GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2006-10 > 1161767782


From: "Alister John Marsh" <>
Subject: Re: [DNA] 12 Marker Matches Within Surname Groups..
Date: Wed, 25 Oct 2006 22:16:22 +1300
In-Reply-To: <7.0.1.0.0.20061024213331.10daaa30@wells.org>


Orin

You said...

"I have voiced my opinion before that 12 markers is pretty much
useless unless your intent is to prove two people are NOT
related. Even with the same surname."

I think a lot depends on context. Particularly if some very rare off modal
marker scores are involved. I once found a partial haplotype on Sorenson
with only 2 markers to compare to FTDNA markers. I confidently (..rashly??)
matched that to a family, based only on those two markers, as the markers
just happened to be rare off modal scores. When a full set of markers were
eventually obtained, the match went from 2/2 to 25/26. Perhaps I was just
lucky that time. Clearly 2 markers are better than none, 12 are better than
2, and 67 better than 12.

I had a case of 10/12, which included 4 mutation steps. I predicted that a
relationship was possible, as geographic location matched, a rare marker
score matched, and a recLOH looked possible on DYS385. Markers 13-37 when
they were later obtained only had 1 mutation step, on the fastest mutating
marker. Bridging haplotypes were also later found which made relationship
seem more credible.

I have a case of a 12/12 match of a surname variant, which I am almost sure
means no match. The geography feels wrong, and the markers are near modal.

I have a case of an 11/12 match, with paper trail connection, which I wonder
about. The mismatch is on the distinctive family marker, and other markers
are near modal.

I don't very often think in terms of 10/12, or 12/12. or "4 mutations on 12
markers". I am looking for matches on distinctive, off modal, slow markers,
and I add that to the surname, geographic location, and paper trail context,
to see what sort of "gut feeling" I get. Are "gut feelings" admissible
evidence?

Perhaps about 10%+ of my project participants have something sufficiently
distinctive in the first 12 markers, to make a 12/12 match look interesting.
For the others, 12/12 would be to various degrees into a "zone of
uncertainty".

John.



This thread: