ERROR: [bcf_sync] incorrect number of fields..

[vinodhsri]

Has anyone encountered this error ('[bcf_sync] incorrect number of fields (0 != 5)') and found a solution already:


vcfutils.pl splitchr -l 50000000 ucsc.hg19.fasta.fai | xargs -i echo samtools mpileup -I -C50 -m3 -F0.0002 -DSuf ucsc.hg19.fasta -r {} -b bam.list | bcftools view -bcvg - /> part-{}.bcf

[bcf_sync] incorrect number of fields (0 != 5) at 0:0
[afs] 0:0.000
xargs: echo: terminated by signal 13

Appreciate your advice.

vsri

[sflink]

I get the same error message with a command like:

nohup samtools mpileup -uf ref.fa s1.bam s2.bam

but it runs fine without the 'nohup'. Have you tried the command sequence without the pipes?

[Mark.hz]

I ran into this problem as well, using mpileup/bcftools (samtools-0.1.18) to call variants from 19 bam files.
I willl try to run it wiiltout nohup.

[Mark.hz]

Still have this problem when doing bcftools view for a bcf file (mpileup generated):[bcf_sync] incorrect number of fields (6 != 5) at 7:1396330564
Any help?

[colindaven]

This is weird. I've used Samtools a lot before on bacterial references and never had this problem, yet on some human exomes it happens. Mpileup runs fine and then I repeatedly get

[bcf_sync] incorrect number of fields (0 != 5) at 0:0

Reinstalling/updating samtools and bcftools doesn't seem to help.

[brentp]

usually this happens if your bcf file is truncated.

[colindaven]

Actually, I think it wasn't truncated, but something added by nohup.

When I don't use nohup to run the mpileup command bcftools seems to work fine.

[Mark.hz]

I still got the error message without nohup: "[bcf_sync] incorrect number of fields (6 != 5) at 7:1396330564"

[salturki]

I just had the same error and by adding '-u' option to 'samtools mplieup' part of the pipeline , this solved the issue in my case.

[kweber2]

Make sure you're using the correct reference genome

In case anyone else encounters this same problem ("[bcf_sync] incorrect number of fields (6 != 5)..."), here's another possible solution:

I was having this problem with multiple files, but then I realized I was using the wrong reference fasta as the argument to the mpileup -f option. It wasn't the same reference fasta that was used to create the bam files. When I switched to the correct one, the problem went away.

[wangli]

I also had the problem when using nohup. Without nohup. it goes well.

[mmartin]

The problem occurs when nohup is used with a Unix pipe.

I've just sent this reply to the samtools-help mailing list.

nohup merges stdout and stderr of the command that it runs. bcftools then
doesn't know what to do with the input data since it also contains messages
such as "[mpileup] 1 samples in 1 input files".

To avoid the problem, you can run the command like this:

nohup samtools mpileup ... 2> stderr.txt | bcftools ... - > out.raw.bcf &

But I strongly suggest you try the 'screen' program. It is less fragile and
much more powerful than nohup. This is the briefest introduction I could find: http://www.askbjoernhansen.com/2006/...ix_screen.html

[Elsie]

I'm getting this error with this command
samtools mpileup -D -S -u -f hg19.fasta child.bam father.bam mother.bam | bcftools view -vcgT trioauto -s sample.txt - > samtoolsCLR.vcf
The resulting vcf file just has the INFO and FORMAT ## lines, then:
#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT
[bcf_sync] incorrect number of fields (0 != 5) at 0:0
[afs] 0:0.000
I am using the correct fa reference file.
My samples file looks like:
child.bam
father.bam
mother.bam

Any advice appreciated.

[Elsie]

Solved my own problem, simple typo in samples file I missed.

[Sacrolfur]

I had the same problem. I think the problen is that bcftools works with bcf-file but you don't specify for samtools to output bcf. So you should add to samtools mpileup command argument -g.

Hope this was helpful.