search

mskcc / facets

Algorithm to implement Fraction and Copy number Estimate from Tumor/normal Sequencing.
144 stars 67 forks source link

snp-pileup input vcf #89

Open janka2012 opened 6 years ago

janka2012 commented 6 years ago

Hi! I have WES with ~30x coverage, matched normal-tumor samples and I am interested in somatic mutations. Unfortunately, it is not clear for me, what kind of vcf file should I use as an input for snp-pileup. If it can be an unfiltered vcf file directly from let's say Mutect or if I should somehow filter my variants (e.g. only somatic) and use filtered vcf file as input for snp-pileup. Should I use only data for tumor sample? (I have two columns in resulting vcf file, one for normal sample and one for tumor smaple). When I am filtering only somatic variants, I am getting like 100-500 variant and with this data, I am getting weird results - the number of segments is the same as the number of chromosomes (gbuild = hg38). I've tried also to play with ndepth and cval parameters, but the results were mostly the same. Thank you very much for suggestions.

veseshan commented 6 years ago

FACETS is a copy number algorithm. The VCF file used should be population germline SNP data for instance dbSNP. FACETS may not work or give you unreliable answers if a VCF from Mutect is used. Please read the vignette and/or manuscript to familiarize yourself with what the algorithm does.

janka2012 commented 6 years ago

Oh, now I get it. I was so deeply focused on the other parts of analysis, that I totally missed this, even if I went through your publication and vignette at least 2 times. Once again, thank you.