GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2007-07 > 1184231637


From: "Elizabeth O'Donoghue" <>
Subject: [DNA] Bottlenecks - R1b1c7 as a calibration tool - The " final "version
Date: Thu, 12 Jul 2007 10:13:59 +0100
In-Reply-To: <REME20070711190523@alum.mit.edu>


Elizabeth wrote:
> John, if this is the case, what constitutes a 'bottleneck'? I've always
> gotten the impression that rates need to take into account possible
> bottlenecks in a population when calculating TMRCA, hence using father-son
> stats are inadequate for considering back, say 2000 years.

John C wrote:
'A true bottleneck occurs at the MRCA of the existing population. This
isn't limited to cases where all members of the population but one
died out at some point in the past -- you get the same effect if all
other lines die out or daughter out at various later times, leaving
only the one line to survive to the present. In other words, you need
to consider more than just the disasters in correcting the simple
model of linear increase in the variance. Another way of putting it
is that the MRCA is almost surely not the same person as the founder
of the population in question. Unfortunately, if you don't already
know the details of the population history well enough to figure out
who, when, and where the MRCA was, you probably don't know enough to
calculate the corrections, either.'

Ken N wrote:
'So called bottlenecks, if they randomly hit the population, do very little
to population variance or permit much more "genetic drift" until the
population gets knocked down to single digits.'
------------
Alas, I feel confused again. A true bottleneck could have happened any time
in a particular family line, surely. But in a larger population, such as
these Irish Gaels I'm always talking about, or the Irish Type III, or the
NWIMH, a particular bottleneck or two would not really affect the overall
"genetic drift" of the group as a whole, if they were to have arrived in
Ireland 2000-3000 years ago +/-, would it? So what other factors would make
it necessary to "fudge" from father-son mutation rates to arrive at a likely
TMRCA?

Thanks very much. Elizabeth

This thread: