Need non-uniquely-mappable locations

[sjneph]

We need to add in regions of the genome that aren't uniquely mappable. Eric's hotspot2 deals elegantly with missing data and these are non-zeroes.

[sjneph]

Looks like there is a way to pass in regions to exclude (unmappable locations, for example). I tend to believe that this particular input should be mandatory, but at least there is an option to do it (and we should use it internally and at the DCC). Note that we do not include the badspot heuristic from hotspot1. I'm unsure how prevalent or real the microsatellite effect is.