GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2007-05 > 1178711583


From: Vincent Vizachero <>
Subject: Re: [DNA] Possible reporting error on SMGF - DYS463=19
Date: Wed, 9 May 2007 06:53:03 -0500
References: <KHEKIJEABJGJEKDPFEDMIENEDJAA.elizabethod@eircom.net>
In-Reply-To: <KHEKIJEABJGJEKDPFEDMIENEDJAA.elizabethod@eircom.net>


On May 9, 2007, at 4:52 AM, Elizabeth O'Donoghue wrote:

> Am I calculating this correctly? If the mutation rate is 1/600
> generations,
> doesn't that mean 1 mutation over 15,000 years (at 25 yrs per
> generation)?
> If so, why is there *bound* to be a variant, particularly one so
> disparate,
> within this subclade, which has been estimated to be less than
> 3,000 yers
> old?

There are some simplifying assumptions in this example, but this is
the gist of it:

Assume that the 3,000 year age is correct. Using STR variance to
estimate the age of an STR-based cluster is an exercise fraught with
peril, but let's just assume the 3,000 years for this discussion.
Assume 30 years per generation (probably more accurate and certainly
easier to calculate in this example).

3,000 years = 100 generations.

Assume that DYS463 has a mutation rate of 1/600 (one per 600
generations).

If every pair of men had the founder (100 generations ago) as their
MRCA, then you would expect 1 person in 6 to have a mutation at
DYS463. If you sampled 40 men, you would expect to find about 7 of
them with a mutation at DYS463.

Of course, in reality some lines may have experienced reversing
mutations ( e.g. from 23 to 22 and back to 23). Also, maybe the
mutation rate is a little slower than 1/600 as John Chandler suggested.

Making corrections for these effects would cause the observed
variance to be less than the simple calculation made here but not
dramatically so. Maybe instead of 7 mutations among 40 men you would
find 3 or 4. I would call that variant.

Vince


This thread: