VCFtools doesn't keep any variants using GATK output

Question

[fcarolinebe]

Hi guys!

I'm having trouble using vcftools to filter snp through GATK output.

For information, I used the command HaplotypeCaller to make SNP-calling of 12 samples; I used the CombineGVCFs to join the 12 VCFS and make the joint call after merging the vcf files; I used the VariantFiltration for a initial filtration.

Now, I want to make a filtration using VCFtools, for this, I used the command :

vcftools --vcf output_gatk.vcf --remove-indels --recode --recode-INFO-all --out SNPs_gatk_only

But, after the filter, the VCFtools doesn't keep any variant, like this:

After filtering, kept 12 out of 12 Individuals
Outputting VCF file...
After filtering, kept 0 out of a possible 68599004 Sites
No data left for analysis!
Run Time = 438.00 seconds

Am I doing something wrong? My VCF input has a header like this:

#CHROM  POS     ID      REF     ALT     QUAL    FILTER  INFO    FORMAT  T1R2-6H T1R3-24H        T1R3-6H T1R6-24H        T1R7-24H ...       
LG1      1       .       G       <NON_REF>       .       PASS    END=14  GT:DP:GQ:MIN_DP:PL      ./.:0:0:0:0,0,0 ./.:0:0:0:0,0,0 ./.:0:0:0:0,0,0  ...
LG1     15      .       A       G,<NON_REF      .       PASS    END=31  GT:DP:GQ:MIN_DP:PL      ./.:0:0:0:0,0,0 ./.:0:0:0:0,0,0  ./.:1:3:1:0,3,19 ...

Answer 1

[Pierre Lindenbaum]

you're working with a GVCF file. A gvcf should be first post-processed with "gatk GenotypeGVCFs"

[fcarolinebe]

Thank you very much !! I didn't pay attention to this detail... Now it worked!!