Archiver > GENEALOGY-DNA > 2009-12 > 1260858975

From: "Tim Janzen" <>
Subject: Re: [DNA] R-U152 and R-L21 on the European Continent
Date: Mon, 14 Dec 2009 22:36:15 -0800
In-Reply-To: <>

Dear Al,
These are tough issues with no easy answers. My impression from
doing quite a few TMRCA estimates on young subclades (from the same surname
or from groups like the Clan McDonald) is that using a 50-marker haplotype
dataset is fine and gives "reasonable" results. The issue starts to get
more thorny in my opinion as you start getting into subhaplogroups or
subclades that are (or appear to be) 3000 years old or more. See the
message below I posted in July. Note that even at 1500-2000 years the TMRCA
estimates with the 50-marker panel (old version) are starting to become
distinctly younger than those given by the 24 slower marker panel (new
version). I think that the simplest answer for this is saturation of the
variance of the fast mutating markers. The separation between the TMRCA
estimates with the 50-marker panel and the TMRCA estimates using the
24-marker slower mutating panel becomes greater and greater the further one
goes back in time.
I fully agree we need more research on this whole topic. One thing
that would be helpful would be to do TMRCA estimates in many different
haplogroups marker by marker. If this were to be done, I think that it
could help point out inaccuracies in some of the mutation rates or point out
haplogroups where the mutation rates don't seem to be consistent with each

-----Original Message-----
From: Tim Janzen [mailto:]
Sent: Thursday, July 09, 2009 10:40 PM
To: 'Didier Vernade'
Subject: RE: [DNA] R-SRY2627+ dating

Dear Didier,
I haven't done an intraclade TMRCA estimate for R-M222 recently. I
just ran one and got these results for 143 67-marker haplotypes downloaded
from the FTDNA R-M222 on June 27 (after removing extra haplotypes for any
surname that had more than one haplotype):

1857 years (new version)
1744 years (YHRD version)
1436 years (old version)

If I reduced the number of haplotypes to 50 (randomly chosen) I got these
1892 years (new version)
1655 years (YHRD version)
1574 years (old version)

If I reduced the number of haplotypes to 25 (randomly chosen) I got these
2058 years (new version)
1901 years (YHRD version)
1653 years (old version)

-----Original Message-----
[mailto:] On Behalf Of Al Aburto
Sent: Monday, December 14, 2009 6:37 AM
Subject: Re: [DNA] R-U152 and R-L21 on the European Continent

Oh, I get it. For the "younger clades" (as if I knew automatically where
to place the dividing line between "younger" and "older") the fast
mutators are not saturated so it is ok to use them?? I wonder if the
fast mutators are saturated then how does one know they are accurate at
all? Perhaps the fast mutators are wrong even for the "younger
clades"? If the fast mutators were not saturated (or wrong) for the
"younger clades" then wouldn't that fast mutation rate be accurate to
use no matter what (young or old clade)? Maybe someone should do some
detailed study of the mutating properties of all markers and just throw
out the saturated ones instead of all those except the slowest 24 in the
67 marker set? Certainly there seems a huge amount of background
research is needed into the mutating properties of all the Y-STRs...

This thread: