GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2008-01 > 1200155078


From: Thomas Krahn <>
Subject: Re: [DNA] Markers - How many and which ones?
Date: Sat, 12 Jan 2008 10:24:51 -0600
References: <2dff56a0801120740y409c2924x3b93b3503ea632e0@mail.gmail.com>
In-Reply-To: <2dff56a0801120740y409c2924x3b93b3503ea632e0@mail.gmail.com>


Colin Ferguson wrote:
> In a recent thread Ken Nordtvedt expressed surprise that the John
> Chandler's average mutation rate for FTDNA markers 1-37 was as high as
> 0.049. John offered the following: "Yes, it surprised me, too. It is
> dominated by the extremely high rate for CDY, which counts doubly in
> the average."
>
> Whoever chose to include CDY in the set did a good thing.
Kudos to Alan Redd and the team of scientists at the U of Arizona who
discovered this marker and made it available to us.
> Back up for
> a minute to just a few years ago when 37 markers was state of the art.
> Had that person chose some other marker with a more typical mutation
> rate John's analysis would have led to a number more like 0.0028. In
> turn they would have used Ann Turner's MRCA calculator and computed a
> median MRCA for a 2 marker mismatch to be 13.3 generations. But that
> didn't happen so instead we use 0.049 and our answer is 7.6
> generations, nearly a factor of 2 less.
>
> So now I wonder is there some other marker that has been left out of
> the 38-67 set that would have had a similiar effect?
Definitely DYF399. This marker is even more volatile than DYS724/CDY and
has 3 alleles in regular haplotypes.

> And more to the
> point, how many more markers are needed to make the average mutation
> rate converge to some stable value?
>
Depends on what you mean with "some". The biggest impact on the high
variance of the mutation rate estimation is still the number of reliable
studies of reported mutations in confirmed pedigrees.

I hope this helps,

Thomas

This thread: