GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2003-12 > 1070402668


From:
Subject: Re: [DNA] Genetic Distance calculation - Comments re MacGregor and a further question
Date: Tue, 2 Dec 2003 17:04:36 -0500 (EST)
References: <1ec.13e271f9.2cf0ed2f@aol.com> <002b01c3b1c7$f967e610$799c89d9@helen> <REME20031201154816@alum.mit.edu> <00bc01c3b86f$b78fef90$3fb50818@c452380a>
In-Reply-To: <00bc01c3b86f$b78fef90$3fb50818@c452380a> (ecbeaty@comcast.net)


Earl wrote:
> With the Beatty surname group we have a consideration which you didn't
> address, and which affects the results. We have a very clearly defined
> common ancestor (CA) for about 3/4 of the Beattys tested. I include in this
> group two who are 3 mutations away from CA. However by the principal of
> maximum parsimony these two have a more recent common ancestor only one
> mutation away. And after another mutation in their common line there is
> another testee who is one generation from CA. Thus with these 3 samples
> there is only 4 mutations, leading me to conclude that squaring the 2 threes
> gives much more weight to these 2 than is realistic. Can we assume that
> these problems of recent common ancestors average out?

I don't quite follow that description. Are you saying that two people
each match the ancestral haplotype at all but one locus, and differ by
three steps at that locus? If so, your deduction that there was an
intermediate ancestor who differed by only two at that locus could be
wrong, since the change could have been all in one three-step mutation.
(Or have you actually observed someone with the intermediate haplotype?)
The DNA comparison of the third person mentioned is also not clear to me.
Is he the same as the first two, except for one additional step of
mutation at the same locus? Or is it one other mutation at some other
locus?

At any rate, if you deliberately sought out possible kin of the first
subject who had a difference of three, then you have indeed introduced
a bias in the statistics. The same would be true even if it wasn't
deliberate, but simply resulted from the network of cousins who
recruited each other into the DNA project. As the study grows, you
can expect these biases to average out, but the ride can be "bumpy."

John Chandler

This thread: