Archiver > GENEALOGY-DNA > 2004-12 > 1104262531

Subject: Re: [DNA] MRCA = 800 AD
Date: Tue, 28 Dec 2004 19:35:31 +0000


Unfortunately I must admit to using some sort of "subitization" (I think that term is known in math) approach to estimating likelihood (intuitive guessing more or less).

Jean and Mr. Sinclair have DYS385a,b = 11,15. I know that this is rare, and in a case where the paper trail points to the same ancestor this will tilt by "relative certainty factor" toward accepting the hypothesized connection.

Can you, Doug or John or others here who are math whizes devise something more systematic so we could factor marker value rarity into a decision - making algorithm of some kind?

David F.

-------------- Original message --------------

> This issue of how sharing values or off-modal values at the very slow
> mutating markers is fascinating. No one to my knowledge has explored
> quantitatively how this should sway our interpretations. Clearly for some
> "after the fact" situations, where a mutation occurs should not be
> considered. But suppose the question is posed this way: Two people differ
> from each other at some small number n out of total N markers, but they also
> share a very rare, off-modal repeat value for their haplogroup at some
> particular marker. How does this latter information change our
> probabilities of something concerning the two individuals?
> Ken

This thread: