Add some way of finding transcripts of interest across dataset

[ptgolden]

One example, from 4 March lab meeting, is to figure out abundances that differ from one treatment to the rest. That is, given five treatments (A, B, C, D, E), could ask the question:

Given the transcript abundances at treatment A, in which treatments are those transcripts expressed at the same level in treatments B and C, but different levels at D and E?

This is actually kind of like the first filtering mechanism we had, where you would draw trend lines for what you were looking for.