Re: Part 1 (of 3): What are major aspects of evolutionary theory?

anon1@xxxxxxx wrote:

> Regarding my explanation of probability of running off the fixation or
> extinction ends of an allelle population frequency axis, and what
> happens to the mean near the end, etc.:
>
>>why the repetition of basics?
>
>
> Because somebody had basically accused me of not understanding the
> basics, so I had to explain what I thought the basics were so you could
> check if I had it right. By my spending 20 minutes writing up the
> explanation, so you could spend 1 minutes reading it, instead of you
> spending 20 minutes telling me the basics I already knew, I saved you
> 19 minutes. :-)
>
>
>>Once again you use the term "group selection" in a non-standard way.
>
>
> I suppose the standard way really means group-of-whole-animals selection,
> or group-of-whole-single-celled-critters selection, right?
> So what should I call:
> - Group-of-DNA-bases selection (trapped together in gene)
> - Group-of-genes selection (trapped together in linkage group)
> - Group-of-chemicals selection (trapped together in abiogenesis proto-cell)
> - More generally, group-of-anything-doesn't-matter-what selection?

I don't know. Just don't call it group selection, because that term is
already taken.

>>>>Or, to put it more reasonably, with present technology it's impossible
>>>>to tell whether the allele is being selected or is hitchhiking.
>>>
>>>Correct.
>>
>>Not actually correct; with present technology we can manipulate alleles
>>so as to separate pieces of haplotype blocks.
>
> What do you mean "manipulate"? Read genome in each daughter cell,
> without killing it, to decide whether it has the exact allelle you want
> to keep or not, or introduce something that attaches to correct allelle
> and protects it from some cleavage agent you introduce next, or what?
>
>
>>Group selection refers to selection operating directly between
>>populations, above the level of individuals. It has nothing to do with
>>summed selection on linked alleles.
>
>
> (See my question above what to call selection that acts on other groups
> of some things.)
>
>
>>>>>we might directly observe deaths of
>>>>>individual cells/organisms, perform autopsies on each to determine
>>>>>cause of death, and thereby attribute some allelle of some locus at
>>>>>fault for the death, and by compiling statistics of such
>>>>>causes-of-death we might then calculate selection pressure for each
>>>>>non-neutral locus. But that's not yet feasible, ...
>>>>
>>>>Actually, similar sorts of studies are quite feasible with current
>>>>technology.
>>>
>>>Wow, that's news to me. Where'd you read about it? I haven't yet seen
>>>anything in _Science_ that implies such technology exists yet.
>>
>>I wasn't talking about the cute little example of microbe autopsies,
>>here, but about experiments that separate the effects of single site
>>variants.
>
> If two different single-site variants result in the same visible effect
> in phenotype, how do you separate them?

By sequencing? We can assume that members of a clone have the same
variant. We can select inbred lines to have any variant we like.
Site-directed mutagenesis can change particular bases at particular sites.

>>As I recall, our disagreement began when you listed three lines of
>>evidence for common descent, and then used them as if they were three
>>lines of evidence for natural selection. That is in fact a disagreement
>>about science.
>
> The same evidence has two effects. First, it shows blatantly that
> descent with modification happened, and that species-splits happened.
> At that point the cause of all that is unexplained. Second, somebody
> contrives a hypothesis that might explain those happenings, and it's
> tested to see whether it is compatible with all the data. That's where
> natural selection comes in. So my check-list is to see how far along
> this line of evidence and theory somebody is willing to agree. Do you
> agree that neo-Darwinisn is the only theory that satisfies all the
> evidence, or do you claim that some completely different theory would
> also explain it, such as committees of angels?

Not clear what you mean here by "neo-Darwinism". It's difficult to show
past natural selection. It can be done in some cases, but not in the
bulk of cases. We have no other current explanation, but the evidence
that natural selection caused birds to grow wings, for example, just
isn't there.

>>>where there are
>>>haplotype blocks averaging 10k bases long that haven't been split since
>>>the pre-human / human bottleneck.
>>
>>So you've found the average length. Note that this is about half the
>>length of the average human gene, so that recombination between and even
>>within loci is not particularly affected by these blocks.
>
> The block structure per se I agree doesn't prohibit recombination
> between adjacent genes and sometimes within the same gene.

In fact it's pretty much required by that block structure, isn't it?

> The word
> "locus" (plural loci) is too vaguely defined to know what you mean by
> "within loci". Is a "locus" a single location, i.e. a single DNA base,
> just that we don't usually measure it that accurately? Or is a "locus"
> a span of thousands of DNA bases all a single "locus"?

That's a silly question. If a locus were a single base, recombination
within it would be impossible. It should be clear that by "locus", here,
I mean "gene".

> However adjacent haplotype blocks might recombine only very rarely,
> like once in a thousand years, so that after 400 thousand years (since
> the population bottleneck that reset the haplotype clock in humans)
> each block-block boundary has recombined at least once, but over a span
> of a few hundred years the linkage might be so strong between adjacent
> haplotype blocks that they are 95% co-evolving all that time and only
> 5% chance of a single recombination the whole span of time.
>
> So from generation to generation there might not be any significant
> recombination within a single gene, and even if all recombinations
> within different allelles of a single gene were immediately fatal it
> wouldn't cause significant loss of reproductive capability for one
> parent who had both allelles and hence whose children each had a
> fraction of a percent chance of the fatal recombination.
> .

What you need here is some reason why recombination within genes should
be rarer than recombination between genes. Since, on the average, at
least half of block boundaries must not coincide with gene boundaries
(because of the relative sizes), that would seem hard to explain.

.

Relevant Pages

  • Re: Part 1 (of 3): What are major aspects of evolutionary theory?
    ... I suppose the standard way really means group-of-whole-animals selection, ... to decide whether it has the exact allelle you want ... >> haplotype blocks averaging 10k bases long that haven't been split since ... The block structure per se I agree doesn't prohibit recombination ...
    (talk.origins)
  • Re: Part 1 (of 3): What are major aspects of evolutionary theory?
    ... > I'm not sure when exactly the _Science_ summary appeared, ... > inventory of where recombination takes place. ... If natural selection is still working ... > site on *every* allelle that block totally neutral, ...
    (talk.origins)
  • Re: Part 1 (of 3): What are major aspects of evolutionary theory?
    ... we attempt to detect selection by looking at nucleotide diversity ... > haplotype blocks averaging 10k bases long that haven't been split since ... length of the average human gene, so that recombination between and even ...
    (talk.origins)
  • Re: Part 1 (of 3): What are major aspects of evolutionary theory?
    ... I'm not sure when exactly the _Science_ summary appeared, ... inventory of where recombination takes place. ... If natural selection is still working ... site on *every* allelle that block totally neutral, ...
    (talk.origins)
  • Re: The Theory of Evolution is a mathematically irrational belief
    ... google the following terms HIV and recombination. ... Because in the environment created by drug cocktails, no one mutation is ... The question about mutation and selection are there any selection ... And selection on one gene can potentially affect many genetically correlated traits. ...
    (talk.origins)