GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2005-06 > 1118970304


From: "Glen Todd" <>
Subject: RE: [DNA] Overall Genetic make-up----comparing Y DNA/mtDNA tests with BioGeographical tests
Date: Thu, 16 Jun 2005 19:05:04 -0600
In-Reply-To: <200506162334.j5GNYunw030545@mail.rootsweb.com>


> I don't agree with you on this. I feel that mtDNA and YDNA
> "fingerprints" do show your ethnic backround ON THAT
> PARTICULAR LINE BEING TESTED. And
> actually, as more data comes in from world populations, and

The problem with that particular thesis lies in recombinant mathematics.
Y-DNA and mtDNA lines are in so far as we currently know the ONLY ones that
come down to us more or less intact. That means that if I go back to,
say, the American Revolution, I have two known lines out of a thousand. I
have paper trail records on most of the rest, but that's all. Until such
time as we can get direct male line or female line cross-references on
EVERYBODY in that tree, which will require an increase in the sizes of the
databases of several orders of magnitude, or until some computational
breakthrough allows us to 'reverse engineer' the recombinant process, we
can't 'add up' all the different lines to get some sort of 'ethnic sum'.

Firms like DNAPrint maintain that they can indeed arrive at a breakdown of
'genetic percentages' using an entirely different approach (and I have NO
wish to debate their methodology or results again), but even so, one line a
dozen or so generations back will simply get lost in the statistical noise.
For instance, I test out as "100% European" (whatever THAT means). Does
that mean that I can be absolutely certain that there isn't a genetic
contribution from an African Roman Legionnaire or an Oriental merchant
somewhere fifty generations back? Of course not. The best that they
can say is that there is no recognizable contribution within the parameters
of the test.

In short, your contention is theoretically accurate, but realistically
unachievable.

Glen


This thread: