Quote:
Originally Posted by Stitcher
Reliability, Schmiability
GenomeWeb News, Tue, 24 Mar 2009, 09:10 AM PDT
Daniel MacArthur points out that 23andMe seems to be testing people for rare variants associated with Parkinson's disease and he wonders why those results (the company noted recently that it hasn't found anyone with those variations) haven't been reported back to the customers. He's guessing that they are concerned about the reliability of the results.
http://www.genomeweb.com/blog/reliability-schmiability
|
I read MacArthur's post and subsequent today posts on the site. The problem is with accumaulating enough negative controls to avoid giving a really large number of false positive results for a gene with less than 0.1% occurence in the general population.
My interest, as one with familial PD, is determining whether I, and possibly my children and their children, carry the most common of the PD-related gene mutations, vis, LRRK2 G2019S. 30-50% of carriers of this change (a glycine replaced by a serine amino acid) develop PD indistinguishable from "sporadic" or "idiopathic" disease. A few other mutations in this same gene (PARK 8) and in other genes associated with PD are much rarer. This puts them at greater risk for showing up with false positive results.
Once 23andme gets the 10,000 PDrs on file, they will be able to generate better statistics on the rarer ones, and possibly reveal additional PD-predisposing genes.