Samtools - huge mpileup file!

Question

[lcc1844]

I have a .bam file which contains exome capture data and is 10GB.

I used the following command to make an mpileup:

samtools mpileup -E -uf hg19.fa file.art.bam > file.mpileup

It took several hours of making the mpileup file and it got to over 100GB in size and my computer ran out of storage! So I obviously stopped this command running to start again. Why was the pileup so large?

My intention upon getting the mpileup file was to do variant calling using:

bcftools view -cg file.mpileup > file.vcf

Are these the right options for me?

Thank you!

[lh3]

Use pipe...

[Istvan Albert]

We're in the pipe five by five:

[lcc1844]

I stopped unzipping the file because my computer ran out of space. I must have done something wrong somewhere because the .vcf should surely be much smaller than fastq and bam!

[Devon Ryan]

Pipes are your friend: zcat file.vcf.gz | less.

Answer 1

[Devon Ryan]

A pileup file is text-based, so it's going to be large. Assuming you have the most recent version of samtools, just use samtools mpileup -Euvf hg19.fa file.art.bam | bgzip > file.vcf.gz.

[lcc1844]

Thanks very much, I ran the command you suggested and it took many hours and generated a zipped .vcf file that is 21GB. Is this not larger that what I could expect from one human exome?

[Devon Ryan]

It seems rather large to me, but you could just look at the results to see if they make any sense.