Comparing TFs across different cell types whithin two conditions (healthy/diseased)

[ashenflower]

Hi everyone,

I’m new to single-cell sequencing and transcription factor (TF) inference, so apologies in advance if this question seems a bit naive! I have an annotated single-cell dataset from healthy and diseased donors. I want to explore whether there are TFs specific to healthy or diseased patients and whether certain TFs are specific to particular cell types.

I’ve looked through similar questions and discussions (1,2,3) but haven’t really found a definitive answer in the comments, so I thought I’d ask here directly. Here’s where I’m unsure: should I run the pipeline on the entire dataset first and then analyze TF specificity among groups using a measure like RSS? Or would it be better to run the pipeline separately for subgroups in my data?

Also, if I were to run the pipeline on the whole dataset, would it make sense to then look at TF specificity across cell types within each condition (i.e., healthy and diseased patients analyzed separately)? Is that even a feasible approach?

Thanks so much for your time and any advice you can offer!