samtools mpileup fails

[bair]

Hello all,

I'm using asmtools mpileup for snp calling from eight bam files, I get this following error message

bam2bcf_indel.c:181: bcf_call_gap_prep: Assertion `n_types < 64' failed.
[bcf_sync] incorrect number of fields (7 != 5). Corrupted file?
[afs] 0:16195.241 1:1.758 2:87.006 3:1.843 4:16.294 5:0.378 6:3.137 7:0.193 8:4.107 9:0.037 10:0.005 11:0.000 12:3.000
13:0.000 14:2.008 15:0.118 16:10.874

Any helps?

Thanks

[lh3]

I thought this should never happen, but it happened... This says that at a position there are more than 64 types of indels. For now, you may use "-I" to disable indel calling.

[bair]

I thought this should never happen, but it happened... This says that at a position there are more than 64 types of indels. For now, you may use "-I" to disable indel calling.

Thank you heng,

But I still get outputs, there are SNPs and INDELs in vcf file, can I trust them? Does samtools ignores the problem INDELs and complete the job, or stop running?

[lh3]

No, your results are truncated.

[bair]

No, your results are truncated.

Yes, you are right. I get more variants with -I option on chr:MT.

[gaffa]

I am also getting this error. I wonder if it would be possible to somehow sort out positions with excessively many indels before running mpileup.

[Haiqing]

Hi! lh3,

I got similar error:
[bcf_sync] incorrect number of fields (0 != 5). Corrupted file?
[afs] 0:0.000

What does this mean?

Thanks

[Haiqing]

Another error message when do the first step for SNP call using samtools:
samtools mpileup -uf reference.fna sample.sort.bam |bcftools view -bvcg - > var.raw.bcf
[mpileup] 1 samples in 1 input files
samtools: bam_plcmd.c:596: group_smpl: Assertion `id >= 0 && id < m->n' failed.
Aborted
:-(

Any suggestions?

[lh3]

You have read groups undefined in the header

[Haiqing]

Thanks! It works now!

BTW, I also tried the old "pileup" for SNP/INDEL call with my deep sequencing data (50000+). Somehow, the pileup out only count 8000 read depth (column eight of the pileup output) for the first reference base. And then add one more read for every next base(like 8001 for second reference base, 8002 for third reference base, and so on). Any suggestions(I already give the -d60000 option to increase the max read depth. It works for mpileup, but not for pileup)? Thanks.

[oudacontrol]

How did you fix the read group issue? I am getting the same error: "bam_plcmd.c:596: group_smpl: Assertion `id >= 0 && id < m->n' failed" ...

[Haiqing]

I think I just removed these @RG header from SAM file(I do not use the read groups in my case).

[Hkins552]

Is there anyway to avoid using the -I option? I would still like to call the INDELs on my file. My error is:

[bcf_sync] incorrect number of fields (6 != 5) at 0:33736084
[afs] 0:27544.812 1:16105.314 2:36880.874

[jamminbeh]

How can I implement the -I option? I've tried:

/u/local/apps/samtools/0.1.11/samtools mpileup -I -uf hg19.fa accepted_hits.bam

and

/u/local/apps/samtools/0.1.11/samtools mpileup -ufI hg19.fa accepted_hits.bam

and I still get the same error:

[bcf_sync] incorrect number of fields (8 != 5). Corrupted file?

[aatu18]

Hi,

How did you solve this error, samtools: bam_plcmd.c:596: group_smpl: Assertion `id >= 0 && id < m->n' failed ? I am repeatedly getting this error! I have a merged and sorted bam file, which I am using to call SNPs using bcftools. i merged multiple sorted bam files using a "rg.txt" file I created..

Where should I look first ?

Thanks for any help!

aarti