GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2010-01 > 1263129293


From: "Anatole Klyosov" <>
Subject: Re: [DNA] R-L21 possible areas of origin
Date: Sun, 10 Jan 2010 08:14:53 -0500
References: <mailman.495.1263110451.29889.genealogy-dna@rootsweb.com>


> From: "Tim Janzen" <>

> I thus decided to analyze the
> currently available 67-marker haplotypes from the FTDNA R-21 project at
> http://www.familytreedna.com/public/R-L21/default.aspx?section=yresults to
> see if the current data supports any particular theory in regards to the
> area of origin of R-L21. The current data shows that R-L21 has the
> highest
> variance in England. Below are intraclade TMRCA estimates using various
> marker sets for groups of haplotypes from various geographical regions.
> The
> TMRCA estimates for samples from Great Britain and Ireland are slightly
> older than the TMRCA estimates for samples from Continental Europe.


Dear Tim,

I have analyzed R-L21 haplotypes in October, 2009. There were 509 of
67-marker
haplotypes and 770 of 25-marker haplotypes available. I have pooled all of
them and
composed a haplotype tree for the 67-marker haplotypes. It was perfectly
"smooth" and
symmetrical, and clearly pointed out at an only one common ancestor for the
whole tree
(which was then confirmed quantitatively, see below). Of course, "the only
one" is in terms
of DNA genealogy, since it might have been two or three brothers, with their
father, etc.,
however the base (ancestral) haplotype was just one. The "linear method"
(that is, the
"mutation counting" method) gave 3575+/-370 years from a common ancestor
(there were 2924 mutations in the first 25 markers in all the 509 of
67-marker haplotypes,
that is 2924/509/0.046 = 125 generations w/out correction for back
mutations, that is 143
generations w/correction, that is 143x25 = 3575 ybp).

To verify this figure I have applied the logarithmic method, which does not
consider mutations,
hence, it is free from "mutations overcounting", symmetry or asymmetry of
mutations, etc.
All the 770 of 12-marker haplotypes contained 49 identical, base haplotypes.
Hence,
[ln(770/49)]/0.022 = 125 generations w/out correction for back mutations,
that is 143
generations w/correction, that is 143x25 = 3575 ybp.

As you see, it is exactly the same figure as that one for the "mutation
counting" (linear)
method. This means that appearance of mutations in this set of haplotypes
and the resulting
disappearance of base haplotypes from the dataset follow the classical
first-order rate process
(including the back mutation counter-process).

For more fun, let's see how many of 25-marker BASE haplotypes suppose to
stay after those
125 generations (w/out correction for back mutations, see above). To be
ridiculously precise,
it should be 2.4 mutations (of course, it is "mathematical" number), since
ln(770/2.4)/0.046 =
125 generations.

In fact, those 770 of 25-marker haplotypes contained 2 base (identical to
each other) haplotypes.

In summary, present-day carriers of R-L21 haplotypes have a common ancestor
who lived
in the middle of the 2nd millennium BC, 3575+/-370 years ago.

Anatole Klyosov

***************

Tim wrote:
> 52 samples from Continental Europe:
> 50 markers: 3171
> 10 YHRD markers using YHRD mutation rates: 3052
> 24 slow markers: 4900
>
> 209 samples from the Great Britain and Ireland:
> 50 markers: 3233
> 10 YHRD markers using YHRD mutation rates: 3296
> 24 slow markers: 4910
>
> 15 samples from Wales:
> 50 markers: 2898
> 10 YHRD markers using YHRD mutation rates: 2535
> 24 slow markers: 5189
>
> 48 samples from Scotland:
> 50 markers: 2963
> 10 YHRD markers using YHRD mutation rates: 3387
> 24 slow markers: 3614
>
> 94 samples from Ireland:
> 50 markers: 3220
> 10 YHRD markers using YHRD mutation rates: 2934
> 24 slow markers: 4130
>
> 52 samples from England:
> 50 markers: 3350
> 10 YHRD markers using YHRD mutation rates: 3582
> 24 slow markers: 6905
>
> 15 samples from Germany:
> 50 markers: 3065
> 10 YHRD markers using YHRD mutation rates: 2615
> 24 slow markers: 3996
>
> 12 samples from France:
> 50 markers: 3285
> 10 YHRD markers using YHRD mutation rates: 3986
> 24 slow markers: 3447
>
> 16 samples from Scandinavia:
> 50 markers: 2528
> 10 YHRD markers using YHRD mutation rates: 2490
> 24 slow markers: 5780


This thread: