Tavi Nathanson

Certainly seems like that's the easiest thing for now: namely using `epidisco` for everything that `cohorts` doesn't do.

Recent thoughts from @hammer (correct me if I'm paraphrasing incorrectly): `epidisco` for anything often generated, `cohorts` for anything exploratory. Our current strategy doesn't quite fall into that description, since we're...

From @hammer: > Cohorts and Epidisco share some functionality. It would be nice to centralize the discussion of how we'll ensure they generate consistent results and ultimately separate concerns. >...

@arahuja yeah those were the main ones I was thinking.

Travis didn't fail because of `requirements.txt`: `mhctools>=0.3.0,

The latest Topiary will probably break Cohorts as well: https://github.com/hammerlab/cohorts/blob/master/requirements.txt#L4

Except `isovar` is still frozen? Worth re-opening for that?

At first glance I like that idea @jburos 👍

Also see https://github.com/hammerlab/bladder-analyses/issues/4#issuecomment-222227666 (sorry to link to a private repo).

The latter; I didn't think too hard about the exact grouping mechanisms for each count, and I know different grouping/de-dup'ing strategies have gotten us into trouble before.