GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2010-02 > 1266711070


From:
Subject: Re: [DNA] FTDNA will introduce Family Finder (cM limit)
Date: Sat, 20 Feb 2010 19:11:10 EST


It sounds like it you might be seeing lots more matches, actually, since
there are some regions of the chromosome where 500 SNPs are in a region with a
low recombination rate. This would be more akin to deCODEme's display at 1
Mb or less, where you'll have many regions of similarity with every
reference sample in the Human Genome Diversity Project panel.

23andMe did use 500 SNPs in their original work presented at the American
Society of Human Genetics conference last fall, but increased it to 700 (in
combination with the cM threshold) to rule out a few problematic sites, I
believe. Presumably the FTDNA project administrators will have some guidance
about the significance of matches in the low cM range.

It's a bit hard for me to tell at this stage of the game, but I think
23andMe only predicts a maximum likelihood score for 6th to 10th cousins after
you share with a person. My lowest ranked matches in the 5th cousin bin cover
the range 4th-10th.

Ann Turner

In a message dated 2/20/2010 12:42:13 PM Pacific Standard Time,
writes:

> Ann, looks to be a combination of both with 500 SNP in series being the
> criteria for FTDNA. I am reasoning that the 500 SNP criteria will show more
> matches, no?
>
> I am told that Family Finder has the exact same resolution as 23andme but
> their interpretation past 5th cousins is "Distant Cousins" rather than 6th
> to 10th cousins or Distant Cousins.
>
> Since Project Admins. will have extra tools to locate descendants of a
> distant ancestor, it looks like showing centimorgans for all of matches that
> are from >1 to 10+ with SNP info per 'block' is part of this feature.
>
> Haven't seen real life examples.


This thread: