KnownSites for non-human data - generate them with UG or HC?
I am working on non-human data and want to use GATK's Haplotype caller to call variants. I am using the protocol that your team recommends.
So far I have reached the indel realignment stage and want to do base recalibration next. Since I do not have known sites, elsewhere in the forum you advise doing initial round of SNP calling on original un-recalibrated data.
Should I do Is this initial calling with UnifiedGenotyper on the indel realigned BAMs, or since I want to use Haploytype Caller downstream, should I use this walker instead?
Thanks for your input!