like you knew that they really do.
>
> Many of the proposed steps in flagellar evolution models require dozens
> of such multi-character changes. Not one of them has ever been shown
> to occur in real life. We aren't talking about evolving a new species
> here or any such major macroevolution change here. We are just talking
> about evolving a single step in the flagellar evolution pathway beyond
> very low levels, just one. It has never happened and it never will.

As I have pointed out, scientists have eliminated a protein that is
required for flagellar formation and *evolution* produced a new
solution that substituted an entirely different protein that fullfilled
the function.
>
> > > You continually misrepresent my position like
> > > this despite my correcting you about such strawmen caricatures over and
> > > over again.
> > >
> > > For the umpteenth time, evolution can start with any functional or
> > > non-functional sequence you want.
> >
> > And it remains a fact that *you* *never do* start with any functional
> > sequence and you object to anyone who starts their evolutionary steps
> > with any sequence that doesn't meet your mathematical fantasy of large
> > neutral gaps.
>
> Not true. I use the flagellar evolution models proposed by you guys.
> These models do start with functional subsystems - which is perfectly
> fine. It is just that the next steppingstone function is always too
> far away from what there is to start with at such high levels of
> functional complexity.

One mutational event (a chimeric protein formation) is too much? And
how does this jibe with your criticism of the ebg to betagalactosidase
result?

> > *You* always start with a random non-functional
> > sequence.
>
> Not true. I always start with a functional sequence and see how long
> it would take to find the next closest functional sequence at a higher
> level of minimum size and/or specificity.

I have *never* seen you propose that cytochrome c did not have to
reinvent the cytochrome c-like heme-binding moieties. That it could
acquire that function from a pre-existing protein (namely cytochrome
c1). Quite the opposite. Your math tells me that you think that
cytochrome c had to re-invent heme-binding. Either you do not
understand your own model or you are lying about it now.

> > And when I, or anyone else points out that the 'function'
> > that evolves, be it galactosidase 'function' from ebg, or cytochrome c
> > 'function' from cytochrome c1 or flagellar 'function' from a chimeric
> > fliG connecting the motor and the rotateable pore, you criticize us
> > because these do NOT require your mathematical fantasy of long neutral
> > (non-functional) walks from random, non-useful sequences.
>
> The minimum requirements for the galactosidase function are less than
> 400 fairly specified residues. Cytochrome c only requires a minimum of
> less than 80aa. Such functions are at a low level of functional
> complexity.

And those numbers are utterly irrelevant to how these proteins evolved.

> Flagellar motility, on the other hand, requires many
> thousands of fairly specified residues. The problem here is that none
> of the proposed steps, to include your chimeric FliG connecting the
> motor to a rotatable pore, have ever been shown to actually evolve in
> real life.

Other chimeric proteins have been observed. What you have to show is
that it is impossible to generate the link between a motor and the
rotateable pore by generating a chimeric protein. I would hazard to
guess that a similar, but different, protein exists in the archaean
flagella.

> The gaps involved are just too big. In fact, what seems
> like such a small jump to you is actually a huge canyon statistically -
> not crossable in trillions of years of average time.

Again. I am saying that there are no evolved systems that require
crossing such canyons *in the absence of functional intermediates*.
You have not presented any such systems. I have pointed out, in the
case of cytochrome c and in the eubacterial flagella, how a single
mutational event could produce a new selectable *function*. I would,
obviously, not think that this *function* would be optimized to modern
levels of function. But once you have the selectable *function*,
natural selection would rapidly lead to optimization for that function.
>
> > You dismiss
> > *real* evolution (descent with modification) in favor of a strawman
> > mathematical fantasy that every one agrees would not happen. Your
> > fantasy never NEEDS to happen. It is your ignorant idea that evolution
> > proceeds by the mechanism ennumerated in your mathematical fantasy that
> > I am criticising; evolution does not proceed by the mechanism
> > ennumerated in your mathematical fantasy. It proceeds via descent with
> > modification and extensive borrowing of pre-existing functional
> > moieties.
>
> At higher levels the required moieties simply do not exist in ratios
> that are large enough to continue up the ladder of functional
> complexity. Evolution starts stalling out at very low levels and never
> makes it up the ladder very far. That's what really happens in real
> life experiments. It's a fact.

Which only says that evolution did not occur by the strawman argument
you present. I certainly agree that if there are, in a particular
cell, no useful intermediates or only intermediates that require even
modestly large number of mutations, all of which would have no or less
utility than the starting point, evolution will not produce a result in
those conditions. The fact remains that *when* evolution does occur,
it does so by mechanisms similar to the ones I describe, not the ones
you describe. That you *assert* without evidence that certain modern
systems cannot have functionally useful intermediates (without having
the modern function) does not make that so.

> > > It is just that at higher levels of
> > > functional complexity, beyond those functions that only require a few
> > > hundred fairly specified residues at minimum, the multi-character gaps
> > > simply get to large to cross this size of a practical eternity of time.
> >
> > And, again, what is the relevance of talking about these "higher levels
> > of functional complexity" if those higher levels of *functional*
> > complexity arose by modification of a pre-existing sequence that both
> > *had* useful functional utility and also *had* most of the sequences
> > needed for the new 'function'.
>
> You simply assume that if subsystems of a particular function exist
> with independent functions that it would be a fairly easy thing to
> bring these subsystems together to form a larger more complex united
> system of function. Well, it just isn't that easy. At higher levels
> such a process would involve trillions of years of average time. I
> know this is a bummer for you, but that's the way it is.
> Statistically, there's just no way around it.

A single lucky mutation (such as a partial duplication or a chimeric
protein) does not require trillions of years. Once such a mutation
becomes fixed because it produces a useful function and that function
becomes optimized, it will *appear* like it took trillions of years.
But it doesn't. Selection for optimization occurs quite rapidly on a
geological timescale.

> > Cytochrome c did not have to reinvent
> > heme-binding sites for its new 'function'. It *borrowed* them from a
> > pre-existing protein (namely cytochrome c1). There is no evolutionary
> > event that I am aware of that NEEDS to to cross the multi-character
> > gaps you fantasize exist.
>
> Tell me how to get to each of the next steppingstones in flagellar
> evolution without any multi-character changes. It just can't be done.

Bloody hell. I started with two predcursor systems that had
independent utility. Neither of them had a motility *function*. These
two subsystems are linked by a single protein that binds and interacts
with both. That is a prima facia case in which one would propose a
chimeric protein that links together into a single protein the binding
components of two proteins that currently interact with the two
systems. Creating this chimera would introduce the *function* of
motility. The changes after that point would optimize the *function*
of motility (to the exclusion or down-grading of previous *functions*
like protein transport that the independent systems performed).

> After a while of honestly thinking about it one is simply forced to
> recognize the necessity of multi-character changes beyond very low
> levels.

I have honestly thought about it. And no, evolution does not and has
not involved crossing such large non-functional gaps in sequence space.
No serious biologist proposes that it did.

> They are there and they do indeed grow with increasing minimum
> requirements of size and/or specificity.

No they don't 'grow' with increasing size or specificity. When you
borrow moieties like heme-binding, you do not need to re-invent it.
And it is quite clear that only small features of large proteins are
actually involved in function. As proteins get larger, their
functional subparts almost always become a smaller percentage of the
total. This is evidenced by the rate of evolutionary sequence change.
Large proteins evolve faster than small ones in general, because large
proteins have proportionately more selectively neutral positions.

> >Your large gaps are a mathematical fantasy
> > that no one in his or her right mind would think actually happens.
>
> Not true. Show me where there are no multi-character gaps in flagellar
> models.

You show me where there are these multi-character gaps that must be
crossed through totally non-functional sequence space. Remember that
your claim is that there are thousands of changes that must occur
before there is *any* selectable function of *any* of the proteins
involved in flagellar *function*. Your claim is, then, that there are
no functional intermediates possible from the structures that form the
flagella. Yet you admit that the motor and the rotateable
protein-exporting pore have independent utility.

> > You
> > *hypothesize* that these large gaps exist because you have an argument
> > against *that* particular model of evolution. But you falsely claim
> > that that model is what *evolutionary biologists* claim as well. It is
> > not.
>
> I know it's not what popular biology claims, but popular biology is
> wrong. The multi-character gaps are indeed there. Prove me wrong.

Again, ebg for betagalactosidase. The obvious role of cytochrome c1 in
the evolution of cytochrome c. The experiments that show that a
currently existing protein required for flagellar utility can be
replaced, not by some sequence that started out as a random sequence
without function, but by a different protein that already had functions
that would be needed for a new activity. Show me a single protein that
did appear that you can demonstrate must evolve by crossing large
selectively non-functional gaps.

> > You have not demonstrated or presented a single *function* that
> > did evolve that *required* crossing your large gaps.
>
> That's because functions that do evolve are always at low levels of
> functional complexity (requiring no more than a few hundred fairly
> specified residues at minimum) where the ratio of beneficial sequences
> in sequence space is high enough to be fairly close to what is already
> in the gene pool library.

Finally we agree. Functions that *do* evolve always do so by a low
number of mutational changes that modify one functioning and useful
protein into a functioning and useful protein that has the new
function. In short, evolution does not work by ocean crossing that
drastically changes structure by many individually useless steps. It
works by island hopping from one related structure to a new nearby one
(or by combining two islands into one). New islands have new
properties, sometimes not drastically different (e.g., cytochrome c1 to
cytochrome c), sometimes modestly different (ebg to betagalactosidase),
some times more dramatic (motor plus pore to flagella).

> > Hell, I have, and
> > I am no genius, presented workable ideas about how a rotateable
> > flagella could arise from a single chimeric protein fusion mutation
>
> Not without significant multi-character changes involved you haven't.

A single chimeric protein fusion does not involve multi-character
changes. It involves change by a single event. That is not "multi" in
any sense. It is, in fact, the absolute minimum number of mutational
changes.

> > and
> > how cytochrome c could arise by a single partial duplication. Neither
> > *function* existed before these mutations; but it would exist
> > afterward.
>
> Cytochrome C is at a much lower level of complexity than is flagellar
> motilty.

Irrelevant. If cytochrome c had to evolve by your mechanism whereby
all 80 fairly specified residues had to arise simultaneously in a sytem
where a protein having 79 of the 80 was completely functionless, that
by itself would doom any evolutionary pathway. You need to demonstrate
that no combination of a protein having any 79 of these sites had any
selectable function.

> > > Cytochrome C, by the way, only requires a minimum of about 80 fairly
> > > specified residues.
> >
> > This number of 80 is only relevant in your mathematical fantasy that
> > cytochrome c arose by starting with a random sequence and proceding in
> > a non-functional random walk until cytchrome c *function* happens to be
> > landed on.
>
> Not true. Show me how you could make cytochrome c smaller or more
> flexible and still maintain its function to at least some minimally
> selectable degree. You see, it is a fact that the cytochrome c
> function requires a minimum size and specificity of arrangement before
> its function can be realized - even a little bit. This is not fantasy.
> It is fact.

Irrelevant. Cytochrome c did not arise from a random sequence. I
propose that it arose by partial duplication of cytochrome c1. That
may mean that the original cytochrome c may well have had *more* amino
acids than the modern one does.

> > In my model, the evoluton of cytochome c's unique
> > *function* does not require changing any of the amino acids involved in
> > heme-binding. That was borrowed from cytochrome c1. It probably does
> > not require changing any amino acids involved in interaction with
> > cytochrome c1. It may not require changing any amino acids involved in
> > the interaction with cytochrome a. It may require a few changes to
> > produce a selectable redox between that of cytochrome c1 and cytochrome
> > a. Once a selectable utility is obtained, of course, selection will
> > rapidly optimize it.
>
> This says nothing about the minimum size and/or specificity
> requirements of cytochrome c or the cytochrome island to which it
> belongs.

I have indeed discussed the specificities required for the specific
*function* of cytochrome c (to shuttle electrons between two other
cytochromes). It requires 1) the heme-binding function that is common
to many proteins, including proteins not involved in electron transport
(hemoglobin and myglobin, for example). *All* the heme-binding
proteins are related. And because heme has chemical activities that do
not require proteins but can be modified by proteins, this is a perfect
situation whereby any protein (regardless of its other functions) that
interacts with heme through an available histidine could be the
starting point for optimization of binding by selection leading to
proteins specialized for binding heme. And by duplication and
specialization, different types of heme-binding proteins performing
different functions can arise. That leaves the small patches of amino
acids that are involved in the affinity for cytochrome c1 and
cytochrome a. The electron transport is actually a function of the
heme rather than the protein. But this is hardly surprising. The
other compound that is involved in shuttling electrons between
complexes is not a protein at all. It is a quinone.

> Showing that CC is indeed very close to something else in the
> genome has nothing to do with its minimum requirements. It only has to
> do with suggesting that it may be evolvable given a certain
> pre-existing starting point of other cytochromes - which is true.

Well, duh. The evolvability of cytochrome c is what you are arguing
against! And here you are saying that the evolvability of cytochrome c
has nothing to do with its minimum requirements as you define it. I
agree. The minimum requirements you present has absolutely nothing to
do with the way that cytochrome c evolved. Your minimum requirements
are merely smoke and mirrors to try to claim that cytochrome c evolved
via a random walk from a random sequence.

> Getting to this cytochrome island cluster in the first place is another
> issue.

Yes it is another issue. It is another issue because you are talking
about the evolution of the *specific function* that cytochrome c serves
in the cell. That *specific function* did not arise from an ocean of
functionless sequence. It arose by a small hop from a nearby island
that already had most of the *structures* needed for cytochrome c
*function*, but did not have cytochrome c's specific *function* of
shuttling electrons between two other cytochromes in different
proton-pumping complexes. [Well, it is possible that cytochrome c1 may
have had some abillity to act like a cytochrome c, but if it did, it
was a poor substitute because it was bound in this large complex rather
than being freely motile.]

> But, again, CC is at a relatively low level of functional
> complexity. At higher levels the steppingstones get more and more
> widely spaced.

Do you have any real evidence for this assertion? How do you determine
"level of functional complexity" separate from "structural or sequence
complexity"? If you can't measure levels of *functional* complexity,
you have no point.

> > > Obviously then, it is not beyond my boundary of
> > > evolvability. Its fairly high degree of minimum specificity
> > > requirements certainly create sizable gaps around its island cluster of
> > > closely related sequences (to include other similar cytochromes in the
> > > ETC which are all on the same island cluster), but not sizable enough
> > > to stall evolution out given Darwinian time scales.
> >
> > Again, the number you present only makes sense if your model is that
> > all 80 sites have to evolve by a random walk of non-functional
> > intermediates.
>
> I never said that all 80 sites have to change by random walk. You
> continually assert that this is my position, but I've never said this
> and my statistics do not depend on this sort of assertion. This is
> just a strawman tactic on your part.

Your math indeed does always assume that all 80 sites have to evolve
without any selectable intermediates. That is the only way that you
can generate a number by taking 80 to some power as a measure of
probability. And even that completely ignores any distinction between
mutations that are absolutely necessary for a change in function and
those that optimize a function. Your calculations are GIGO unless you
assume what I am telling you that you assume.

> All I'm saying is that there is indeed a minimum requirement for such
> functions. For CC this minimum is about 80 fairly specified residues.
> Even at this relatively low level, there are gaps. The gaps are not
> nearly as large as they are at higher levels, but given the fairly high
> minimum specificity requirements of CC and the other cytochromes in the
> ETC, this island of closely related sequences does indeed have gaps
> between its own island cluster and all other sequences in the average
> gene pool. These multi-character gaps may only be a handful of changes
> wide and are therefore theoretically crossable in evolutionary time
> frames, but they are still there.
>
> > That argument *is* a strawman argument. It is a
> > mathematical fantasy that no evolutionary model even considers.
>
> You are arguing against a strawman version of my position of your own
> creation. My real model is very much in line with how evolution is
> supposed to work.

Your model is absolutly not how evolution is thought to work! It is a
mathematical abstraction that is seriously flawed in all aspects.

> > > <snip rest of strawman>
> >
> > I am most certainly NOT making a strawman of your argument.
>
> Yes, you are.
>
> > Your
> > argument, the math that you present, *claims* that the only way to
> > evolve cytochrome c is to start with a sequence that has NONE of the 80
> > "fairly specified residues" and proceeds, by a neutral random walk
> > through sequence space that has NO function, until it accidently lands
> > on the rare island of cytochrome c function.
>
> Not true. My model actually predicts that at lower levels of
> functional complexity, especially below levels involving a minimum of
> only a few hundred fairly specified residues, that many if not most of
> the required minimum will already exist pre-formed in a given gene
> pool. Out of the 80 required minimum 75 may already exist to the
> degree of specificity needed, leaving a gap of only 5. Such a gap is
> indeed crossable in evolutionary time frames.

This is the very first time I have ever seen you say this. But notice
that you are simply waving your hands wrt your numbers. Now a gap that
you have been claiming was 80 is now only 5. But I have presented a
mechanism by which cytochrome c as a selectable unique *function* can
arise in a single step, albeit maybe not at modern levels of function.
And there *can* be levels of cytochrome c *function*.

> However, when you move up to those functions that require a few
> thousand fairly specified residues, the gaps grow so that the average
> distance to the next closest beneficial island cluster is no longer 5,
> but 50.

Or, in specific cases, one. Or 5. What makes you think that size of
the *real* gap is related to the number of "fairly specified residues"?
If a protein has all the features needed to perform a function but
one, and that only requires a single change, what makes you think that
this changes with the size of the protein? But, again, I have never
heard you say this before. It is always 80 or thousands of changes,
with the math to indicate that.

> When the required multi-character difference is only 50 out of
> thousands, such a small percentage difference may not seem like much at
> all. It seems like such a small gap; so small as to be insignificant.
> But, it is actually devastating to evolutionary progress, requiring
> trillions upon trillions of years of average time for even a huge
> genetic pool to cross.

Or it would be, if the gap were that large. But you have presented no
evidence that the size of the gap increases exponentially with the size
of the protein or the number of proteins or whatever you call
"functional complexity". My point is that your math, which does assume
the larger numbers, is dead wrong. I am glad that you agree that you
should never have made calculations using those numbers.

> > Either you don't
> > understand the math you are using (that is, you are presenting an
> > argument without even understanding what you are saying) or you are not
> > being truthful. Being generous, I assume your ignorance over your
> > perfidity.
>
> Thanks ; )
>
> > And MY point is that such a *hypothetical* mathematical argument is
> > irrelevant to the way *real* evolution works. It is nothing but an
> > abstract mathematical exercise signifying nothing of importance or
> > relevance to real evolutionary pathways. I fully agree that *if*
> > evolution worked the way *you* claim it does, in *your* mathematical
> > descriptions, evolution would be next to impossible, and that includes
> > the evolution of cytochrome c. But your model is not the way that
> > evolution has ever been proposed to act. Your model is a strawman. It
> > is most certainly not descent with modification.
>
> Evolution has indeed been proposed to act using random mutation and
> function based selection over time. Evolution is supposed to change
> what is already available with tiny modifications over time to produced
> novel functions at higher and higher levels of functional complexity.

Yes. Notice that bit about the tiny modifications. More tiny
modifications are needed to produce more complex *structures*. But the
steps must be functional and selectable in their own right.

> That is how evolution is supposed to work. The problem with this
> mindless model is that at higher levels the neutral gaps do indeed
> outpace the ability of such mindless mechanisms to keep up.

So you claim.

> This
> stalling out of evolutionary progress is demonstrable in real life
> experiments. It's a fact.

It can be demonstrated that when you arrange conditions so that such
large gaps must be crossed they can't be. I agree. But the evolution
that *happened* is not the evolution that didn't *happen*. And the
evolution that *happened* did not have to cross these hypothetical
large gaps.
>
> > Cytochrome c 'function' did not arise by *your* proposed mechanism.
>
> If CC's function did evolve, it most certainly evolved via your
> proposed mechanism -which is the same as my proposed mechanism for how
> evolution is supposed to work.

No it is NOT the same mechanism. In my mechanism, cytochrome c evolved
its specific function by modification of a pre-existing cytochrome (c1,
specifically). You have NEVER, in any of your discussion of the
evolution of cytochrome c even mentioned this mechanism. Instead you
talk about 80 "fairly specified residues" changing purely by chance,
with math that indicates that there are no intermediate functional
states, until it winds up at the teleologic end point of modern
cytochrome c. That is an entirely different mechanism.

> We are proposing the same mechanism
> Howard. We really are. You're painting of my position is nothing more
> than a strawman - a caricature that isn't real. The problem is that
> this mechanism of yours, your own mechanism mind you, depends upon
> closely spaced steppingstones all the way up the ladder. It really
> does.

Absolutely. And for systems that *did* evolve, that is *how* they
evolved. Of course, as my examples point out, I consider mutational
events that are not always simple point mutations, such as duplication
and divergence, partial duplication, and chimera formations. Those are
the real mechanisms that produce most of the radical "new" functions.
Point mutations primarily change the function to the point of
optimization.

> These steppingstone pathways just aren't that close anymore as
> you move up the ladder. They really do get farther and farther apart
> so that the random walk times required to get from one steppingstone
> island to the next do indeed move into the trillions upon trillions of
> years beyond the few hundred fairly specified residue minimum.

You have no evidence to support this. And the fact that most molecular
systems clearly arise by modifications of pre-existing systems and make
wide use of independent protein moieties and changes that only involve
small parts of proteins, such as those involved in forming
heterodimers. Hell, the fact that so many different binding
*functions* can be generated by random mutation of only a few dozen aa
residues in immunoglobins says that one does not need to change
thousands of the aa residues in immunoglobins to produce proteins with
radically different specific *functions*.

> > No
> > one thinks it did. Cytochrome c probably arose by the mechanism I
> > described, and arose from a molecule that already had the heme-binding
> > moiety, already had the cytochrome c1- affinity, may have even already
> > had the cytochrome a affinity. And the mutation (as in single
> > mutational event; in this case a partial duplication) may have itself
> > produced a molecule with the right redox potential (not the *optimal*
> > potential of present cytochrome c, but a selectable and useful redox
> > potential). In short, I am telling you that your argument against the
> > model you propose as the way proteins evolve is correct. But it is
> > irrelevant because no one thinks that is the way evolution happened. It
> > is, IOW, a strawman argument. It is an argument that you, indeed, have
> > a counter to; if evolution worked or had to work that way, it would be
> > impossible. But evolution is not (and never was) thought to work that
> > way.
>
> The way you propose evolution to "work" only works if the
> steppingstones never get very far apart as you go up the ladder. But,
> they do start moving apart.

Evidence. Not based on a hypothetical. Based on a real protein that
did appear. What evidence do you have that that specific protein
evolved in a way that required crossing a large non-functional gap?

> The average random walk times do start to
> increase, exponentially, beyond very low levels of minimum size and/or
> specificity requirements. That's a big problem for *your* ToE.

I see no reason, in theory or even in hypothetical terms, to justify
your idea that the size of the random walks needed increase
substantially as a function of protein size or specificity
requirements. Do you have any actual evidence that it does? The size
of the gap is merely the number of changes in a pre-existing protein
needed to generate the new one. I see no reason, either from
experimental evidence of even as an abstract idea, why such gaps cannot
be small regardless of protein size or current function.
>
> Sean Pitman
> www.DetectingDesign.com

.

Relevant Pages

  • Re: Harshman, Felsenstein & Dembski
    ... >probability of the information content of a protein being generated by ... >probability is very low and the chance theory inadequate. ... that the odds against formation of a useful cytochrome is 10^-137. ... CytoC function is around 1e65 for 100aa sequence space". ...
    (talk.origins)
  • Re: The Starting Point Problem - for Howard Hershey
    ... where does Nature really come into ... there was the new technology of heritable protein sequences. ... cytochrome c to change its sequence of amino acids. ...
    (talk.origins)
  • Re: Sean Pitmans Misleading Statements Regarding Bias and
    ... misnaming "the number of effectively invariant sites in cytochrome c" ... as "average gap size" and then extending that measure of "sequence ... protein of the appropriate length (or "minimum structural threshold ...
    (talk.origins)
  • Re: The Theory of Evolution is a mathematically irrational belief system
    ... cytochrome c is the same protein ... in humans, flies, and yeast, even though its sequence differs. ...
    (talk.origins)
  • Re: Of Mice and Straw Men
    ... But you claim that it evolved from a random sequence by a random ... evolution will never make it off this island to any other at the same ... Functions like lactase, nylonase, cytochrome c, and other such ... The neutral gaps ...
    (talk.origins)