GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2008-05 > 1211415098


From: "Tim Janzen" <>
Subject: Re: [DNA] S21/S28 Split+m223 stuff
Date: Wed, 21 May 2008 17:11:38 -0700
In-Reply-To: <298538.28235.qm@web86601.mail.ird.yahoo.com>


Dear Alan,
I share the same concerns that David and you do that Ken's current
method produces an age for the common MRCA for S21 and S28 (3780 years per
his message yesterday) that is simply too young. I don't know how this
situation is going to settle out. Perhaps weighting the markers will change
the age calculations and will result in an older age for the common MRCA for
S21 and S28.
I thought I would share some additional calculations using an
updated version of my spreadsheet that pertain to this whole situation that
I have been doing over the past day based on the latest data from the FTDNA
haplogroup projects and from David Faux's S28 project (downloaded
yesterday). The calculations include variations in which I ignored the data
for all of the multi-copy markers (DYS 385, DYS 395, DYS 413, DYS 425, DYS
459, DYS 464, YCA II, and CDY) and where I retained all the data. The ages
using Ken's method are designated as "KN". The ages using James Heald's
method are designated as "JH". The ages where all of the multi-copy markers
are retained are designated as "+M" and the ages where all of the multi-copy
markers were ignored are designated as "-M". I also calculated DYS 389B
after subtracting DYS 389i from DYS 389ii and included DYS 389B in the
calculations. Based on the results it would appear that removing the data
for multi-copy markers doesn't generally change the results more than about
5% on the average.
I only have 8 R1b1b2 (M269+, S116-, S21-) haplotypes at this point:
T982M, 5VU5V, ANSRV, CDNKN, DZ7XA, DZRMK, KV7Y2, T2WDH in Y search. I have
included the results below for these haplotypes. I would be happy to run
calculations for a larger R1b1b2 (M269+, S116-, S21-) dataset if someone
happens to have a larger dataset.
I think that it is interesting that the MRCA for R1b1c9 (S21+) is
younger than the MRCA for R1b1c10 (S28+) even though we know that the S21
SNP occurred before the S28 SNP did. This suggests that the S21 lineage
came through some type of genetic bottleneck one or two thousand or possibly
more years after the S21 SNP occurred.
Note that we still have the issue in that James Heald's method
results in older ages than Ken's method does. Ages calculated using James'
method are about 50% older than the ages calculated using Ken's method.
Perhaps that won't be as much of a problem after each of the markers is
weighted.
Also note that we still have at least somewhat of an issue with
mutation rates. Data using only 12 markers frequently gives older dates
than does data using 37 and/or 67 markers. Hopefully, mutation rates can be
refined at some point in the future allowing for more accurate calculations.
It should also be noted that the age calculations using the ASD
method in Ken's spreadsheet exactly match the age calculations using
variance method in my spreadsheet. This is not surprising since the ages
should have matched.
I haven't yet had a chance to calculate 95% confidence intervals and
I haven't yet had a chance to follow up on John Marsh's suggestion to run
the calculations on datasets involving the fast mutating markers and those
involving the slow moving markers.
No, matter how this settles out I think Ken has done an excellent
job of summarizing all of the theoretical aspects of these age calculations
and has really helped move this issue forward.
Sincerely,
Tim Janzen

Project JH -M KN -M JH +M KN +M # of haplotypes # of markers
R1b1c6 M167+ 6678 2629 6531 3267 32 12
R1b1c6 M167+ 2336 2349 3843 2243 28 37
R1b1c6 M167+ 5714 2811 5958 2700 20 67

R1b1c7 M222+ 1804 1617 1736 1573 416 12
R1b1c7 M222+ 1599 1501 1608 1143 341 37
R1b1c7 M222+ 1786 1600 1863 1351 135 67

R1b1c9 S21+ 4110 3486 4050 3539 190 12
R1b1c9 S21+ 4808 3465 4224 2727 180 37
R1b1c9 S21+ 3349 3075 3462 2431 120 67

R1b1c10 Kerchner 6655 4001 6937 4876 62 12
R1b1c10 Kerchner 5928 3352 5401 2537 59 37
R1b1c10 Kerchner 6224 3589 5924 2850 43 67

R1b1c10 Faux 5451 3834 6300 5187 166 12
R1b1c10 Faux 5015 3299 4614 2486 131 37

R1b1b2a2 S116+ 4077 3557 4796 2866 72 37
R1b1b2a2 S116+ 4519 3782 4798 3090 55 67

R1b1b2 S116-S21- 18952 6020 16589 5768 8 12

-----Original Message-----
From:
[mailto:] On Behalf Of Alan R
Sent: Wednesday, May 21, 2008 1:47 PM
To:
Subject: [DNA] S21/S28 Split+m223 stuff

Has a MRCA for R1b1c* S116 -ve folk been calculated? That would be
interesting to see as they could be a remanant of the R1b population before
the explosion that sent certain R1b lineages into huge demographic
ascendancy. Their distribution may also be a clue as to the extent of R1b
prior to these lineages.

I like Mr Faux still find it very hard to reconcile the late dating with our
normal understanding of demographic turning points in European prehistory
derived from archaeological and environmental data.
Alan

This thread: