Restarting edgePy

[bc2zb]

There was some discussion on the r-bioinformatics slack about restarting this, so let's get this moving. Reviewing some of the previous work, two issues that need to be worked out are covered in the papers below as highlighted in the table here. As far as I can tell, these methods are not covered anywhere in python, but I need to check PyDESeq2 to see how they handled it. I'm not as intimately familiar with DESeq2 as I am edgeR and limma, so I don't know if DESeq2 does these dispersions differently.

https://academic.oup.com/biostatistics/article/9/2/321/353777

https://academic.oup.com/bioinformatics/article/23/21/2881/372869