Details about plot SCENIC results with cytoscape

[DiracZhu1998]

Dear authors, thank you for bringing us such a wonderful toolkit.
I have successfully implemented pyscenic, the retained TF-target and weight on regulon were retrieved and as input to cytoscape for plotting network. However, there are some details which haven't been described in many papers:
1). If a TF links too many genes, what thresholds for the linkage should be used to filter
2). The GRNs have a hierarchical structure but if we simply merge regulons, for example in my case, Olig2 and Sox8 are two important TFs, the first one is regulated by Sox8, thus, the Olig2 was merged into Sox8, and Olig2(+) TF cannot be found after merging regulons. In this case, what should I do? Do I need to skip the "merge step" if I want the hierarchical structure?
3). If I want to analyze all regulons together (if they show cooperative behavior on regulating a similar set of target genes), what percentage of overlaps between two regulons' target gene set should be retained?

Thank you for your help!
Best wishes,
Yuanzhen

[DiracZhu1998]

for question3, maybe I should also consider PPI?