Skip to content

pySCENIC GRN Specifications

Process Details

  • Name: pySCENIC GRN
  • Process UUID: 4e16292e410b4ebdbf07aaca9ce0757b
  • Process Group: SingleCell

Overview

This process implements pySCENIC (Python Single-Cell rEgulatory Network Inference and Clustering) for gene regulatory network (GRN) inference from single-cell RNA sequencing data. The workflow identifies transcription factor (TF) targets, validates these regulatory relationships using cis-regulatory motif analysis, and quantifies regulon activity across individual cells through AUCell scoring.

This process is implemented in Bash, which invokes Python scripts for integrating pySCENIC outputs and exporting results.

Key Functionality

  • Gene Regulatory Network Inference: Identifies co-expressed gene modules and potential TF-target relationships using GENIE3 algorithm
  • Motif Enrichment Analysis: Validates TF-target relationships by checking for enrichment of TF binding motifs in target gene promoter regions
  • Regulon Activity Scoring: Quantifies the activity of each regulon (TF and its validated targets) in individual cells using AUCell methodology

Input/Output Specification

Inputs

Required Inputs

  • Output File

    • Description: Output file specification for the analysis results
    • Format: out

  • Input Directory

    • Description: Directory containing reference databases including feather files for motif analysis
    • Format: directory

  • TF Lists File

    • Description: Text file containing transcription factor gene names for network inference
    • Format: txt

  • Expression Matrix

    • Description: Single-cell expression data in loom format containing gene expression counts
    • Format: txt

Outputs

  • Results Archive

    • Description: Compressed archive containing adjacencies matrix, regulons list, and AUCell activity matrix
    • Format: zip

  • Integrated Results

    • Description: Loom file with integrated pySCENIC results including regulon activities and metadata
    • Format: loom

Parameters & Settings

These parameters can be adjusted in the Foundry UI when running this process.

  • Mask dropouts?

    • Description: Using this option excludes zero-expression cells when calculating TF-target correlations, focusing only on cells where both genes are expressed. This affects how the correlation between TF and target genes is calculated
    • Available options: false (default), true

  • AUC threshold

    • Description: Fraction of the ranked gene list in each cell that is used when computing the AUC (Area Under the Curve) score for each regulon
    • Default value: 0.05

References & Resources

  • Tool Documentation: Contact the team for details on integrate_pyscenic_output.py
  • Related Papers: Aibar, S. et al. SCENIC: single-cell regulatory network inference and clustering. Nature Methods 14, 1083–1086 (2017). https://doi.org/10.1038/nmeth.4463