I have some known variants called by a different caller, further confirmed by Sanger sequencing. I've noticed HaplotypeCaller (HC) will call variants at this location when they are in a heterozygous state, but often will not call them when they are homozygous. I've attached a screenshot. The gene is PKLR, the sequencing platform was Haloplex, which can introduce tiny random holes in the data. The two variants are frequent polymorphisms and fall on either side of this Haloplex hole. One other thing about this particular run is that we ran HC with only 3-4 samples. After we noticed this problem we tried to increase the batch size to 8 but it was still not called. Any ideas? Let me know if I can provide any further information!
↧