Which variant filering process is better for high-coverage data?
I'm currently analysing 4 high-coverage (>30X) mammalian non-human genomes.
I've implemented GATK's/Picard's dedup, realigner, fixmates, and called the variants using UnifiedGenotyper, followed by BaseRecalibrator.
Given there is no available database with variants called on other relevant samples, I'm wondering which variant filtering method is better to use:
1) Variant quality score recalibration
Thank you very much!