From: John Taylor <taylorjs@uv...> - 2013-06-14 01:42:56
I understand that poorly-mapped reads would provide unreliable data about SNPs and that the default MQ=10 (in the vcfutils.pl step) makes sense for this reason. But, let's say you want dp (read depth data) for regions of a genome where there's variation in sequence copy number. If the ref seq has two copies of a sequence and the read data come from a genome with four copies, you won't know it (from a regional increase in dp), because the reads that map to that region will be excluded. In an attempt to fix this, I set MQ=0. But, there are still SNPs/reads in the onscreen tview output that do not show up in the .vcf file (tab delimited table). Can anyone tell me why not? That is, what step/parameter in the bcftools + vcfutils.pl analysis (that generates the .vcf file) is eliminating reads I see in the samtools tview __.bam __.fa output?
Get latest updates about Open Source Projects, Conferences and News.