Reducing memory consumption

Question

Reducing memory consumption

Opened this issue 3 years ago · 1 comments

Hi,
I am trying to run the preprocessing step of nebula on a BAM with long reads, in order to later on call it on short read BAMs. However, I am running out of memory: with the entire BED file, 450GB of RAM was not enough.
If I take only a fraction of the BED it works, but I am not sure how to then do the merge. Will Nebula consider it as multiple samples?
Thanks

Answer 1 · 2021-09-17T21:05:35.000Z

Hi,

Nebula is not meant to be used and has not been tested with long read (PB or ONT) BAM files with high-error rates. This may result in excessive amounts of kmers being extracted.