Running SCENIC by Cluster (Two Conditions)

[MujeebQadiri]

Hello!

First of all, thank you to the developers of this tool! I have a question regarding the workflow of my analysis -- I am looking at the differential regulon activity of samples A and B. I am mainly interested in 4 cell types which I have annotated in both samples, and I was wondering if I can run SCENIC by cell type (8 runs total) or should I run it by sample (2 runs total).

Thank you,
Mujeeb

[MujeebQadiri]

Hello everyone!

I proceeded with running SCENIC by combining the two conditions by cell type (WT cell type 1 vs KO cell type 1) for each cell type. I have a follow up question -- should I control for the cell counts in each condition when running SCENIC (run it with ~500 cells each). My worry is the regulon target gene definitions can be majorly influenced if one set of cells is much higher in count (500 cells vs 1500 cells). Does anyone have any experience with this?

Thank you!
Mujeeb