CRSO

CRSO infers combinations of cooperative driver genetic alterations in individual tumors using a rule-set optimization framework implemented as an R package.

Key Features:

• Inference of Cooperative Alterations: Infers combinations of genetic alterations that cooperate to drive tumor formation in individual patients.
• Rule-set optimization (not co-occurrence): Uses a rule-set optimization approach that departs from traditional statistical co-occurrence methods to identify cooperating drivers.
• Application across TCGA cancer types: Applied to 19 TCGA cancer types, identifying an average of 11 core driver combinations per cancer type that typically involve 2–6 alterations and account for approximately 70% of samples.
• Identification of known and novel combinations: Recovers well-documented driver combinations missed by other methods and nominates novel combinations correlated with clinical outcomes across multiple cancer types.
• Discovery of therapeutically relevant synergies: Reveals synergistic interactions such as those in NRAS-mutant melanoma and core combinations involving NFE2L2 across four cancer types implicating the NRF2 pathway.
• Implementation: Provided as an R package for computational analysis of cooperative driver rule sets.

Scientific Applications:

• Personalized oncology: Enables identification of patient-specific cooperative driver combinations to inform targeted treatment hypotheses.
• Biomarker and target discovery: Supports discovery of novel driver combinations and pathway-level synergies (e.g., NRF2) that may serve as therapeutic targets or biomarkers.
• Comparative cancer genomics: Facilitates cross-cancer analysis of core driver rule sets across multiple TCGA cancer types.

Methodology:

CRSO applies a rule-set optimization algorithm implemented in R and explicitly departs from statistical co-occurrence methods to identify combinations of cooperating driver alterations.