GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2005-09 > 1125704180


From: (John Chandler)
Subject: Re: [DNA] I1a Buchanan 'match' on Sorensens. Mutation possibilties?
Date: Fri, 2 Sep 2005 19:36:20 -0400 (EDT)
References: <20050902213945.2123.qmail@web81805.mail.mud.yahoo.com>
In-Reply-To: <20050902213945.2123.qmail@web81805.mail.mud.yahoo.com> (messagefrom Robert Stafford on Fri, 2 Sep 2005 14:39:45 -0700 (PDT))


Bob wrote:

> It is best to use the "forward" part of the TMRCA calculation to
> test scenarios, such as you presented. It is really all most people
> need to use. Ann Turner has provided a calculator for this.
>
> http://members.aol.com/dnafiler/MutationCalculator.exe
>
> With a 23/28 match, 36 transmission events (3 generations per
> century, doubled for the two legs) and a rate of .0023, I get a 5%
> probability that the two have a common ancestor in 600 years. Not
> particularly high, but still a possibility.

I think what you meant to say is that with 36 transmissions on 28 loci
with a mutation rate of 0.0023, you get a 5% probability of exactly
5 mutations, neither more nor fewer. It would be better to calculate
the probability of getting five *or* *more* mutations if (as in this
discussion) the question was whether five mutations might be too many
to expect. That's 9%.

> Testing the additional 11 markers FTDNA might provide some
> additional information. A 34/39 match would raise the probabilty to
> 11.6%.

Probability of N mutations in 39 loci over 18 generations for two lines
assumed rate=0.0023
0 0.03958
1 0.12784
2 0.20641
3 0.22218
4 0.17936
5 0.11584
6 0.06234

In other words, the probability of 5 or more mutations is 22%.

John Chandler

This thread: