GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2005-07 > 1120242128


From: (John Chandler)
Subject: Re: [DNA] mtDNA results - Helena
Date: Fri, 1 Jul 2005 14:22:08 -0400 (EDT)
References: <dd.29357684.2ff5f585@aol.com> <fef9eee40507010830292c0ff0@mail.gmail.com>
In-Reply-To: <fef9eee40507010830292c0ff0@mail.gmail.com> (message from John SWalden on Fri, 1 Jul 2005 11:30:25 -0400)


John wrote:
> I case the answer is yes and I use 16001 to 16569 and 00001 to 00574, for a
> total of 1143 base pairs
> for HVR1&2 [it is not clear from the abstracts that these studies used those
> base pairs]
> and I use 3E-07 for the mutation rate.
> And I look at an MRCA at 10 generations back.
> The value from the Poisson calculation for one mutation is 0.0007
> or about one chance in 1400.

The correct answer is about 10 times bigger. Could you have slipped
a decimal point somewhere?

> If I double that to 20 generation back or about 500 years I get 0.0135
> or one chance in 75.

That's more like it.

> If I look for the 2.5% point [half the 5% left from the 95% ]
> That happens at 37 generations or about 925 years.

This 2.5% point is not what you are looking for -- it's just the time
at which the probability of getting exactly one mutation is the
specified level. The 95% confidence interval for the TMRCA is the
interval such that the probability is 5% that the time lies outside
the interval for the given number of mutations. Taking the number
of markers to be 1143 and the mutation rate to be 1e-6, I get a 95%
confidence interval of about 3000 years ago to 70,000 years ago for
one mutation.

John Chandler

This thread: