The problem is that Java treats the genotype fields as integers. Therefore, it thinks you are dividing two integers, and performing a/b where b>a will always result in a value of 0. You need to multiply the numerator and denominator by 1.0 to con…

Two questions:
1. How could a single read have nearly identical likelihoods for two different alleles? After all, a single read at a given variant position will have a base pair (or multiple base pairs, in the case of an insertion) with a particu…