jodyphelan/pathogen-profiler

make bwa-mem output reproducible upon changing number of threads

Closed this issue · 1 comments

According to lh3/bwa#121, the bwa-mem output is not stable upon changing the number of threads.

Maybe it would be good to set -K 10000000 as suggested by lh3 to make the results reproducible here:
https://github.com/jodyphelan/TBProfiler/blob/57d54be5069aa2c34c463d4576a55240dab17f9f/pathogenprofiler/fastq.py#L61

Not sure if it also affects bwa-mem2.

Hi @pmenzel,

Thanks for letting me know, I'll put this in.