GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2003-11 > 1068330035


From:
Subject: Re: [DNA] 25/25 marker Y-DNA test results match example with very different looking surnames -- some possibilities
Date: Sat, 8 Nov 2003 17:20:42 -0500 (EST)
References: <3FAC735B.7050904@kerchner.com>
In-Reply-To: <3FAC735B.7050904@kerchner.com> (message from Charles on Fri, 07Nov 2003 23:38:51 -0500)


Charles wrote:
> 1. Mr. Zimmerman and Mr. Carpenter share a common male ancestor which
> predates the use of surnames. The Y chromosome on all 25 markers has
> remained stable and unchanged on all the 25 markers tested for many
> hundreds of years. This is certainly possible. A question for the
> statisticians here. What are the odds of this happening for two direct
> male line descendants of a common male ancestor who lived 1000 years
> ago, assuming 20 years per generation? What are the odds of a change not
> occurring in the direct male descendant lines over a time period of 1000
> years?

Well, for the sake of realism, as well as convenience, 25 years per
generation is a better number to use. The probability of zero
mutations in that time (for 25 markers, with an average mutation rate
of 0.002) is about 2%.

> 2. Mr. Zimmerman and Mr. Carpenter share the same Y chromosome markers
> because over the last 1000 years their respective ancestors who
> originally had 5 differences in their Y chromosome markers have mutated
> in such a way that their Y chromosome haplotypes converged to the same
> set of Y markers. This could happen. Which leads to another question for
> the statisticians here. With what we know about the mutation process
> over time, is it possible to compute the odds of this happening?

This question is not a theoretical one, like the first question. The
answer would depend a great deal on the detailed distributions of all
the haplotypes within a few steps of the final outcome, as well as the
assumptions you might want to make about where these ancestors lived
(which would be reflected in the detailed distributions, of course).
If the areas in question have had major population shifts in the past
1000 years, then it's ultimately impossible to make even a decent stab
at an answer.

John Chandler


This thread: