how necessary is to do VariantFiltration after VQSR?
I have ~200 whole genomes sequenced at 30x and am currently having an issue after HaplotypeCaller, which in chromosome Y most of the GT of the samples are not supported by many reads (around 4 or so). The problem comes when I apply VariantFiltration DP < 8 reads which sets to no call most of the GT in chrY, and post QC removes variants with > 5% missing call rate which removes all my variants in chrY!!
Hence I was wondering how important is to apply VariantFiltration providing that I already performed VQSR and I have a huge dataset. From the slides and doc I understood hard filtering should only be applied when VQSR can't but I guess a variant supported by just 4 reads cannopt be trusted?