GENEALOGY-DNA-L ArchivesArchiver > GENEALOGY-DNA > 2011-01 > 1293999989
From: "Diana Gale Matthiesen" <>
Subject: Re: [DNA] Recommendations for DNA testing
Date: Sun, 2 Jan 2011 15:26:29 -0500
> From: Debbie Kennett
> Sent: Sunday, January 02, 2011 1:28 PM
> I don't think 12 markers are sufficient to prove or disprove any genealogical
> I've had 12/12 marker matches with similar surnames fall apart with the
addition of more
> markers. Conversely I had a 10/12 mismatch which became a 35/37 match.
A 12-marker test can easily debunk a connection -- and a low correspondence at
12 markers often means you are looking at people in different haplogroups. But
12 markers can also be enough to support a match, though I'll grant that is
usually not the case. It depends how rare the haplotype and surname are.
CARRICO is an uncommon surname. Paper genealogy (including a summary of their
entire 1850-1930 U.S. census record) indicates that every CARRICO in the U.S.
prior to the 20th Century probably descends from one man: Peter CARRICO, the
1674 immigrant to Maryland. So far, Y-DNA testing is supporting that view. The
closest matches to their modal haplotype, so far, are 5/12, 17/25, and 21/37.
If someone with the surname CARRICO, or any name that could be considered a
phonetic equivalent, has an 11/12 or 12/12 match with this haplotype, I'm going
to accept that they have a connection to this family. I still advocate 67
markers, but not to substantiate a connection to the family, but, rather, to
look for branching markers.
And, yes, I've had the same experience, where a 10/12 match gets better with
more markers -- which, as we know, isn't usually the case. I use this group
from the GIBSON project to demonstrate several points, this being one of them:
Note how #G-6 has a 10/12 match with Lineage Group III-A modal haplotype, which
then turns into a 34/37 match, then a 64/67 match (I'm counting the change at
CDYb as one multi-step event). Please also note how the value of DYS413a splits
Lineage Group III into A and B subgroups. If just a few people had tested 67
markers, we might mistake DYS413a for a private mutation, instead of the branch
marker it turned out to be. You cannot predict whether 67 markers will be
helpful until *after* you've tested them.
In every case, there is a judgment to be made as to what level of testing is
needed to accomplish the goal. The answers are simply not black and white.