GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2008-06 > 1213025673


From: "Ken Nordtvedt" <>
Subject: Re: [DNA] Generations2 program and the CMH-12 extended haplotypes...
Date: Mon, 9 Jun 2008 09:34:33 -0600
References: <936471110806080753k1912d8e7s787d45c8d71666a@mail.gmail.com><009f01c8c97a$e40205b0$6400a8c0@Ken1> <484C0D1B.90109@sbcglobal.net> <00f801c8c988$87da6030$6400a8c0@Ken1><484C81C9.7080509@sbcglobal.net><484D4B0C.9010308@sbcglobal.net>


----- Original Message -----
From: "Alfred A. Aburto Jr." <>
To: <>
Sent: Monday, June 09, 2008 9:23 AM
Subject: Re: [DNA] Generations2 program and the CMH-12 extended haplotypes
...


> Oops!!!
> I misread Ken's spreadsheet results for the 67 marker J1 Cohen case!
> There are two results given. The first is G*(A), and the second is an
> estimate of G(A) which is the MRCA age! Previously I incorrectly used
> G*(A).
>
> Here are the correct results:
> (1) MRCA age of J1 Cohen's, 29 haplotypes of 67 markers, 30
> years/generation: 2100 years ago
> (2) MRCA age of J1 Cohen's, 66 haplotypes of 37 markers, 30
> years/generation: 2300 years ago.
>
> The MRCA ages are virtually the same! This gives one confidence that we
> are near the correct age for the J1 Cohen's (the one and only group of
> Cohen's in J1 as far as I know). Still it would be valuable to have
> other J1 Cohen's test 67 markers and upload their results to the Cohen
> DNA Project and/or the Y-DNA J Project.
> Al
>
When I see your database I get concerned that you are selecting against some
of the members of this clade. If you freeze a large number of markers on a
modal value in the selection process, you are eliminating members of the
clade who happened to have a mutation on one of the defining markers. This
gets more important if your defining markers happen to be middle or fast
markers, or if you use a large number of markers to select from the
database.

I try to use the minimum number of slow markers that I can to isolate a
clade. If I am forced to use a larger number of fast markers in the
identification of a clade, then I one by one let one of those defining
markers join the "wild card" set of markers to see how many haplotypes are
picked up by that process.

What one is really trying to do is include those haplotypes and only those
haplotypes which have a clear probability advantage of having descended from
the clade founder versus having descended along some other line of the
parent (encompassing) population.

Ken




This thread: