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

Nic wrote:

> John Harshman wrote:
>
>>Nic wrote:
>>
>>
>>>John Harshman wrote:
>>>
>>>
>>>>Nic wrote:
>>>>
>>>>
>>>>
>>>>>More drift than I thought
>>>>>John Harshman wrote:
>>>>>
>>>>>
>>>>>
>>>>>>Larry Moran wrote:
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>>On Fri, 2 Dec 2005 09:41:22 -0800, anon1@xxxxxxx <anon1@xxxxxxx> wrote:
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>>>The mechanism is called random genetic drift. The affected alleles
>>>>>>>>>may be beneficial, neutral, or detrimental with respect to natural
>>>>>>>>>selection. The frequencies of all three types of allele can be
>>>>>>>>>influenced by random genetic drift so it's not appropriate to
>>>>>>>>>refer to the mechanism as "neutral drift."
>>>>>>>>
>>>>>>>>Ah, yes, I was thinking of which factor dominates. If there's no
>>>>>>>>selection, drift dominates, and it's truly neutral, whereas if there's
>>>>>>>>lots of selection, selection dominates, so you can mostly ignore drift.
>>>>>>>>But to handle the between cases, where there's selection but there's
>>>>>>>>also drift of comparable amount, so you must not ignore either, your
>>>>>>>>point is well taken.
>>>>>>>>
>>>>>>>>By the way, it seems to me that random genetic drift due to sampling
>>>>>>>>error happens only during meiosis, so it doesn't apply to asexually
>>>>>>>>reproducing cells. Is that basically correct?
>>>>>>>
>>>>>>>
>>>>>>>No, ... not even close.
>>>>>>>
>>>>>>>Imagine a population of one million single-cell organisms that divide
>>>>>>>by binary fission. This could be mitosis if the cells are eukaryotic.
>>>>>>>The population size doesn't change from generation to generation so
>>>>>>>half of the daughter cells die before reproducing. Sometimes both
>>>>>>>daughter cells die and the lineage comes to an end. Sometimes both
>>>>>>>survive and the allele frequency of that lineage increases.
>>>>>>>
>>>>>>>Allele frequencies in the population change over time due to the
>>>>>>>random survival of the daughter cells. That's random genetic drift.
>>>>>>>The founder effect is another example of random genetic drift.
>>>>>>>
>>>>>>>The same effect applies to sexually reproducing organisms. Some
>>>>>>>die by chance and some survive by chance. The classic textbook case
>>>>>>>is flowers that are killed in a mudslide. Other good examples are
>>>>>>>squirrels that are run over by a car and innocent Middle Eastern
>>>>>>>citizens who are blown up by suicide bombers.
>>>>>
>>>>>-
>>>>>
>>>>>
>>>>>
>>>>>>How about this as a definition of genetic drift?: differential
>>>>>>reproductive success not correlated with genotype.
>>>>>
>>>>>
>>>>>Maybe, but I think there is an enormous amount of drift even if all
>>>>>misadventure could be eliminated say, by having a guardian angel
>>>>>assigned to each organism.
>
> -
>
>>>>>Differential reproductive success due to what you've got at one locus
>>>>>is random success for what you've got at all the other loci. Good
>>>>>allele A gets killed because time is up for bad allele B at some other
>>>>>locus. Probably happens a lot more often than mudslides.
>
> -
>
>>>>True if you consider only one locus at a time. But we realize that
>>>>selection happens to entire organisms, and fitness is a term that really
>>>>applies to entire genomes. In one-locus models, the implicit assumption
>>>>is that the genetic background is otherwise uniform in the population.
>>>>Multi-locus models define the fitnesses of genotypes, even if they are
>>>>only the sum of allelic fitnesses. That's not drift.
>>>
> -
>
>>>It's not drift, or it's no drift?
>>>What I mean by that question is, where there is selection, you can say
>>>there is no drift, but I think that is just a way of saying the drift
>>>component averages out to zero.
>>
>>No, not true. Selection and drift are two different processes that can
>>operate simultaneously.
>
> Apologies for the confusion. Where you said "That's not drift." I
> understood what you meant. That's the big picture, and it isn't drift
> in anybody's book - it's natural selection. The small picture is of
> allele A's evolution suffering a set back, even though selection for it
> will prevail in the long run. I was trying to make the point that one
> *could* call that a case of drift in opposition to selective pressure.
> Further below, you agree it's OK to talk like that.

No, I don't think I do. You are talking here about opposing selection
pressures. It's not drift. Nobody says its drift, or treats it as drift.

>>Of course any drift component always averages
>>out to zero, since there is no force pushing it in any direction. But of
>>course, also, no single case is average, and an allele subject only to
>>drift has only two possible fates: extinction or fixation, with the
>>relative probabilitie of each at any moment equal to 1-p and p,
>>respectively, where p is the current frequency.
>>
>>
>>>I thought the point under discussion
>>>here is whether the averages taken by nature (over population, and over
>>>species life time) are such that we can consider that drift averages
>>>out to zero in all but the (near enough) completely unselected cases.
>>
>>I can make little sense of this.
>
> Again, my fault for the confusion. The above was intended to mean no
> more than what I put in my next paragraph.
>
>>Drift and selection operate
>>simultaneously to affect the probability of fixation or extinction of
>>alleles. Whether drift or selection is the most important parameter in
>>any given case depends on both the selective value of the alleles and
>>the size of the population.
>>
>>
>>>I think it is valid to ask whether drift in typical population sizes,
>>>over typical species life times, is capable of overcoming weak to
>>>moderate selective pressure.
>>
>>Yes, it is, depending on your particular values of population size and
>>"weak to moderate". A text on population genetics would provide you with
>>some nice equations that would tell you the exact relationships among
>>these parameters.
>
> What if your drift definition was changed from: differential
> reproductive success not correlated with genotype.
> To: differential allelic reproductive success not correlated with that
> allele?

Then you would have a definition of drift that everyone else would
consider selection, and you would need different equations, which you
could also find in a population genetics text. Are the loci unlinked?
What is the nature of their interaction?

> How prevalent are neutral polymorphisms which are currently drifting?

Most sites in the human genome are evolving neutrally.

> I had always assumed they would be quite rare and short lived simply
> because the so called molecular clock seems to work, and that wouldn't
> be much use if things were taking 10s of millions of years to become
> fixed or go extinct (or were taking a length of time which is a strong
> function of population). I now think my reasoning may be flawed in
> this.

Indeed it is. If there's a molecular clock, it works precisely because
of drift. However, it has a huge variance.

.

Relevant Pages

  • Re: The Cost of Substitution [possible REPOST]
    ... I.e. the time to fixation under even rather weak selection will typically be ... just as it does under drift. ... If we know that an allele will fix then under a neutral ...
    (sci.bio.evolution)
  • Re: The Cost of Substitution [possible REPOST]
    ... I just said that your point about fixation taking ... longer in larger populations under drift also applies to selection. ... Drift can lead to the allele sometimes not reaching fixation at ...
    (sci.bio.evolution)
  • Re: Standards of Evidence
    ... >>treats the external environment as constant - selection occurs, for instance, ... how diversity is affected by environment size. ... Or it may be that one speciation occurs we ... >>coefficient or ordinary drift changes a population, ...
    (talk.origins)
  • Re: The Cost of Substitution [possible REPOST]
    ... directional selection. ... I just said that your point about fixation ... taking longer in larger populations under drift also applies to ... Drift can lead to the allele sometimes not reaching fixation at ...
    (sci.bio.evolution)
  • Re: A critique of the BBC aquatic ape programme and the transcript.
    ... >> efficiency differences were broadly due to selection and that swimming ... If drift is the ... > as evidence. ... Whether or not they've received proper criticism ...
    (sci.anthropology.paleo)