Pull 3' bias patch from kachulis/rnaseqc

[agraubert]

No description provided.

[kachulis]

Thanks @agraubert. The patch I implemented is a bit of a sledgehammer, there are definitely more precise ways to fix this that still allow the metrics to be calculated based on <10 genes. But it doesn't seem to me that the bias metrics would be particularly meaningful in those situations anyway, so not sure it's worth spending any significant time to do that.