The frontline support team will be offline as we are occupied with the GATK Workshop on March 21st and 22nd 2019. We will be back and available to answer questions on the forum on March 25th 2019.
possible issue with VCF files made by GenotypeGVCFs in 3.2-2??
I'm trying to follow best practices and also use the latest version of GATK/Queue. I've created 2 GVCF files, and I've run GenotypeGVCFs on them to create a VCF file. These are the commands:
java -jar -Xmx8g /home/sgfriede/gene_apps/gatk-3.2-2/GenomeAnalysisTK.jar -T GenotypeGVCFs \
-R /gpfs_share/sgfriede/ref_data/canFam3.fa \
--variant /gpfs_share/sgfriede/poodles_addisons/cocoa_wilt/gvcf/cocoa_wilt.sorted.dedup.haplotypecaller.g.vcf \
--variant /gpfs_share/sgfriede/poodles_addisons/ellie_traska/gvcf/ellie_traska.sorted.dedup.haplotypecaller.g.vcf \
From this, I get the output file no problem.
However, when I try to run VQSR on this VCF (affected_poodles.vcf) file I get the following error message:
ERROR MESSAGE: Line 5082: there aren't enough columns for line 0,3:3:9:71,9,0 0/0:2,0:2:6:0,6,61 (we expected 9 tokens, and saw 2 ), for input source: /gpfs_share/sgfriede/poodles_addisons/affected_poodles/20140722/vcf/affected_poodles.vcf
I think (but I could be wrong!!) I've narrowed this down to something with the output format from GenotypeGVCFs in GATK 3.2-2, because when I run VQSR on the same file but created using GenotypeGVCFs with GATK 3.1-1, VQSR runs just fine (and I'm calling the VQSR walker using 3.2-2).
It is also possible I'm doing something wrong here, but I thought I'd see if any of the experts might have an idea.
Appreciate your help as always!