search

aertslab / pySCENIC

pySCENIC is a lightning-fast python implementation of the SCENIC pipeline (Single-Cell rEgulatory Network Inference and Clustering) which enables biologists to infer transcription factors, gene regulatory networks and cell types from single-cell RNA-seq data.
http://scenic.aertslab.org
GNU General Public License v3.0
420 stars 179 forks source link

[results]TFs with zero expression level in some cells have not zero RAS in same cells #318

Closed xiaobaicainiao666 closed 3 years ago

xiaobaicainiao666 commented 3 years ago

When comparing the gene expression level and RAS of TF, there is a problem: Some TFs have zero expression level in some cells, but have not zero RAS(regulon activity score) in same cells. Is that normal? I think that zero-expression TF's corresponding regulon activity is meaningless in these cells. Is it because the calculation of RAS(regulon activity score)? Or because I didn't run pyscenic in the right way? Thank you very much!

xiaobaicainiao666 commented 3 years ago

One more question: This is the output of pyscenic ctx. What do these number mean? I guess it is -log10(Pvalue) showing relationship between TF and targets. Bigger number means stronger evidence of TF-target? image

cflerin commented 3 years ago

If a TF has zero expression in all cells, it will not be used for GRN inference, otherwise you will get results using the cells that do have expression values.

The regulon activity is not directly related to this though. After creating the regulons, each cell is scored based on the activity of the genes in that regulon (the AUCell step). So it is not directly based on TF expression alone, but the entire regulon.

The numbers after the target genes is the importance score from the GRN step.