Samtools mpileup

[bioinf newbie]

I am using samtools mpileup on my data.
I have used bwa for alignment of my NGS whole genome:
bwa sampe ref.fa read?.sai read?.fq.gz > aln.sam
After conversion to bam, and sorting using samtools I have used mpileup.
But it is not giving the desired output. This is the output i get.

[mpileup] 1 samples in 1 input files
<mpileup> Set max per-file depth to 8000
[afs] 0:0.000

Thanks in advance

[AWeller]

The output you get is just the usual report samtools gives while running, not an error message.

Nobody will be able to answer your question with this little information, sorry.
Which command did you use for samtools mpileup?

Cheers
Andreas

[swbarnes2]

Those steps sound right, did you do any eyeballing to make sure nothing's gone horribly wrong in your previous steps?

Have you glanced at the sam? Have you looked at your .bam on IGV?

Does your ref.fa.fai look okay? Does the mpileup output file have anything in it at all?

[bioinf newbie]

The output from mpileup is empty.
The command I gave for mpileup is :
[root@B Picard_processed]# samtools mpileup -uf /mnt/data/Genomics/Gaurav/Genome/Broad_Homo_sapiens_assembly19.fasta 47_singles.sorted.bam | bcftools view -bvcg - > 47.raw.bcf

I checked my sam and bam files. they look ok..so is my ref.fa.fai

But my alignment files from bowtie2 goes through mpileup without any problem.
This issue is for the aligned files from bwa.

[AWeller]

You tell bcftools to only give you variants, maybe something is wrong with the snp calling, e.g. your quality is so low that it doesn't call any snps or there just are no snps?
Does samtools produce meaningful output if you just let it give you the raw basecalls (without using bcftools)?

[nanto]

mpileup makes some tricks to people. I had so many problems with it, so I downloaded samtools version with pileup option. There should be no problems then.
Try samtools 0.1.15

[swbarnes2]

I'm a little surprised that mpileup command works. I would think it would make a pileup file, not something suitable to put through bcftools.

But looking at the pileup is also a good thing to do.

[nanto]

it can generate pileup format if you're interested in it. I'm using pileup format, cause I got used to parsing it in shell script :P But I was also doing some work with bcftools, but it's not very useful to me, cause SNPs weren't my concern

[bioinf newbie]

This is strange!
I had the same sample aligned using bowtie2 and bwa.
For the bowtie aligned sample, mpileup works well..
[mpileup] 1 samples in 1 input files
<mpileup> Set max per-file depth to 8000
[bcfview] 100000 sites processed.
[afs] 0:81900.939 1:9681.844 2:8417.218
..........................
but for the bwa aligned data:
[mpileup] 1 samples in 1 input files
<mpileup> Set max per-file depth to 8000
[afs] 0:0.000
[root@B samtools-0.1.15]#

Because it is the same sample, it cant be the quality of data that is causing the problem here right?
tried samtools 0.1.15 pileup too, to no avail.

[nanto]

like i said mpileup makes some tricks, and also I had problems with bwa-sw with long reads or large files (thought it was core dump but even on a big machine it failed) so i started using pileup option and this works quick.

Look for samtools version 0.1.15 or older, they have all of them available on their download site.

EDIT:
You can also download galaxy and try mpileup from there (if you have confidential data use local galaxy at 8080 port)

[bioinf newbie]

yes, i tried pileup from samtools version 0.1.15
no change in output..
[afs] 0:0.000
[root@B samtools-0.1.15]#

[nanto]

so you have to dig in data from bwa.

Check if reference is indexed and you are using this reference file in mpileup, use sorted bam file, make sure you have generated bam file with headers, count lines in sam file to know if any reads get lost, compare with flagstat bam file and bam sorted file, if anything didn't go wrong on this step, and write a message to samtools mailing list, sometimes they response with something more than: blablablablabla do it yourself

[bioinf newbie]

got it figured out. looks like it was my ref.fa.fai that was "corrupted". the one i was using was downloaded from a site. generated a new index file for reference. all is well.
many thanks for the help.