Isovar undercounting supporting reference reads

[ihodes]

Looking at a variant with 81 RNA reads spanning, isovar is saying 70 are reference and 9 support the alternate allele. Looking in IGV I see 72 reads supporting reference; looking around, I notice that two (ref) reads start at the point of the variant: looks like Isovar misses those?

[ihodes]

@julia326 this might be a good one for us to work on this week

[ihodes]

Did we figure this out?

[iskandr]

Isovar throws away reads which don't align to reference positions before and after the variant. Planning on fixing this eventually as part of a rewrite of LocusRead.

[ihodes]

In the screenshotted case, they do appear to align to the reference, though?

[iskandr]

The read with the variant position on its end will get dropped.

[iskandr]

Related to #14