Re: Is "(Random Mutation) + (Natural Selection) => Complexity" Scientific?

On Jun 27, 4:05 pm, shlimazel <aashkin...@xxxxxxxxxxx> wrote:
Well, other than the claim that "we assume random mutation" (it is, after all
an observation, not an assumption), I'll proceed with this as a premise.

Interesting. Aside from alleles that produce minor differences such as
variation in eye color or nose size, what do you see, after the fact,
in cooperative proteins of a cascade or hormonal pathway that is the
hallmark of random mutation rather than, say, some unknown
deterministic event such as self-organization?

First, the generation of new cascades or hormonal pathways are
evolutionarily rare events. The vast, vast, vast, vast, majority of
speciation or evolutionary events involve relatively minor changes in
the rates and timing of gene activity rather than changes that produce
new 'enzymes' or even new 'proteins'. This is the type of variation
that *exists* via random mutation (and there have been any number of
experiments that demonstrate that mutation is random wrt need) in most
populations and allowed us to exploit it to produce all the
domesticated animals we have, including differences as great as that
betwen a chihuahua and a Great Dane, in a geological millisecond.

Second, new cascades are typically obtained by duplication and
divergence. So are 'new' proteins like the beta globins of hemoglobin
(all the variants are initially from a duplication and divergence of
an ancestral alpha globin). Another good example is the aldosterone
receptor that arose before aldosterone existed by two changes in aa in
a duplicate. None of these 'new' functions requires inventing
*anything* from scratch or some random sequence.

Ah. Well, explanations should be as simple as possible, but in this attempt
you've rather stretched the limit of possibility. While the mutations
that produce genes may be random, their accumulation in the genome of a
population certainly is not.

Not so. There will always be two independent factors to each
probability: the probability that the random mutation occurs in the
first place, and the probability that it would then become fixed in
the population. Even for beneficial mutations, there can be chance
accidents, mating sampling errors, etc., that stop propagation. There
is no certainty that all beneficial mutations become fixed. But if you
disagree, assign the fixation probability factor a value of unity.

The probability of a beneficial mutation becoming fixed is a function
of its relative benefit. The probability of a selectively neutral
mutation becoming fixed is a function of its frequency in the
population. Yet the vast majority of change in genomes is due to
neutral mutation fixation. Selection is primarily a conservative
force *reducing* the rate of change in genomes. But when selection
favors change, because the probability of fixation is a function of
relative benefit, such change will be extremely rapid relative to
neutral drift. Again, the *evidence* clearly indicates that there is
more than enough existing variation in populations to allow the
observed rate of morphological change. Change involving new
'biochemistry' is more idiosyncratic in nature and is the rarest form
of evolution. In essentially all cases, such new biochemistry arises
by modification of (often a duplicate) pre-existing protein that has
some structural relationship to the new function.

An event with such a small probability is unlikely to have occurred!
So the assumption that random mutation is the driving mechanism for
variation is suspect.

No. It isn't. It's observed.

Again, how? See question above.

Do you have evidence that the rare creation of new 'biochemistry' like
the aldosterone receptor or beta globin is somehow beyond the capacity
of random mutation to create? For most evolutionary change, the
existence of dog breeds tells you about how much variation exists in
populations that would be exploitable in tiny fractions of geological
time. Also, the relatively minor (and mostly due to neutral drift)
differences in DNA sequence between chimpanzees and humans points to
only a small need for random mutation that would not be included and
already exist in the 'normal variation' in populations.

If evolution were entirely driven by mutation, you might have some point
(the argument would still be too simple to be useful, but it would be closer).
You've neglected entirely the role of selection in populations.

From the above, you should now know that I didn't mean that evolution
is entirely driven by mutation, since we discussed inclusion of
fixation probability above. I guess I should have made that clearer in
my earlier post, since someone else also thought I was excluding
natural selection.

Your evidence that the observed rate of change in genomes is
impossible given the rate of mutation is...?? In fact, the rate of
change in genomes is actually *less* than would be accounted for by
neutral drift. That is because most selection is necessarily
conservative.

Alias shlimazel


.

Relevant Pages