GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2010-02 > 1265484768


From: "Lancaster-Boon" <>
Subject: [DNA] Variance Assessment of R:U106 DYS425Null Cluster
Date: Sat, 6 Feb 2010 20:32:48 +0100


Hi Ken

Yes, this is also what I would like to understand. My question is hoping to
prompt an explanation of why Anatole thinks your example was unrealistic.

What I understand from his responses so far is that he claims that normally
the work he does involves data sets where both the family tree and the
starting haplotype are clear and obvious and therefore the reconstruction of
those, which is apparently his first step, effectively has no tolerance
because it is just right.

I can understand that if you are absolutely certain of the phylogeny and the
starting haplotypes at the nodes, then age estimations will have lower
tolerances.

But based on my experience, and I've worked with some of the same data sets
Anatole has worked on, the family tree structures of many clades are
absolutely shots in the dark. This in other words means that you NORMALLY do
not know anything about any common ancestors within the group except the one
right at the stem.

Is this also how others see it? To me this APPEARS to be the big difference
between Anatole's assumptions and those used by yourself, and I guess
defining the differences is the big question on everyone's lips now. There
have been way too many unclear discussions where this question has gone
unanswered by anything other than "if you do not already know then there is
no point" type responses.

Let me say that I might be misunderstanding all sides of the discussion
(there have been some less than clear back-and-forths) but I hope not! :)

Anyway thanks to both your and Anatole for trying once again, and I hope the
discussion continues!

Best Regards
Andrew

---
From: "Ken Nordtvedt" <>
Subject: Re: [DNA] Variance Assessment of R:U106 DYS425Null Cluster
Date: Sat, 6 Feb 2010 08:19:30 -0700
References: <CB3B852E26C64182B4C95D818AEB3D5F@PC>

The simple set of short haplotypes was given only for purposes of
understanding what "counting mutations" is supposed to mean in Anatole's
method.
It was not about eventually evaluating a statistical confidence interval.

I maintain that you can not count mutations because they happened in the
past and would require an inference about the whole tree history. One can
only count outcomes of mutations in the sample set of final haplotypes seen
today, after making an interence about a founding haplotype. So my goal
right now is to just find out what Anatole is counting, whatever he wants to
call it.


This thread: