Invalid character '%' in 'FREQ' FORMAT field

[JoKw]

Hi,

When I used the ensemble method, I encountered the following error:

CalledProcessError: Command 'set -o pipefail; bcftools reheader -h /home/test/sample/work/varscan/samplecall-annotated-header.txt /home/test/sample/work/varscan/samplecall.vcf.gz | bcftools view | bgzip -c > /home/test/sample/work/bcbiotx/tmpB22Zux/samplecall-annotated.vcf.gz [W::vcf_parse] contig '1' is not defined in the header. (Quick workaround: index the file with tabix.) [E::vcf_parse_format] Invalid character '%' in 'FREQ' FORMAT field at 1:11181327 Write failed, wrote -1 instead of 14820 bytes. ' returned non-zero exit status 255

I had a look at the Varscan VCF file and it seems that the error was caused by the use of "%" in the FREQ field in the VCF file

Any fix for that? Thanks

[chapmanb]

Sorry about the issue, it looks like there is something problematic with the varscan output file that bcftools does not like. Would you be able to share the output of:

tabix /home/test/sample/work/varscan/samplecall.vcf.gz 1:11181327-11181327

and we can determine how best to work around it? Thanks much for the help debugging.

[JoKw]

1 11166713 . T C 0 PASS ADP=168;HET=1;HOM=0;NC=0;WT=0 GT:GQ:SDP:DP:RD:AD:FREQ:PVAL:RBQ:ABQ:RDF:RDR:ADF:ADR 0/1:154:168:168:122:46:27.38%:3.9712E-16:55:54:88:34:34:12

On Mon, 24 Apr 2017 at 22:04 Brad Chapman notifications@github.com wrote:

Sorry about the issue, it looks like there is something problematic with the varscan output file that bcftools does not like. Would you be able to share the output of:

tabix /home/test/sample/work/varscan/samplecall.vcf.gz 1:11181327-11181327

and we can determine how best to work around it? Thanks much for the help debugging.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/chapmanb/bcbio-nextgen/issues/1914#issuecomment-296678472, or mute the thread https://github.com/notifications/unsubscribe-auth/ATW-fSJpcknPpte7lRmpRBhIxd6iaqk_ks5rzKvQgaJpZM4NGGvy .

[chapmanb]

Thanks so much for following up with the problem samples. This helped me identify the underlying issue: we have code to fix FREQ but it was only being used on tumor/normal samples, not tumor-only. I pushed a fix to the latest development version (bcbio_nextgen.py upgrade -u development) which should resolve the problem. You can re-run your current project but should remove the varscan directory and checkpoints_parallel directories to let bcbio re-run the VarScan variant calls with the post-calling fix. Hope this gets everything working cleanly for you.