When I ran it on several bam files in parallel it seg faulted. On error message indicated insufficient virtual men (this is on a 48Gig machine). So rerunning a single process now I note that it eats up more and more memory while reading in the READS. Could anybody give an indication of how the process scales? Is there a way to calculate how much mem I will need for a given number of READS in the sam/bam?
Just wondering if I should move this step to a cluster where I might use 120Gigs or even more.
Thanks
Philipp
by Philipp Schiffer on 29.01.2014 - 10:23
solved
Hi!
Think I solved the issue. There was a problem with the READS not being indexed with -1/-2, which I guess will give a problem with filterBam.
Thanks!
Philipp