Haplotype Size and HMM

[mtkelleher]

Hello,

I was wondering if the current HMM model takes haplotype size into account when finding a path through the graph. From what I can tell, it doesn't, but I could be wrong! If it doesn't, has that approach been tested before? I’d be curious to see how it performs, especially since I’ve noticed some switching between what we believe to be correct-incorrect-correct calls, where the middle potentially incorrect call corresponds to a much larger haplotype.

Thanks,
Micah