GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2007-09 > 1188738002


From: <>
Subject: [DNA] genetic drift
Date: Sun, 2 Sep 2007 09:00:02 -0400
References: <ca5.17e1ed33.3409fa2f@aol.com> <000d01c7ec7b$babc8cc0$6401a8c0@HP><000901c7ec9e$cefca440$6400a8c0@Ken1>
In-Reply-To: <000901c7ec9e$cefca440$6400a8c0@Ken1>


> -----Original Message-----
> From: On Behalf Of Ken Nordtvedt
> Sent: Saturday, September 01, 2007 9:49 AM
> To:
> Subject: Re: [DNA] Haplogroup I in Europe
>
>
> ----- Original Message -----
> From: <>
>
> > Bottom line: even without selective pressure, the number of alleles will
tend
> > to be reduced by chance extinctions because a lost allele cannot come back
> > into play.
>
> And new alleles are being produced by mutations. It's a bit more
> complicated.

Yes, of course there are new mutations, but they're rare, especially in small
populations, so unless there is a selective advantage to them, their low
frequency makes them likely to be lost. In small populations, the tendency is
for most alleles to move to a frequency of either 1 or 0 in the population.
This isn't simply my opinion, it's basic population genetics. The equations are
out there, you just have to Google the topic. In fact, the entry at Wikipedia
is pretty good:

http://en.wikipedia.org/wiki/Genetic_drift

essentially making the same point I did about how the tossing of a coin (or die)
is not a good analogy for allele frequencies (viz., "allele frequencies in real
populations are not probability distributions; rather, they are a random
sample..."). The coin maintains its original probabilities for every set of
tosses because the coin is not changing. But in genetics, the starting point is
changing with each generation, which is why I said, to be a valid analogy, you
would have to whack a side off the die each time a number failed to appear in a
set of throws, such that, eventually, you'd end up tossing a one-sided die.

Even a population that has all alleles equally frequent in Generation I will not
have them equal in Generation II, and the frequencies become less equal with
each subsequent generation, until all alleles are lost, but one. The best
opportunity for an increase in variability is for two or more formerly isolated
populations to merge and/or for the population size to increase, so that new
mutations and rare alleles are not so easily lost. The human population is in
this condition now: formerly isolated populations coming into contact, with an
exploding population, essentially bringing genetic drift to a halt, except in
the dwindling number of small populations that remain isolated.
>
>
> The smaller the population, the more pronounced drift will be.
>
>

Yes, thats what I said: "In large populations, the effect is virtually nil,
but... the effect is highly significant in small populations..."

Diana




This thread: