GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2010-03 > 1269347847


From:
Subject: Re: [DNA] Re : Usefulness of SNP tests
Date: Tue, 23 Mar 2010 08:37:27 EDT


I have some real-life data for two parent-child trios, and your synthetic
child is right on target. For one set, the child's haplotype (set of alleles
on one chromosome) could be determined for 84% of the SNPs; the other set
was 94%. From there, you can split the parents' genotypes into two sets
(passed on vs not passed on). Long stretches will correspond to haplotypes, but
you won't be able to determine the crossover points. I still found this useful
in my family study of an inherited hearing impairment.

PSOR1 and PSOR2 are not genes -- they are regions containing several genes.
No SNPs have been assigned an rs (reference SNP) number. You might be
interested in this article about PSOR1:

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC379192/pdf/AJHGv71p554.pdf

Ann Turner


In a message dated 3/22/2010 11:11:53 PM Pacific Daylight Time,
writes:

> Sorry for being slow to reply. Last weekend I had another look at the
> Mendels on 23&Me as a dry run to try to work out what else might come out
> of
> testing. Being smallish I chose chromosome 17 and put their downloaded
> results into a spread sheet. From a total of 15,310 SNPs approx 10,500
> were
> homozygous, 4,800 hetero and 40ish discarded. I then went on to create a
> fictional child made up from the parents SNPs and found that 1842 sites
> were
> double heterozygous ie unresolvably ambiguous. Meaning that it was
> possible
> to correctly assign almost 88% of the alleles to each chromosome of the
> child. With that in mind, I'm also thinking that by comparing my 2 sons
> results it should be possible to assign specific SNP sequences to both my
> own and wife's chromosomes based on regions of correspondence between
> them.
> Does this make sense? And does 23&Me or FTDNA do this for you anyway given
>
> only 2 generations?
>
> As I said earlier I'm not hopeful of too much coming out of the medical
> SNPs. If I'm wrong please let me know. (OT: 23&Me say they can provide
> information on psoriasis susceptibility. But when you look at the Raw
> Results for the Mendels 2 of the major suspect genes PSORS1 on Chr 6 &
> PSORS2 on Chr 17 have no SNPs listed because the genes can't be positioned
>
> they say. I do hope we will get *all* the raw data in future).


This thread: