Re: Back in the good old days.

Perplexed in Peoria wrote:
"John Harshman" <jharshman.diespamdie@xxxxxxxxxxx> wrote:
Perplexed in Peoria wrote:
"John Harshman" <jharshman.diespamdie@xxxxxxxxxxx> wrote:
Perplexed in Peoria wrote:
"John Harshman" <jharshman.diespamdie@xxxxxxxxxxx> wrote:
Perplexed in Peoria wrote:
"John Harshman" <jharshman.diespamdie@xxxxxxxxxxx> wrote:
Perplexed in Peoria wrote:
"spintronic" <spintronic@xxxxxxxxxxx> wrote:
when spin argued against there being "Junk D.N.A".

There were a lot of big mouths on T.O claiming 95% of our genome was
Junk.

http://www.sciencedaily.com/releases/2008/11/081104180928.htm
Yep, there were, and I was one of them. This is a very interesting
article, Spin. Thx for posting the link. Here is the abstract
<http://genome.cshlp.org/content/18/11/1752>
but unfortunately the article itself requires a subscription.

I'm not sure whether this article really supports the belief that
a lot of that "95% junk" actually has a function. In discussing this, it is important to first answer the question: "Cui bono?" Or,
"'Functioning' in whose interest?" Because what the "95%-ers"
like me have been saying all along is that all that 'junk' probably
doesn't serve a function for the organism that carries it. But
we have always conceded that much of the 'junk' is likely to
be 'selfish DNA' - which does nothing positive for the organism
but does function for the 'benefit' of the DNA sequence itself.
Transposons, for example.

But this article is suggesting that some of the junk has a function
of a different kind - a function benefiting a different entity. The
article claims that many binding sites for transcription factors
got to where they are today by riding on transposons. This
is interesting. Because people like me readily agree that the
binding sites of the transcription factors are functional for the
organism, and therefore not 'junk'. But we have been arguing
that the transposons are functional only for themselves, and
hence are junk. The article, however, argues that the transposons,
by transporting the transcription factor binding sites, have been
acting to accelerate evolution in various lines of mammals, and
hence that they have a function beyond their own selfish interests.

Well, yes they do, but if you look carefully, it cannot be called
a function which benefits the organism. The acceleration of
evolution is something which only shows a benefit in the
long term, and as Keynes tells us, in the long term the organism
is dead. It seems that the beneficiary of those 'junk' transposons
would have to be an entity with long term interests. I would say
that the beneficiary is the species (or deme). Conceivably, though,
the beneficiary might be some kind of Author of the evolutionary
process. The "Author" has a purpose in "Mind" for evolution -
and arguably so does the species. But the individual organism
doesn't 'want' to evolve, and, in fact, CAN'T evolve. So these
evolution-enhancing transposons simply cannot be said to serve
a function for the benefit of the organism. Therefore, I would say, they are still 'junk'.

Put more succintly, certain random mutations (of which transposon insertions are one class) are occasionally beneficial, and so subject to fixation by selection. If they benefit the organisms that contain them, that is. I don't see where one gets an author (big or little a), or even a benefit to the species, out of that.
Imagine you have a gene which (depending on which allele) turns
a particular class of mutations on or off. Is the 'on' allele beneficial
to the organism, and will it therefore increase in frequency in the
population?

Well, the answer to that question is going to depend on the properties of that class of mutations. So, just so we have some
definite numbers, assume that the 'on' allele allows 1 mutation
per generation, and that the mutations are 90% neutral, 6%
detrimental (s=0.1) and 4% beneficial (also with s=0.1 (or should I say -0.1?) ).

So, it would seem that the 'on' allele does not benefit the individual
(on average) so it should be purged from the population. And
simple models which ignore genetic linkage endorse this conclusion.
However, it will occasionally happen that the 'on' allele will be
closely linked with one of the beneficial mutations which it permits.
This complicates the analysis.

AIUI, the result is going to depend on the details of the population
structure. It is a kind of group-selection problem. The 'on' allele is an altruism gene. It produces benefits which eventually spread
to the entire population, but the carrier of this gene suffers detrimental
consequences which more than compensate.

Regardless of whether you yourself have the 'on' allele, it is to your
advantage if your neighbors have that alleles. Because if your
neighbor is unlucky and transmits a detrimental mutation to his
offspring, that offspring likely dies in childhood, doing you no
harm. But if your neighbor is lucky and transmits a beneficial
mutation to his offspring, that offspring reaches maturity and
possibly breeds with your own children.

To say it another way, in a sexual population, you want low
fitness in your pre-selection neighbors - because they are
your competitors. But you want high fitness in your post-selection
neighbors - because your grandchildren will share those genes.

So how to get low fitness pre-selection, but high fitness post-selection? By having high variance in fitness (pre-selection).
And that is just what the 'on' allele accomplishes. It benefits
the population, though it slightly harms the individuals who
bear it.

This analysis all seems dubious to me. Won't your children be in the same position vs. your neighbor's children as you were vs. you neighbors?
Yes. Why is that a problem?

You have the group selection changing directions between generations, don't you? I don't see any group selection argument here that would promote fixation of "on".
How is selection changing direction? I just said (agreeing with you, I thought) that the situation is the same in every generation.

As to the group selection argument, I can use any of the standard ones.
Maynard Smith's haystacks, Hamilton's rule, the Price equation, the
Sober-Wilson trait group selection model. The only novelty here is
my claim and argument that my hypothetical "on" allele is effectively
an "altruistic gene".
If all this is so, I really don't understand your scenario. I don't understand the difference between pre-selection and post-selection. I had supposed it meant you want your competitors to have low fitness and your children's competitors to have high fitness. But that would be a change between generations.

'Preselection' and 'postselection' refer to two different points in time within the same generation. It is easiest to understand in a model with non-overlapping generations. Start with preselection, just after
the new generation is born. There is Hardy-Weinberg equilibrium
at this point in time. But then selection takes place, killing off the
unfit, and also, incidentally disturbing the H-W equilibrium. Now
you are at the post-selection stage. Still the same generation, since
no mating has taken place yet, but the gene frequencies in the
population have changed. Next you have mating and reproduction
and a new generation (which is back at H-W equilibrium and
in the next pre-selection stage).

You want your pre-selection neighbors to be low fitness because
you compete with them during the upcoming selective phase.
But once selection is done, if you were lucky enough to survive,
you want your co-survivors to be high fitness (higher than yourself,
if possible). Because those folks are about to become your mates.

Then, once your children are born, they will be back in the
pre-selection state, ready for their chance to compete with
their peers.

Your post-selection potential mates are some of the same folks
who were your selective-phase competitors. But a subset.
Because some of your competitors are dead.

I believe I now understand at least part of your argument. It's always dangerous to talk about what "you want" in evolution, because it risks confusion. I think there's confusion here. Apparently you have a model in which there is panmixis within demes but considerable structure among demes. In terms of individual selection, it doesn't matter what your mate choices are, and it doesn't matter if the deme as a whole is increasing in fitness. Deme fitness isn't even defined. What matters is your representation in the next generation. The individuals with the greatest fitness will make the greatest contribution to the next generation. That's all. The pool of mates is irrelevant. It seems to me that there is no increase in your fitness if other members of your deme have favorable mutations; there is in fact a decrease, because their contributions to the next generation are likely to be larger than yours. There may be some increase in deme fitness, if we manage to define the term, but we're back to group selection. This won't explain why "on" would increase in frequency in a deme, only why demes in which "on" happens to become fixed might prosper at the expense of others.

Nor do I understand the connection to epigenetic inheritance.

It has absolutely nothing to do with Cabej and epigenetics. It had
a marginal relationship with Spintronic and hologenomics (which
is the new slogan for the folks who claim that junk DNA is functional).

I sympathise with your confusion - it all seems to run together for
me too.

Yes, I noticed that right after posting. So what does this have to do with whatever Spintronic was trying to say?

.

Relevant Pages

  • Re: Back in the good old days.
    ... lot of that "95% junk" actually has a function. ... certain random mutations (of which transposon ... to fixation by selection. ... Is the 'on' allele beneficial ...
    (talk.origins)
  • Re: Back in the good old days.
    ... lot of that "95% junk" actually has a function. ... serve a function for the organism that carries it. ... to fixation by selection. ... Is the 'on' allele beneficial ...
    (talk.origins)
  • RE: Fundamental theorems, dilemmas, fitness, and information.
    ... > Is there a limit to the rate at which natural selection can ... > increase fitness? ... > allele fixation requires "selective deaths". ... cost of substitution. ...
    (sci.bio.evolution)
  • Re: Back in the good old days.
    ... a lot of that "95% junk" actually has a function. ... Imagine you have a gene which (depending on which allele) turns ... But you want high fitness in your post-selection ... You have the group selection changing directions between generations, ...
    (talk.origins)
  • A Question Of Integrity
    ... >>and that can't be right because selection works at multiple levels. ... Would Prof. Felsenstein please provide ... model of c (Darwinian fitness) where c contains ... the allele with the most pleiotropic effects ...
    (sci.bio.evolution)
Loading