Platypus based joint variant calling

Question

[agoel]

Hi,

I am trying to use Platypus to get joint variation calling on >200 BAM files (tumour/normal tissue samples. Avg. depth of cov. is 80x).I am assigning 8 processors and have 32Gb RAM available. When I try with a couple of samples, Platypus runs fine. But with all samples, it gets stuck. The max. no. of samples I have been able to process in joint mode, is 46. I have tried restricting the search space by providing target regions (--regions param.) as well, but no avail.

Does anyone have experience in running Platypus on as many samples?

Or, with hundreds of BAM, the better option is to use GATK GVCF route?

Any advice would be of great help!

[Benn]

I am not sure, but I guess 32 GB and 8 processors is not sufficient for 100s of BAM files, for any tool available...