GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2009-12 > 1260991115


From: Michael Walsh <>
Subject: Re: [DNA] R-L21 Plus news - downstream branches to be better defined
Date: Wed, 16 Dec 2009 13:18:35 -0600


---> My response:

Your situation may actually be similar to mine. I have no 12 of 12
matches... a couple of 22 or 23 of 25 types, but they didn't make kind
of geographical sense. I upgraded to 37 markers thinking the 22 or 23
types would narrow out to who was really important for me to focus on.
Instead, the lower marker matches dropped off and a couple of new
people appeared. I don't think anyone was closer than 32 of 37. For
a long time I didn't upgrade further because I just thought it was
hopeless. I just started looking for deep ancestry stuff but people
kept asking for specific results at key STR's like DYS492, which I
didn't have.

Finally, I went ahead and did the 67 marker upgrade. My 31 or 32 of
37 matches disappeared, except one. Now that one was 61 out of 67 so
I hardly lost any genetic distance ground with him even though we
added 29 markers to the comparison. Better yet, another person of the
same surname popped up 60 of 67; and another with a different surname
also at 60 of 67. Lo and behold, all three of these people were from
vicinity of Wales. Only 20-30km apart! They are not close
genealogical cousins, but I was quite excited that some of my old
family folklore actually appeared to be true.

I was also excited that I may have a found members of some ancient
clan or tribe. More practically speaking, I know had a precise set of
signature markers (and new surnames/geographies) that I could search
for better matches. In my new, smarter search attempts I did found
several new people to compare with and did find one that is even
closer (61 of 67) so I'm inching along and at least I know I'm on the
right track (or perhaps more appropriately, branch.)

I think why this occurs is that FTDNA's 38-67 marker panels have some
nice slow and medium moving markers that are good "signature" flags.
Some of the first 37 high speed markers like DYS464 and CDYa/b can
actually throw you off track a bit in my opinion as can markers that
"jump" more than one step at a time like YCAIIa/b.

Think of like this. R1b1b2 in Europe has a very "bushy" family tree.
An awful lot of branches have sprouted in a very short period of time
so we have few distinguishing characteristics. Some of the
differences are just "noise" or twigs from different branches crossing
each other. Deep clade testing can help eliminate some of that noise
as does deeper STR testing ... to 67 at a minimum, I think.

The other side of the coin (different from "noise") is making sure you
are on the right branch. You might have some close twigs that seems to
be on another branch but deeper testing shows the differences may be
due to very recent mutations. Extended STR testing helps identify
those people like it did for me.

Mike

---> CeCe Moore <> wrote:

Same with us. He has no close matches...nothing even at 31/37. FTDNA
has him as R-M269; predicted R1b1b2.

> Date: Tue, 15 Dec 2009 19:46:04 -0800
> From:
> To:
> Subject: Re: [DNA] R-L21 Plus news - downstream branches to be better defined
>
> On Tue, Dec 15, 2009 at 7:27 PM, Michael Walsh <> wrote:
>
> > ---> My response:
> > A number of genealogical geneticists have already indicated that at
> > least 67 markers are needed to discern between R1b1b2 people. Most
> > notable, probably, is Dr. Michael Hammer's statements to this effect
> > this March. The problem is that R1b1b2 people's haplotypes look very
> > much alike so more markers are needed to differentiate between people.
> > There are cases of 25 of 25 or 35 of 37 matches that turned out to be
> > in totally different sub-clades of R1b1b2.
>
> This is true of some but by no means all R1b1b2. My father has no false
> matches even at the 22/25 level.


This thread: