qfy.py killed about hap.py

[pkrusche]

What is the memory available on your machine (or to the job when submitted on a queue) / Is is possible that quantify runs out of memory? Especially when using large region bed files, memory requirements can go up.

[teodora-aleksic]

It's got 16GB of RAM. I can try and run it on a larger machine. I'm looking at the memory usage during the execution and it's bellow 2GB during those 2h. It only peaks at the last 2-3 min and then fails.

[pkrusche]

If you run with --verbose, and --keep-scratch + --scratch-prefix <DIR> you can see if quantify produces a (possibly large) TSV file which then eats up memory when being loaded for postprocessing. One possibility is that your bed file contains a fourth column, which will produce a very large number of stratification regions. If this is not intended, speedup + memory reduction can be achieved by supplying a bed file that only contains chrom, start, end columns.

[teodora-aleksic]

It does contain a fourth column. I'll remove it and rerun with the options you suggested.

[teodora-aleksic]

Yep, it was the fourth column's fault. It finished under 2 min now. Thanks!