bcftools issue

[willgrand]

Folks:

I have a mouse sequencing project with multiple .bam files generated.

samtools version 0.1.19-44428cd
bcftools version 1.1-137-g1f79bbd (using htslib 1.1-124-g9b1cb94)

I tried to do the following:

samtools mpileup -g -f mm10.fa mybam.bam > mybam.bcf

The program runs for several hours then generated a 5Gb file mybam.bcf

However,when I tried to use
bcftools view mybam.bcf

it has error:

[E::hts_hopen] fail to open file 'mybam.bcf'
[E::hts_open] fail to open file 'mybam.bcf'
Failed to open mybam.bcf: Success

same error message if I use 'bcftools call -v mybam.bcf' command

anyone has the same issue and is there a solution? the file is just there but seems the program fail to see it

[GenoMax]

Can you try one of the following?

$ bcftools view ./mybam.bcf

or

$ bcftools view /path_to/mybam.bcf

Use the bcftools included in samtools v.0.1.19.

[dpryan]

Mixing different versions of samtools and vcftools is usually a bad idea. If you're going to use a new version of bcftools, use a new version of samtools too. That resolves problems like this most of the time.

[GenoMax]

Missed the fact that OP was mixing samtools versions. Good catch Devon.

[willgrand]

Thanks for the replies, I think this is due to issue with different version and potentially different htslib ...

downloading latest samtools and building it now...

NGS is a beast given all process takes long time to finish

[willgrand]

using latest samtools:

samtools mpileup -g -f mm10.fa mybam.bam > mybam.bcf

got message: [fai_fetch_seq] The sequence "1" not found
[fai_fetch_seq] The sequence "10" not found

googled the error message but got no informative information, should I be concerned? mybam.bam has its mybam.bai file in the same folder...

[dpryan]

Are you sure the chromosome names in the BAM file and fasta file match?

[dpryan]

BTW, if the answer is yes, then try remaking the fasta index (samtools faidx).

[willgrand]

Thanks, I think that is the issue here... will look into it and get back to you

BTW, if the answer is yes, then try remaking the fasta index (samtools faidx).

[Jessica_L]

I'm having the same issue with
"[fai_fetch_seq] The sequence "1" not found"
across all human chromosomes when I'm running mpileup. I am about 99% certain that chr names aren't identical in both files, but I'm not sure how to fix that. Editing my human genome fasta file isn't really an option...

I'm on a linux system, using samtools/bcftools v1.2 and I tried remaking the fasta index already. Has anyone else had this problem before?

[dpryan]

If you're fairly certain that the chromosome names differ between the files then you'll need to change one of them. Changing them in the BAM file just require "samtools view -H foo.bam", changing the names (but not the order!) to be correct, and using "samtools reheader". Alternatively, just download the matching fasta file (presumably from Ensembl or Gencode, given the error message).

[Jessica_L]

If you're fairly certain that the chromosome names differ between the files then you'll need to change one of them. Changing them in the BAM file just require "samtools view -H foo.bam", changing the names (but not the order!) to be correct, and using "samtools reheader". Alternatively, just download the matching fasta file (presumably from Ensembl or Gencode, given the error message).

Thanks! I'm going to give the samtools reheader a try first!