Hi,
I'm trying to generate a GVCF with HaplotypeCaller based on amplicon (PCR) sequencing, and bases towards the end of the reads (~10 bases last bases) consistently get genotype quality of 0, even when the base quality is high. I tried simulating reads, and even if the position is covered by reads from both strands, with a large number of reads, and high base quality the genotype quality of the reference is 0. If there is a SNP in that position, it does get called with reasonable genotype quality, only the reference bases get quality 0. I tried changing various parameters but nothing seems to make a difference? Is there a parameter I'm missing? should I submit a detailed bug report?