GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2008-09 > 1220626385


From: "David Faux" <>
Subject: Re: [DNA] Age of R1b
Date: Fri, 5 Sep 2008 07:53:05 -0700
References: <090520081302.3176.48C12DFC000BCBF200000C68220073483090019C019704050E@comcast.net>
In-Reply-To: <090520081302.3176.48C12DFC000BCBF200000C68220073483090019C019704050E@comcast.net>


Anatole and Vince:

Now that I am off the hot seat, it is interesting to watch you two "dukeing
it out" . Of course I still don't agree with Anatole anymore than I agree
with Vince. As a reason I only need to use one word to encapsulate my
concerns - calibration, or lack thereof.

In population genetics today there is a realization that in order to obtain
valid estimates of of haplogroup ages, or split times between clades,
whatever method is used needs to be calibrated to a date tha can serve as an
anchor point. In Karafet et al. (2008) they used one of the currently
accepted dates for the expansion out of Africa and the split between
haplogroups C and T. They used 70,000 years. The authors acknowledge that
this date is somewhat artibrary and is certainly not to be taken as the
final word. Their calculations could change dramatically with only a slight
shift in the presumed CT division. Their date for the TMRCA of R1 is 18,500
years but the uncertainty factor as reflected in the confidence intervals is
huge (12,500 to 25,700). As far as I know this is the youngest date yet
offered for R1, although the fact that one of the main authors is using
another methodology in their most recent publication does not bespeak of
great confidence in the SNP - based estimates of Karafet.

I don't recall seeing any calibration in either of your estimates. As far
as I can see you are still essentially using vague variance rates which
likely do not apply beyond 1000 years ago or projecting germ line rates back
into prehistory which is considered by population geneticists as a "classic
error" which will distort the date toward a highly artificial younger age.

You can banter back and forth from now till the cows come home but unless
you can tie your work into a series of ancient YDNA samples (not yet
possible), or go back to the time Mount Toba blew its top, your uncalibrated
dates just hang in the air, unrooted.

David K. Faux.


On 9/5/08, <> wrote:
>
> Anyway, Vincent, I think we both are generally on the same page, and - as I
> see - your comments are more personal that factual. Please do not be
> jealous, the field of DNA genealogy is large enough to accommodate both of
> us. And, yes, when you make negative comments, please supply them with
> specific examples and calculations. This would be a courtesy for others. Me
> included.
> Regards,
> Anatole Klyosov
>
>

This thread: