GENEALOGY-DNA-L Archives
Archiver > GENEALOGY-DNA > 2007-03 > 1173433323
From:
Subject: Re: [DNA] Four key STR values are sufficient to identify the
Date: Fri, 9 Mar 2007 01:42:03 -0800
Colin, you are quite right. AH5VY is one of three haplotypes that
don't quite fit. They are otherwise a good match for Scots Modal,
when compared on the full spectrum of markers. UZ2MM was designed as
a filter based on only those STRs that are least variable within the
cluster - that is, based on the most stable STRs. Those STRs that
demonstrate the greatest variance within the cluster were
deliberately omitted, in order to cast a larger net, with the intent
of capturing as many Scots Modal haplotypes as possible to determine
the scope of variability within the cluster. Having achieved that
preliminary database, I then attempted to define the haplotype more
narrowly, based on the most minimal "bikini" haplotype possible which
might still capture the same group of results. The difference is in
how the filters are used. With UZ2MM, I have typically allowed a
genetic distance of not more than three, specifically because I have
already allowed for a wide spectrum of variability within the most
variable markers in the cluster (those that were omitted). If you
then allow for even more variability within the most stable markers,
you tend to lose the cluster altogether, & what you end up with is a
mix of Scots Modal & AMH.
On the other hand, with the bikini haplotype, I have selected just a
few of the most stable markers which are also the strongest, minimal
package that might filter out the cluster. So to use that filter, it
is important to allow a genetic distance of zero, & no more. You
immediately discovered the exceptions of course, one of the three
that I am aware of. The other two are a Talley(HVU9M) & a Smith
(WEPDN). Every filter will have a few exceptions. There are probably
also a few DYS531=11 & a few DYS391=11. Ultimately, it is the genetic
distance from the complete haplotype that determines whether it fits.
The two filters are just for collecting the database. That said, the
STRs that are most reliable in filtering for the cluster may still be
useful clues for the location of potential SNPs. I still think that
Scots Modal haplotypes should be tested for DYS717, so that we know
whether it is a reliable marker or not. Best of luck with your
research, Pat
> Pat,
> Comparing 57 markers, Ysearch ID AHV5Y is at genetic distance of 5
> from
> the Scots Modal UZ2MM yet DYS444 is 12 not 11. Do you know why in the
> first 37 markers, 10 are undefined for UZ2MM? See http://
> tinyurl.com/2dad5d
> --
> Cheers,
> Colin Ferguson
This thread: