GENEALOGY-DNA-L ArchivesArchiver > GENEALOGY-DNA > 2008-09 > 1220551440
Subject: Re: [DNA] Age of R1b
Date: Thu, 04 Sep 2008 18:04:00 +0000
Lawrence Mayka wrote:
>I would very much like to know which data you used for these older dates
(12K years for R1a in the Balkans, 11K years for R1b in Armenia). I presume
that you are not using Ysearch... I hope you are not using YHRD... and a
Of course not. I use haplotypes classified by the authors according to their haplogroups. The Balkans R1a1 haplotypes are published by Barac, Pericic et al, the Armenian R1b haplotypes by Weale et al.
However, all ancient sets of haplotypes split into different branches, from very recent ones to very old ones. One should be able to separate branches, and then to look for common ancestors for each one of the branches. I use the same approach I have just demonstrated to Ken with a haplotype tree with different branches of "I" haplotypes.
Those ancient data series are different in kind from more recent ones, since recent series never contain those bushy, fluffy, loose branches, containing VERY mutated haplotypes, characteristic for descendants of ancient common ancestors.
As you might remember, I appeared here first time with a description how I separated branches from mixed haplotype sets, in order to sort out mutations and calculate them. However, there was an outcry that it is wrong and impossible, and nobody needs it, and one should use only clean subclades. "He uses mixed haplotypes, god forbid!". Well, you are not going to see clean subclades with "ancient" haplotypes.
To count mutations in a haplotype mix stemming from a number of "common ancestors" is like to measure an "average temperature" in a hospital (including a morgue).