samtools mpileup output only contains vcf header

Question

[s102060125]

Hi, I tried to call variant for creating consensus sequence with the following command:

samtools mpileup -C50 -uf $ref_genome $sorted.bam > output.mpileup

but the output contains vcf headers only, no variant was called, and no error message shown. The input bam file was produced by 'bwa mem', with default setting, and was sorted and indexed. The reference genome was also indexed. The commands works properly with a test data I downloaded from SRA. So I really don't know what went wrong. Also I can call variant from my bam file using GATK, so there must be variants presented in my sample. samtools version is 1.7, bwa version is 0.7.1

Thanks!!

Answer 1

[Pierre Lindenbaum]

the output is not a VCF file but a BCF (binary VCF file), that is why you won't see anything but the header.

$ file output.mpileup
output.mpileup: data

$ file regular.vcf
regular.vcf: ASCII text, with very long lines

you'll need to convert the BCF to VCF with bcftools view

[s102060125]

I may not make my point clear. There's nothing under the headers in my mpileup output, not even binary stuff. It's different from the test output I generate with the same command but using SRA data. Also, my 'output.mpileup' didn't return anything but vcf header in the downstream command 'bcftools call', while my test data worked fine.

[Pierre Lindenbaum]

ok, I see

Answer 2

[swbarnes2]

I wonder if the issue is with the chromosome naming...maybe there is a discrepancy between the bam and the ref that GATK can handle, but samtools cannot.