Re: [DIYbio] 23andMe exomes

Hi,

On May 17, 2012, at 19:00 , Madeleine Ball wrote:

> The Personal Genome Project has a couple participants that have donated their 23andme exome data to be public through our IRB-approved open consent process. You can find the VCF files linked as "source data" on the interpretations here: http://evidence.personalgenomes.org/genomes You can also find them linked on individual profile pages, e.g.: https://my.personalgenomes.org/profile/hu97DB4A

this is great. I totally forgot to look at the PGP-data for this, my bad. m(

> The VCF format might be frustrating to you in that it fails to distinguish between "not sufficiently covered to make a genotype call" and "matches the reference genome". (It is theoretically possible to report this, but to date it's only been done in a base-by-base manner, which results in ridiculously large files.)

This sounds bad. So the 23andMe-files also report those SNPs which haven't been called as a match to the reference? An example out of one of the VCF-files of the PGP:

…
1 753405 rs61770173 C A 203.64 […excluded…] GT:AD:DP:GQ:PL 0/1:17,16:34:99:234,0,489
…

So in this case the genotype is given as 0/1, an unphased C/A, but it also could be that the allele which is given as C wasn't called at all? But there is also information about the read-depth (DP), Genotype-Quality (GQ) and a phred-scaled likelihood (PL), so I could use those to determine how accurate the genotype-calling was, couldn't I?

I should mention: For openSNP we currently only aim to read the known SNPs (those with Rs-IDs).

Cheers,
Bastian

--
// Bastian Greshake
// Zehnthofstraße 36
// 55252 Mainz-Kastel, Germany
// cell: +49 176 213 044 66
// web: www.ruleofthirds.de






--
You received this message because you are subscribed to the Google Groups "DIYbio" group.
To post to this group, send email to diybio@googlegroups.com.
To unsubscribe from this group, send email to diybio+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/diybio?hl=en.