saeedfc
commented
4 years ago Hi, I am not a contributor to this package. But If I may suggest, I will start with feeding the Normalized matrix of the whole data(integrated data) (RNA assay, slot = "data") to SCENIC.
Onc you generate the results, get the AUC matrix. In this matrix you have the 'activity scores' for each transcription factor for each cell (Transcription factors as columns and cells as rows). You can add this to the meta.data of your seurat object. Then you can visualize using ViolinPlots with split.by argument for each transcription factor to see the difference between Treated and Control. Once you add the AUC output to your seurat meta.data, there are a lot of possibilities.
If I understand it correctly, to make a comparison possible, you should feed the whole matrix together and not separate and run SCENIC once.
You may also look at the regulon specificity score functionality available with PySCENIC. After adding the AUC output in the meta.data, you may split the AUC matrix (treated vs control), compute regulon specificity score for each AUC matrix separately and try to see the differences.
to compute regulon specificity score you need AUC matrix (treated and control) the cluster information of your seurat clusters. So you need meta data also split between treated and control. Since the AUC essentially xome from the single analysis, this can be a good comparison
Hope it helps
singcell
PDhingra2016
I have control and treated samples and I have generated Differential gene expression list from Seurat integration tool. I wanted to determine the gene regulatory network associated with the genes that are differentially expressed between conditions. How can I use SCENIC tool when I have two different dataset comparing two different conditions. Do I need to add samples separately ? Do you have workflow for this kind of analysis.