mosdepth coverage so slow - fast way to get coverage of restricted regions WES
0
1
Entering edit mode
12 months ago
cocchi.e89 ▴ 150

I am trying to check the coverage in few regions in some WES samples. I really need very few regions, this is the BED file:

2   198266460   198266618
2   198266703   198266860
2   198267274   198267556
2   198267667   198267765
2   198268303   198268494
9   5069919 5070058
9   5073692 5073791
9   5078300 5078450
17  74732949    74732969
20  31021081    31025147


I tried bedtools genomecov but it was slow, I then searched around and found few posts speaking of mosdepth that, based on their README, should take:

mosdepth --by capture.bed sample-output sample.exome.bam

For a 5.5GB exome BAM and all 1,195,764 ensembl exons as the regions, this completes in 1 minute 38 seconds with a single CPU

and mine command on a 8.6 GB WES .bam is already taking more tahn 25 minutes! Here the command:

mosdepth --by ${restrictedBed} restricted-output${fileGenome}

and similarly with option -t 3 that should use 3 threads.

Am I not getting something? Is there a fast way to get the coverage of a limited set of regions from a BAM??

Thank you very much in advance for any help

mosdepth coverage exome WES • 526 views
1
Entering edit mode

Check with top if the tool is running at 100% CPU or if any bottlenecks are slowing it down.

0
Entering edit mode

Are you doing the analysis on a laptop? Time comparisons are not meaningful unless there is metadata for hardware provided. Trying to compare a 15W U series laptop CPU to a 150W Xeon monster is unfair on many levels.

0
Entering edit mode

Sorry, on a server machine with 48 CPUs and 128GB RAM

0
Entering edit mode

And the file is not on a slow remote disk or something? If IO or network are limiting then fast tools won't make much difference