mpileup on whole genome bam file

[adrian]

Hi:

I am writing to know users experience about mpileup snp calling with human whole genome bam file.

my 'de-duped' sorted bam file is ~55GB in size.

When I ran mpileup on these files, it has been running continuously for 4 days on a cluster node with 16 processor 64gb node.

For some bam files:

the output file of the node job has this following information:

[bcfview] 26800000 sites processed.
[afs] 0:91696.193 1:4741.654 2:3562.153
[bcfview] 26900000 sites processed.
[afs] 0:89785.449 1:6797.768 2:3416.783

Similarly the error output file has this information:

Warning: no access to tty (Bad file descriptor).
Thus no job control in this shell.
Terminated

I have two questions in this context:

1. Is my process complete or this is a genuine error that mpileup is not finished.

2. What is the better practice of calling SNPs using mpileup for human whole genome sequence data (where a bam file is over 50GB in size - if size is a concern).

Thanks
Adrian

[swbarnes2]

bcfview] 26800000 sites processed.
[afs] 0:91696.193 1:4741.654 2:3562.153
[bcfview] 26900000 sites processed.
[afs] 0:89785.449 1:6797.768 2:3416.783

That part looks normal. Isn't TTY your command window?

Why don't you try making a smaller .bam with samtools view, and seeing if mpileup can get through that?

[adrian]

thank you. that is a really good idea. Since I have two replicates I will separate every chromosome for both files and run mpileup chromosome by chromosome.
How can I have a bam file for chromsome with header?


Here is what I did (just for chr21)

samtools view s6.ddup.bam chr21 > s6.chr21.ddup.bam


-h print header for the SAM output
-H print header only (no alignments)


Can I do this? Is this valid?

samtools view -bh s6.ddup.bam chr21 > s6.chr21.ddup.bam


thanks
Adrian

[svl]

samtools view -bh s6.ddup.bam chr21 > s6.chr21.ddup.bam

That does the job indeed. Afterwards index the new BAM file (samtools index) and you're good to go.

I get the tty warning as well, doesn't seem to be any problem..