GENEALOGY-DNA-L Archives
Archiver > GENEALOGY-DNA > 2006-02 > 1140815473
From: <>
Subject: Re: [DNA] SNP Test Necessary: WEBB-BIRNBACH R1b SuperX Haplotype Study Report
Date: Fri, 24 Feb 2006 21:11:13 -0000
References: <BKEPIIDHHKEPCMDIEBKBKEOFCOAA.andrew.en.inge@skynet.be>
Andrew
that's precisely the point that John's missing.
Yes, we can prove without any doubt that recurring SNPs must exist in large
numbers because the number of human males alive today exceeds the number of
sequenceable bases of the Y chromosome by a factor with several 0s in it,
and each one of us will have a SNP type mutation with high probability in
from the last two or three generations of our paternal line.
BUT
Currently whilst we are testing a limited number of SNPs the chance of
someone with a recurring SNP (e.g. being S28+ but not by descent from the
presumed original founder) actually coming forward to be tested for that
same SNP are infinitesimal.
I may by chance be M170+ (very small chance - 1 in a million?) but I am not
going to be tested for that SNP because I have no reason to believe I am in
hg I and in fact I now know I am in hg R1b1c6. So it's not an issue.
Sometime in the near future we will start sequencing larger sections of Y
chromosome. Then we will start detecting recurring SNPs occasionally. But
they will very easy to identify - because whilst the chances of being
"identical by state" at one single base are small but not too small, the
chances of two men being identical by state at *two* different SNP sites
must be infinitesimal. The rest of the sequence will show very clearly when
a shared mutation cannot be due to a shared ancestor. So it still won't be
an issue.
The issues - as John points out in his last sentence - are going to be down
to lab errors, contamination or poor quality samples, and SNPs like P25
whose mutational properties mean they are less likely to give reliable
haplogroup determinations.
Gareth
----- Original Message -----
From: "Andrew and Inge" <>
To: <>
Sent: Friday, February 24, 2006 8:37 PM
Subject: [DNA] SNP Test Necessary: WEBB-BIRNBACH R1b SuperX Haplotype Study
Report
> John Chandler wrote:
>
> There is a point you're missing. We often speak of SNPs as if they
> were unique events, and, indeed, for population genetics they can
> usually be treated as such. However, for genealogy, they just are
> not. We can say quite confidently that there are right now living in
> the world hundreds of R1b males who are S28+ due to recent parallel
> mutations, rather than inheritance from the "original" S28 founder.
> Obviously, mere hundreds of males can be safely ignored for the
> purposes of studying the migrations of whole peoples, and the
> likelihood of finding one of these males in your own surname project
> is very small -- but "small" is not "zero". And this is just the
> beginning. The chances of a lab error or contaminated DNA sample
> or other reporting error are also small but non-zero.
>
>
> John,
>
> Surely the coincidence of both being M269 narrow down the chances of have
a
> parallel SNP a little bit?
>
> Best Regards
> Andrew
This thread: