GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2002-02 > 1013544600


From: "John F. Chandler" <>
Subject: Re: [DNA] Distribution of alleles
Date: Tue, 12 Feb 2002 15:10 EST
In-Reply-To: DNACousins@aol.com message <8a.13e6c757.299a82c1@aol.com> of Tue, 12 Feb 2002 07:38:01 -0700


Ann wrote:
> There's an article describing the Y STR database by Roewer et al which
> includes the 30 most common haplotypes:
>
> http://www.eva.mpg.de/genetics/Roewer_FSI2001.pdf

Gee, if I had known about that article, I might not have gone to the
trouble of probing the database. Of course, the article is a snapshot
of the database when it was only half its present size (and had no
contribution at all from the British Isles).

> The haplotype John pulled out of the database is one of the 30, but there are
> others which are even more common. It might be interesting to look at this
> again with Nancy's charts in hand.

The trouble with the table of haplotypes in the article is that it gives
only the top 30. If you want to "collapse" the table to get the numbers
for a 5-locus haplotype (rather than the full 9-locus haplotypes shown),
you're out of luck. If you add all the entries in the table that give a
5-locus match with the AMH, for example, you will be missing the very
large part of the AMH total that didn't make it into the top 30. In the
same way, you can't tell the frequency of the 5-locus haplotype that I
found. The very same mechanism that makes the AMH the most frequent
5-locus haplotype, virtually smothering the alternate haplotype I found,
is operating even more strongly in this table. There are only 944
samples represented in the top 30, out of some 4000 in the database at
the time. We know that the value 10 is much more frequent than 11 for
DYS 391, but I added up the counts within the top 30 list and found 474
with 11 and only 470 with 10.

John Chandler


This thread: