GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2009-04 > 1238933452


From: "Alister John Marsh" <>
Subject: Re: [DNA] Comparison of Chinese, Yoruba,Watson and Venter genome y-snps
Date: Mon, 6 Apr 2009 00:10:52 +1200
References: <200904051042.n35AgVWO030245@mail.rootsweb.com>
In-Reply-To: <200904051042.n35AgVWO030245@mail.rootsweb.com>


Tim,

Many thanks for this assessment.

It seems that different sequencing processes (and or computer analysis
techniques) used for different genomes might be achieving different
standards of accuracy, making it for the moment difficult to get conclusive
results on haplogroup ages using SNPs.

I wonder if the computer software used to cobble together the short
fragments of DNA reads might be reading SNPs as a mismatch with the Hugo
sequence, and discarding them as ambiguous. If a short sequence had 2 or
more SNPs compared to Hugo, it might really throw the computer off the
scent. Perhaps the problem is not wrong reads, so much as segments with
SNPs being discarded as ambiguous at the computer analysis stage.

It seems to me that the closer you get to the present day, there are less
SNPs reported, because having been not widely reported before if at all, the
computer programmes are less confident in recognizing them.

If there are 246 or more SNPs not yet placed on the tree by James, this
could be something significant as far as these age estimates are concerned.

John.

-----Original Message-----
From:
[mailto:] On Behalf Of Tim Janzen
Sent: Sunday, April 05, 2009 10:43 PM
To:
Subject: Re: [DNA] Comparison of Chinese, Yoruba,Watson and Venter genome
y-snps

Dear John,
If you look at James' Genome SNPs Comparisons file you will note
that he lists 188 SNPs between Rows 304 and 491. These SNPs start with
haplogroup P and go down to 3 SNPs that James suggests are private SNPs in
the HUGO sequence. Of these 188 SNPs, all but 8 are found in both the
Chinese sequence as well as in the Yoruba sequence. The 8 SNPs that are in
question are found in the Chinese sequence, but not in the Yoruba sequence.
James categorizes the 180 SNPs that the Chinese and Yoruba male
share that appear to be between the PO node and the HUGO male as follows:
P: 14
P or R?: 4
P+: 42 (derived under P)
R: 8
R+: 15 (derived under R)
R1: 8 (P294, P242, P238, P225, P286, P236, P234, P233)
R1+: 10 (derived under R1)
R1b: 1 (M343)
R1b?: 6
R1b+: 12 (derived between R1b and R1b1)
R1b1+: 10 (derived between R1b1 and R1b1b)
R1b1b: 1 (P297)
R1b1b+: 34 (derived between R1b1b and R1b1b2)
R1b1b2: 3 (S10, S3, and M269)
R1b1b2a: 2 (L49 and S141)
R1b1b2a1: 1 (L51)
R1b1b2a1a: 3 (S127, S128, S129)
R1b1b2a1a2: 1 (S116/P312)
R1b1b2a1a2d3: 1 (S139/L2)
R1b1b2a1a2d3a: 1 (S144/L20)
Z-HUGO: 3

Karafet, et al. (see http://genome.cshlp.org/content/18/5/830)
estimated the age of P as 34,000 years, R as 26,800 years, and R1 as 18,500
years. If we divide 34,000 by 180 then we get 188 years per SNP between the
HUGO male and the haplogroup P node, or approximately 6 generations per SNP.

One problem I see at this point is the fact that only 3 SNPs have
been identified at this point between the HUGO male and the L20 SNP. It
seems probable to me that there would be more than just three SNPs in this
interval. There is at least one SNP just downstream from L51, specifically
L52 (rs13304168) that I can't find in James' spreadsheet. I am not sure why
it wasn't included. Perhaps it is an unreliable SNP.
In any case, if you take the data as it stands and don't include L52,
then there are only 5 SNPs downstream from P312 and only 12 SNPs downstream
from M269. 12 times 188 is 2256 years as a rough estimate for M269 and 5
times 188 is 940 years as a rough estimate for P312. I think that this is
far too low for both M269 and P312. I have previously estimated P312 at
about 3500 years using the interclade variance method (see
http://archiver.rootsweb.ancestry.com/th/read/GENEALOGY-DNA/2008-11/12279941
55). I believe that Anatole Klyosov has estimated M269 to be about 4600
years old using a logarithmic method.
In addition to the 180 SNPs that James categorizes that the Chinese
and Yoruba male share that appear to be between the PO node and the HUGO
male, there are 246 other SNPs that the Chinese and Yoruba male share that
James hasn't yet categorized and haven't yet been placed on the Y SNP tree.
Ideally, these 246 SNPs should be placed on the next major SNP chip. Of
these 246 SNPs, 10 are shared by the Chinese and Yoruba male as well as by
Watson and Ventor. These 10 SNPs may well be eventually placed somewhere
below P312 on the Y SNP tree or they could prove to be errors in the HUGO
sequence. The rest of the 246 SNPs likely are found somewhere on the Y SNP
tree somewhere between the HUGO male and M168. Probably some of these 246
SNPs will prove to be errors in the HUGO sequence.
I think that we need more accurate complete R-M269 Y chromosome
sequences to help us sort some of these issues out. In particular, we are
going to need a highly accurate R-L20 Y chromosome sequence that we can
compare to the HUGO sequence to help us spot more errors in the HUGO
sequence. Ventor's sequence has proven to be useful, but contains a lot of
errors. Watson's sequence while very helpful is also not quite as accurate
as the Chinese and Yoruba sequences appear to be. It is possible that there
are some more downstream SNPs between P312 and the HUGO male in this file of
132 SNPs:
http://freepages.genealogy.rootsweb.ancestry.com/~ncscotts/Y-DNA/Tim%20Janze
n/W_V-shared-SNPs.htm. However, my impression at this point in time is that
at least some of those SNPs are errors and aren't going to be between P312
and the HUGO male (particularly the SNPs between positions 57392694 and
57441017) since only 4 SNPs between positions 57392694 and 57441017 were
found to be shared SNPs by the Chinese male and Yoruba male.

Sincerely,
Tim

No virus found in this outgoing message.
Checked by AVG.
Version: 7.5.557 / Virus Database: 270.11.41/2041 - Release Date: 4/4/2009
4:53 PM


This thread: