GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2010-03 > 1267817834


From: SVass <>
Subject: Re: [DNA] L4 added to ISOGG
Date: Fri, 5 Mar 2010 11:37:14 -0800
References: <mailman.522.1267734672.12642.genealogy-dna@rootsweb.com>
In-Reply-To: <mailman.522.1267734672.12642.genealogy-dna@rootsweb.com>


My exercise had one primary purpose which to demonstrate that L4 was not a private SNP restricted to one family so as to be added to the ISOGG hierarchy. Any comparison to L2+ individuals was only done to be complete. I wanted to compute an approximate age for both the separation of proven Ashkenazi L4+ clusters from each other and their separation from the European L2- clusters. Any European distribution that includes L2+ and/or L20+ individuals should be ignored in this vein. I agree with Lawrence Mayka that I should have run a sensitivity analysis on the data.

I reran by changing the L4 entry to 12 at 492 as suggested and then substituted a second L4 positive, 67 marker sample and then a third untested but in the other cluster with 67 markers.
Original results with 492=14
256 with a SigmaG of 70 for L4+ vs. L2-
211 with a SigmaG of 62 for L4+ vs. L2+
160 with a SigmaG of 47 for L2- vs. L2+
New L4+ L2-, 233 with SigmaG of 66 & 310 with SigmaG of 84

New results with 492=12
143 with a SigmaG of 44 for L4+ vs. L2-
160 with a SigmaG of 54 for L4+ vs. L2+
New L4+ L2-, 168 with SigmaG of 49 & 221 with a SigmaG of 64
New , L4, L4 est 66 with a SigmaG of 27

Finally, I went to the Faux website and copied the four L2- , DYS492 =14 persons with 67 markers to Generations4 I repeated the interclade age calculation to one L4 person and got ages that ranged from 106 to 147 generations with SigmaG ranging from 35 to 43.

Whether the distance between L2- and L4+ is 100 generations or 200 generations is not important for my original purpose. L4 clearly predates the Diaspora!

Sam Vass

On Mar 4, 2010, Al Aburto wrote:

> Very interesting Sam.
> I would try to find and use a few more samples for Generations4, just to be sure the results won't change ...
> Al



This thread: