Verifybamid2 on long read sequencing data

[reshu23]

Dear Team,
I am using verifybamid2 on ONT data. To run verifybamid , at first I convert cram file in to pileup format by using samtools mpileup( with -B option). Then I use pileup as an input to verifybamid2. It is still time taking job. Can I generate pileup format from cram file by using samtools mpileup --positions option? Where I can provide 10K bed file provided in the resource directory to "--positions" parameter. Do you think, it makes sense? Here is the bed file, I am talking about. Thank you.

[Griffan]

Hi @reshu23, yes it should work as long as the result is still in pileup format. Let me know if you encounter any compatibility issues.

[reshu23]

Thank you. Yes, it worked. Earlier cram to pileup conversion with >=35X coverage was taking more than 17 hours. Now with options "q=20 and positions=bed file", it is taking around 4 hours 30 minutes including verifybamid run. Still, it is a long time, but I can not think of any other solution of reducing time.

[Griffan]

@reshu23 thanks for the update. Would you mind posting your example script here so that I can pin it for others who also want to process long reads BAM/CRAM?

[reshu23]

Sure. Here it is:
bedfile="1000g.phase3.10k.b38.vcf.gz.dat.bed
input="Sample.cram"