Are There "Standard" Quality Score Cutoffs For Gatk Vcf'S

Question

[Max]

Are there "canonical" quality score cutoffs for GATK vcf outputs that are more or less standard in the literature? Many papers don't report their cutoffs more mutation calls, values that I've seen include >40 or 50 for read quality (GATK's default) and >20 for GQ (genotype quality scores). The latter seems far too conservative, at least in view of the number of mutations I'm finding per sample vs. those reported for similar tumor samples in the literature.

Answer 1

[Charles Warden]

GATK should add LowQuality flags in the .vcf.

At least for the examples I have seen, the high quality results (labeled with a ".", instead of "HighQuality") should be pretty safe calls and I would just use either the full .vcf or focus on the variants without the "LowQuality" flag.