Would taking the average scores per cluster in the AUC matrix indicate more activity?

[CodeInTheSkies]

Hi there,

I have been able to run SCENIC successfully and have extracted the AUC matrix, which I plan to add to meta.data in Seurat and maybe make FeaturePlots.

I just want to verify if my thinking is right. The AUC scores range from 0 to 1, is that correct?

Also, if I manually compute the average AUC score for a given regulon across a Seurat-defined cluster, then that would indicate the average regulon activity for the whole cluster, is that right? In effect, by using this average, I can then choose the top regulons to plot as feature plots. If I pick the regulons that have higher averages, then I should, most likely, see more cells with more activity in that cluster. Is this correct?

Thank you!