DiscoverSL

DiscoverSL predicts synthetic lethality interactions in cancer as an R package by integrating mutation, copy number alteration, and gene expression data from The Cancer Genome Atlas and constructing a multi-parametric Random Forest classifier.

Key Features:

• Multi-Omic Data Integration: Integrates mutation, copy number alteration, and gene expression data from The Cancer Genome Atlas (TCGA).
• Random Forest Classifier: Constructs a multi-parametric Random Forest classifier to predict synthetic lethal gene pairs.
• In Silico Validation: Validates predicted synthetic lethal genes using shRNA and drug screening datasets from cancer cell line databases.
• Clinical Outcome Analysis: Performs Kaplan-Meier survival analysis to evaluate cases where a mutation in one gene coincides with over- or under-expression of its synthetic lethal partner.

Scientific Applications:

• Therapeutic Target Identification: Identifies candidate synthetic lethal gene pairs as potential targets for targeted cancer therapy development.
• Prioritization for Experimental Follow-up: Prioritizes candidate SL pairs for experimental follow-up using integrated multi-omic evidence and screening data.
• Clinical Relevance Assessment: Assesses clinical associations between mutations, partner gene expression, and patient survival.

Methodology:

Aggregation of mutation, copy number alteration, and gene expression data from The Cancer Genome Atlas; construction of a multi-parametric Random Forest classifier; in silico validation using shRNA and drug screening datasets from cancer cell line databases; and Kaplan-Meier survival analysis for clinical correlation.