GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2005-07 > 1121103292


From: (John Chandler)
Subject: Re: [DNA] upgrading after a 25 marker test
Date: Mon, 11 Jul 2005 13:34:52 -0400 (EDT)
References: <IGEOKAGLHNEKPCKPADIGMELNKCAA.bbailey.lowedna@baileyconnection.com>
In-Reply-To: <IGEOKAGLHNEKPCKPADIGMELNKCAA.bbailey.lowedna@baileyconnection.com>


Bill wrote:
> However, STRS are like walking thin ice, when deep level
> SNP testing would put us on solid ground once and for all.

What you're completely ignoring is the ambiguity factor. SNPs are *not*
unique events, even though it is convenient to pretend they are. This
ambiguity doesn't matter for SNPs at present because SNPs have no
genealogical applications, but a future situation with many more SNPs
available for testing would have correspondingly more ambiguity.

Note that the genealogical power of a set of DNA tests is roughly
proportional to the sum of the mutation rates of all the markers
included. *That* is why more markers are always better and why fast
markers are desirable (up to the point where they're likely to mutate
in the time span of the family trees under study). Since SNPs mutate
about 10,000 times slower than STRs, and there are about 50 STRs
commercially available, it quickly follows that SNP testing won't
become as useful to genealogy as STRs until there are about a million
SNPs. (In the meantime, the number of available STRs will grow, too,
but not by a lot.)

NOTE: this happens to be the same number that I have demonstrated
elsewhere to be the minimum number of SNPs necessary to set up a
haplogroup scheme with genealogical potential, but it is *not* the
same in principle. That fact that it roughly matches says something
very profound about haplogroup schemes in general -- there is no need
to look for the optimum SNPs for creating a "phylogenetic" tree because
the genealogical discriminating power will be the same regardless. In
short, *any* million base pairs will do, as long as they aren't located
in genes or other sequences with genetic functions. Think of it! We
don't actually have to wait for whole-chromosome sequencing!

John Chandler

This thread: