search

natsuhiko / rasqual

Robust Allele Specific Quantification and quality controL
37 stars 20 forks source link

Estimated computation time too long #48

Closed liesaweigert closed 3 years ago

liesaweigert commented 3 years ago

I am currently using rasqual for ATAC data with features extended 50 000kb around each peak. Unfortunately for most of the features the computation is canceled, because the estimated computation time is too long.

Estimated computational time is too long for 8:29443453:29544444 (Sample Size=84, NofrSNPs=249, NoffSNPs=215)...aborted. Estimated computational time is too long for 8:29466273:29567135 (Sample Size=84, NofrSNPs=233, NoffSNPs=200)...aborted. Estimated computational time is too long for 8:29472642:29573108 (Sample Size=84, NofrSNPs=227, NoffSNPs=191)...aborted. Estimated computational time is too long for 8:29530196:29630965 (Sample Size=84, NofrSNPs=210, NoffSNPs=185)...aborted. Estimated computational time is too long for 8:29544525:29645849 (Sample Size=84, NofrSNPs=201, NoffSNPs=185)...aborted. Estimated computational time is too long for 8:29607788:29708459 (Sample Size=84, NofrSNPs=195, NoffSNPs=180)...aborted. Estimated computational time is too long for 8:29611621:29712319 (Sample Size=84, NofrSNPs=198, NoffSNPs=183)...aborted.

Is there a way to fix this or somehow parallalize the computation?

natsuhiko commented 3 years ago

Hi,

You can set '--force' option to ignore this warning. You can also parallelise the computation by set '--n-threads' option.

Although it seems the number of feature SNPs is too many. For ATAC-seq, we usually see 3-4 SNPs per peak. I would recommend to revise your command first.

Natsuhiko