GENEALOGY-DNA-L ArchivesArchiver > GENEALOGY-DNA > 2008-09 > 1222449537
Subject: Re: [DNA] Similarity at 23andME
Date: Fri, 26 Sep 2008 13:18:57 EDT
In a message dated 9/26/2008 9:29:21 AM Pacific Daylight Time,
> This message to those at 23andME . I would like to increase the number
> of genome sharing partner. Up to now the "compare genes" function gave a
> surprising 74% similarity with all Europeans. This figure is puzzling
> and I would like to see if it 's maintained on a larger number of
> participants (13 at this time).
> Similarity is the pair wise comparison over 577000 SNPs of how many are
> similar (same result) between the 2 genome compared. It's very basic but
> the result isn't.
I'll be your partner when my results get back, but in the meantime, I'll give
some similar stats from deCODEme. Several unrelated people have shared their
raw data files with me, and I've also tested two cousins once removed as part
of my study on the genetic hearing impairment that runs in the family. I was
particularly interested in the short arm of chromosome 1.
I compared all SNPs in that region ( ~ 45K), scoring 2 if we shared both
alleles, 1 if we shared one allele, and 0 if we shared no alleles, then calculated
the average similarity. The scores ranged from 1.56 to 1.58 for the unrelated
people, and 1.59 to 1.60 for my cousins. That's a statistically significant
difference, but it doesn't jump out at you because it's so small in absolute
I think the lesson here is that summary statistics show our fundamental
similarities but obscure the particular ways in which we might differ. If you look
at the deCODEme chromosome diagrams I prepared, you can clearly see specific
regions where my cousins and I have much greater similarity than you'd see with
Looking for simple solutions to your real-life financial challenges? Check
out WalletPop for the latest news and information, tips and calculators.