SNP calling in a "case-control" setting

moranmoran The BroadMember

Hello Geraldine,

I wish to look for SNPs that are unique to my case samples, and do not appear in my control samples. To be more precise, their enrichment in the case samples should be significant. Something like a fisher's test.

I have many bam files of control samples, and many of cases. How can I run this?

This is bacterial DNA, so as far as we know, there is a single copy of each chromosome.

Thanks!

Best Answer

Answers

  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie admin

    First, call them all together with UnifiedGenotyper (with -ploidy 1). Then you'll need to work with the variant selection and evaluation tools to generate numbers of presence/absence (which you can do per set of samples). Then run the numbers through your favorite stats package.

  • moranmoran The BroadMember

    When you say all together you mean all samples together? both cases and controls?

    Also, I see that UnifiedGenotyper requires a dbsnp, what if I don't have one?

    thanks!

Sign In or Register to comment.