samtools mpileup Aborted (Core dumped)

Question

[rafa.rios.50]

I'm just trying to use samtools mpileup and it is aborting with out showing any error for some bacterial genomes, with the file given there are two variants and the program dies.

The error is just:

[mpileup] 1 samples in 1 input files
<mpileup> Set max per-file depth to 8000
Aborted (core dumped)

The reference file (indexed with bwa) is: https://drive.google.com/file/d/0BzPjZ1hM-XWPMmxucnpDUjdpUFE/view?usp=sharing The sorted bam file is: https://drive.google.com/file/d/0BzPjZ1hM-XWPUWw4NkdhM0E3Vm8/view?usp=sharing

The call is:

samtools mpileup -uf SaureusTCH1516_nt_genome.fasta 446.srt.bam

To obtained the sorted bam file:

First I trimmed pair end reads by quality and if adapters were found.

Second the trimmed reads are splitted in two different files, one for each end of the sequencing data, resulting in files without quality information.

Third I aligned the reads files to the reference with bwa aln and bwa sampe

Fourth I created the sam file with samtools view | sort

And finally I indexed the sorted bam file resulting in the file given in the link.

The version of samtools I'm using is 1.1 using htslib 1.1

Any hints on what I could be doing wrong, or what may be happening?

Thanks in advance

Answer 1

[Devon Ryan]

You have a corrupted alignment (for read #1 of HISEQ:185:HBB8WADXX:2:1105:8685:78337). I haven't determined what's actually wrong with its encoding (well, it's with the auxiliary tags), but you can simply avoid the problem with

samtools mpileup -Ruf SaureusTCH1516_nt_genome.fasta 446.srt.bam

[John Marshall]

This is the 64th read in the file, and I believe the preceding reads were just lucky. I think the problem occurs due to the QUAL being * and that Rafa has encountered bug #320/BAQ recalculation corrupts records without base qualities, which can alternatively be worked around by disabling BAQ recalculation (if appropriate) with -B.

[Devon Ryan]

Interesting, I guess I've never looked at the code for BAQ (I just noticed that the core dumps due to an assert in skip_aux when it's looking for the RG tag).

[John Marshall]

Indeed. Like I alluded to in the samtools-help thread linked from bug #320, tracking this down involved quite a lot of not-fun with the debugger :-)

Answer 2

[vaibhav.vsk]

I did bwa mem human alignment with single end fasta file. I tried running samtools mpileup & I'm getting same error as mentioned in thread as: Aborted (core dumped)

Command used was:

samtools mpileup -g -u -f /hg19/referenceFile/genome.fa MDS03_T_DRR001284_1111_small_fasta-quality_checked.bam > MDS03_T_DRR001284_1111_small_fasta-quality_checked__mpileup

There's no such exact error cause displayed on stdout.I tried solutions & I could get rid of this error using -R and -B options both independently. However there's difference between the variants obtained in vcf file after bcftools.

How to debug the exact cause and which fix to be used either -B or -R?

[Devon Ryan]

You can debug this by running the command in gdb. This requires that you be familiar with C, though.

[Max Ivon]

Are you sure that you have enough Memory for running samtools with this reference? This may be caused simply by the lack of PC resources. Have you previously successfully performed mpileup with this reference on your PC/server?