Archiver > GENEALOGY-DNA > 2008-06 > 1214238547

Subject: Re: [DNA] Anatole's method
Date: Mon, 23 Jun 2008 16:29:07 +0000

Vince, please, bear with me. I am still in the process of explaining, and just posted
a continuation in the E1b1b thread. As you see, there is something to digest, that is
why I move rather slowly.

I am going to address your question a bit later, however, since you have asked, here it is.

The method, of course, is not a universal one, and there are no universal methods. It works
at different time spans depending on a number of markers in a haplotype. I would not recommend to use it when you have less than 5 base haplotypes in your set. Simply because the ratio becomes very sensitive
with a number of base haplotypes below 5 or 4. One can still use it - just to get a preliminary idea when a common ancestor could have lived. Clearly, if you have less than 5 base haplotypes out of 100, this should be very many generations ago.

In a 100-haplotype list that you have mentioned, 5 base (identical) haplotypes left would point at the following time spans to a common ancestor (but it should be checked with mutations, to be on a safe side) [besides, these all should be corrected for back mutations, and I am giving the corrected figures in brackets]:

6-marker haplotypes - 312 generations [456 gen = 11,400 years]
12-marker haplotypes - 125 generations [143 gen = 3,600 years]
25-marker haplotypes - 65 generations [69 gen = 1,700 years]
37-marker haplotypes - 33 generations [34 gen = 850 years]

In you example with no base haplotype left, it would be more than:

6-marker haplotypes - 480 generations [960 gen = 24,000 years]
12-marker haplotypes - 192 generations [238 gen = 6,000 years]
25-marker haplotypes - 100 generations 111 gen = 2,800 years]
37-marker haplotypes - 51 generations [54 gen = 1,350 years]


Anatole Klyosov

Vince wrote

The "logarithmic" method for dating groups of clades that you
describe is certainly novel.

If, in a 37-marker haplotype set, there are 100 haplotypes and they
are all unique then what is the base haplotype?

This thread: