cordial

cordial infers functional biochemical pathways by identifying genes with correlated dependencies from large-scale RNA interference (RNAi) screens to perform functional pathway inference analysis (FPIA).

Key Features:

• Functional Pathway Inference Analysis (FPIA): Infers pathway membership by correlating anti-proliferative gene properties across large-scale RNAi screens.
• Correlation-based dependency analysis: Quantifies positive correlations among gene dependencies to identify genes likely in a common biochemical pathway.
• Detection of anti-correlations: Identifies anti-correlated dependencies to highlight known or candidate negative regulators such as PTEN.
• Over-representation analysis: Performs over-representation analysis on FPIA-identified genes to enrich and confirm pathway associations.
• Tumor lineage–specific analysis: Extracts cell type–specific networks when applied to specific tumor lineages.
• Benchmarking and validation: Validated using well-characterized oncogenic pathways including PI3K/AKT/MTOR, p53, and MAPK with expected gene associations.
• R package implementation: Implemented as an R package for computational analysis.
• Empirical gene associations: Recovers associations such as AKT1, MTOR, and PDPK1 correlated with PIK3CA, anti-correlation with PTEN, MDM2 and TP53BP1 linked to p53, and MAPK1 and BRAF correlated with MEK1 signaling.

Scientific Applications:

• Genome annotation: Linking genes to biochemical pathways based on dependency correlation patterns.
• Biochemical process elucidation: Delineating components and interactions within signaling pathways such as PI3K/AKT/MTOR, p53, and MAPK.
• Drug target discovery: Identifying signaling members and potential drug targets through correlated dependencies.
• Cancer biology: Characterizing tumor lineage–specific networks to inform cancer biology and therapeutic hypotheses.
• Pathway enrichment confirmation: Confirming biological relevance of inferred networks via over-representation analysis.

Methodology:

Performs correlation analysis of anti-proliferative gene dependency profiles derived from large-scale RNAi screens followed by over-representation analysis, with validation against benchmark oncogenic pathways (PI3K/AKT/MTOR, p53, MAPK).