Improve statistical test methods

[abearab]

Current method:

https://screenpro2.readthedocs.io/en/latest/PhenoScore.html

Statistical test comparing y vs x per each target, T:

$$\text{p-value}(T,x,y) = \text{t-test} \left( \begin{bmatrix}{N_{x}}\end{bmatrix}_{(a,b)}, \begin{bmatrix}{N_{y}}\end{bmatrix}_{(a,b)} \right)$$

(see this wikipedia page: Dependent t-test for paired samples)

(see the link to the implemented tool: ttest_rel, a scipy module)

---

@mhorlbeck used Mann–Whitney U test in https://github.com/mhorlbeck/ScreenProcessing. The t-test approach here was originally extracted from @hanig's notebook in which he processed a V3 screen for his lab.

@AshirBorah has some ideas for assigning empirical p-values relative to negative control guides/oligos, we can discuss it here and implement it moving forward.