GATK version used: gatk/4.1.2.0 Exact command used: gatk BaseRecalibrator - ApplyBQSR - HaplotypeCaller - VariantRecalibrator-ApplyVQSR The following are the parameters I used in BQSR, VariantRecalibrator, and ApplyVQSR:
BQSR: --static-quantized-quals 10 --static-quantized-quals 20 --static-quantized-quals 30
VariantRecalibrator: --max-gaussians 6 -an QD -an FS -an MQRankSum -an ReadPosRankSum -an SOR -an MQ and four training sets from Bundle:
-resource:hapmap,known=false,training=true,truth=true,prior=15.0 $vcfHapmap \
-resource:omni,known=false,training=true,truth=true,prior=12.0 $vcfOmni \
-resource:1000G,known=false,training=true,truth=false,prior=10.0 $vcfGlk \
-resource:dbsnp,known=true,training=false,truth=false,prior=7.0 $vcfDbsnp
ApplyVQSR: -truth-sensitivity-filter-level 99.5
I am using GATK to call SNPs in tumor samples (I try to find all SNPs regardless of germline or somatic). And after I calibrated all SNPs identified by GATK, can I say other positions that don't include in output vcf have no SNPs? I used samtools to remove low quality and flagged reads and to see the depth of some interesting genomic positions that no SNPs detected by GATK. I regard those positions as genotype REF/REF if those sites have good depth. Wonder if it is feasible to do that.
Thanks.