GENEALOGY-DNA-L ArchivesArchiver > GENEALOGY-DNA > 2010-02 > 1266757636
From: "Anatole Klyosov" <>
Subject: Re: [DNA] TMRCA assessments
Date: Sun, 21 Feb 2010 08:07:17 -0500
>From: Robert Stafford <>
>Here is an example from our surname project. It is a documented family.
>16 match the ancestral on FTDNA's 25 markers.
>16 differences from the ancestral.
So, what is a problem? I do not see a statement or a puzzle. Any concern?
On a first glance I do not see here anything unusual from a point of view of
Since there are 16 base haplotypes among the total of 28, then
[ln(28/16)]/0.046 = 12 generations (x 25 years) from a common ancestor, that
is 300 years (approximately, of course).
Since there are 16 one-step mutations in those 28 haplotypes (one apparently
recLOH I count for one mutation), then
16/28/0.046 = 12 generations, the same 300 years.
So what is a bid deal?
With the margin of error it would be 300+/-150 years to a common ancestor
(with 95% confidence), since there are not many haplotypes in the dataset,
good statistics is not there. With about 70% confidence (one sigma) it would
be 300+/-80 years to a common ancestor, around 1710 year plus-minus the
margin of error. What the documents say about the common ancestor?
Among those 16 mutations there is not a single true two-step mutation. This
shows that there is practically no back mutations in the dataset. Even one
would have contributed only 1/16 ~ 6% in the total pool, that is well below
the margin of error.
A: DYS464 = (15,15,15,16) vs. (15,15,16,16) for ancestral
B. DYS391 = 10 vs. 11
DYS439 = 12 vs. 13
C. DYS390 = 24 vs. 25
D. DYS389ii-i = 17 vs, 16
DYS449 = 30 vs. 29
E. DYS389ii-i = 17 vs, 16
DYS439 = 12 vs. 13
F. DYS439 = 14 vs. 13
G. DYS385 = (11,11) vs. (11,14)
H. DYS389ii-i = 17 vs. 16
I. DYS393 = 12 vs. 13
J. DYS447 = 24 vs. 25
DYS459 = (9,10) vs. (10,10)
>K. DYS464 = (15,15,15,16) vs. (15,15,16,16) for ancestral
>L. DYS390 = 24 vs. 25
On Sat, Feb 20, 2010 at 5:50 PM, Anatole Klyosov
> With repeated "inline mutations" originated by a single mutation those
> haplotypes form a separate branch in the haplotype tree. Separate branches
> are calculated separately, hence, there are no "overcounting" mutations
> If you still do not understand, ask me for a concrete example.
|Re: [DNA] TMRCA assessments by "Anatole Klyosov" <>|