Archiver > GENEALOGY-DNA > 2008-07 > 1215642064

From: "Ken Nordtvedt" <>
Subject: Re: [DNA] calibrating for common ancestor - QUESTION?
Date: Wed, 9 Jul 2008 16:21:04 -0600
References: <><><><002501c8e205$158ff950$6400a8c0@Ken1><><004601c8e20d$998fa7c0$6400a8c0@Ken1><005501c8e20f$03509a60$6400a8c0@Ken1><>

I guess you could establish some thresholds and program this whole
correlation matrix process to discover clades. If you were the DIA (Defense
Intelligence Agency) and wanted to extract every morsel of information out
of such data you'd probably do it that way. I am pretty conservative in
establishing clades, so it has always been quite clear that a clade was
there when I identified it (I think I have only thrown away one clade).
That's probably because I mainly was looking for the very ancient clades
whose founders were most often hundreds of generations in tree branch length
from other clade founders, so quite distinct in their founding haplotypes.
About 40 percent of the I1 haplotypes in my research database I have thrown
into I1-ASgeneric --- meaning being I see no robust clade structure in this
population and suspect they all are just descendants of the original I1
founder less the later subclades which popped up in various places.

A marginally perceived clade could not be a yes or no thing. One would
suspect it, but still have doubts. A quantifier would always have border
line cases. Finding signals in noise is that kind of game. Fortunately the
stakes for us are not quite as high as deciding whether a submarine is


----- Original Message -----
From: "Sasson Margaliot" <>
To: <>
Sent: Wednesday, July 09, 2008 4:04 PM
Subject: Re: [DNA] calibrating for common ancestor - QUESTION?

> It is probably possible to have a program calculate the answer.
> Would it be yes-or-no, or some quantitative measure?
> On Thu, Jul 10, 2008 at 12:59 AM, Ken Nordtvedt <>
> wrote:
>> But in the early days I used the above correlation
>> matrix method, and it is always there to fall back upon or to check a
>> questionable clade.
> -------------------------------
> To unsubscribe from the list, please send an email to
> with the word 'unsubscribe' without the
> quotes in the subject and the body of the message

This thread: