Bsmap Alignment very slow

[haloudashu]

No description provided.

[haloudashu]

Hi Team,

I am running Bsmap with the below command. I can see it's really slow.

bsmap
-a ${outdir}/${sample}/${sample}_cut1.fq.gz
-b ${outdir}/${sample}/${sample}_cut2.fq.gz
-d ${genomefa}
-z 33 -S 1 -s 16 -g 3 -n 1 -q 0 -f 5 -p 30 -u -r 1 -v 15
-o ${outdir}/${sample}/bsmap/${sample}_bsmapAligned.bam
2> ${outdir}/${sample}/bsmap/${sample}_bsmap_output.txt

I have 511,750,752 total reads paired. about 112 (Gb) data. I did pre processing. Filtered reads are used for align. it used 32 CPUs and 128GB RAM.
The process has been running for 137 hours and is still not complete. Can you please help with this?

Version: Bsmap 2.90

[haloudashu]

This is computing resource usage

[xi11west]

The gap mode -g is a much slower than ungapped mode. Also -p 30 is unlikely to be much better than -p 12.

…

On Mon, Jul 1, 2024 at 9:53 PM haloudashu ***@***.***> wrote: image.png (view on web) <https://github.com/xi11west/bsmap/assets/174400440/e65b544a-75dc-49b1-8a20-34ab71848a02> This is computing resource usage — Reply to this email directly, view it on GitHub <#1 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AVVMBQUHLYYISLOLGVYRBVTZKII23AVCNFSM6AAAAABKGT6G2SVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEMBRG42TQMRQGQ> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[haloudashu]

Thank you for your response. I tried setting the -g parameter to its default value and conducted tests using 14,596,415 total read pairs, which is approximately 5GB of data. When -g was set to 3, the completion time was 6 hours and 28 minutes, whereas when -g was set to 0 (the default), the completion time was only 1 hour and 2 minutes. Although this resulted in a roughly 5% reduction in alignment rate, it saved a significant amount of time. I think, under the default parameters, whether using -p 30 would result in a 50% time reduction compared to using -p 12?