Archiver > GENEALOGY-DNA > 2003-05 > 1053528323

Subject: Re: [DNA] DNAPrint "failure" results and an possible anomoly (TT in a marker ...
Date: Wed, 21 May 2003 10:45:23 EDT

I am pasting cogent explanations from Matt Thomas of DNAPrint for these two

There really is no anomaly at this marker and it all has to do with the
complentarity of DNA.  Yes, this SNP is listed as G/A in the NCBI database
but that refers to only one strand of the DNA.  On the other strand this SNP
is C/T!  We've found that using the other strand (CT) was more efficient and
handled certain chemistry issues better.  In case people are interested a TT
= AA, TC = GA, and CC = GG. 
In regard to failures we've done a lot of testing on the issue of failures. 
We've taken genotypes that result in a wide range of scores and then
introduce failed markers into them at different rates.  For example we've
introduced a single fail at every marker in turn, and we've introduced
increasing numbers of random (from 5 to 35) fails to find the threshold of
when there are too many fails (significant changes in the scores).  Some
people are more tolerant to fails than others in their scores but we've found
a cut off of 8 to 10 fails is very conservative and still provides accurate
results.  Anyone with more than that number of fails gets re-tested the
following week and if there are more than 15 fails we extract their other
swab as well. 
Hope this clears up things.
Matthew Thomas Ph.D .         Senior Scientist
DNAprint Genomics               

This thread: