bio-gene-regulatory-networks-scenic-regulons

To infer transcription factor regulons from scRNA-seq data, apply the pySCENIC three-step pipeline: use GRNBoost2 for co-expression modules, cisTarget for motif-based pruning, and AUCell to score regulon activity per cell.

The pySCENIC pipeline for gene regulatory network analysis identifies master regulators by deriving co-expression modules with GRNBoost2, pruning them via cisTarget motif enrichment, and scoring cell-type-specific regulon activity using AUCell.

You need preprocessed scRNA-seq data in appropriate input formats such as loom or h5ad files to run GRNBoost2 and AUCell, along with a Python environment with pySCENIC and the required cisTarget motif databases installed.

You can use pySCENIC to identify master regulators for cell-type annotation by scoring TF activity across cells with AUCell, allowing you to cluster cells based on their underlying regulatory programs and distinct cellular identity.

pySCENIC requires motif databases to prune co-expression modules because cisTarget filters indirect regulatory interactions, ensuring the final gene regulatory network contains only transcription factor regulons backed by direct DNA motif binding evidence.

The best way to discover cell-type-specific regulons is running the complete pySCENIC workflow, which moves from broad GRNBoost2 co-expression networks to refined, high-confidence regulons scored per cell to reveal distinct regulatory states.