Strange genotype calling with HaplotypeCaller
I cannot understand some of HC's calls. It is rare, but occasionally HC will call individuals as 0/1 when their ADs are 0,50+ (clearly 1/1). I tried looking at the output bams at those positions, but nothing there would suggest to my human eye that the individual would have a haplotype that included a ref at that position. This is definitely an issue of long range haplotype calling, because the genotypes are "correctly" called when I use a vcf as the input file instead of a bed file containing multiple variant positions in the same region.