RePhine

RePhine-drug: Identification of Drug Response-Associated Transcriptional Regulators

RePhine-drug integrates pharmacogenomic and ChIP-seq data using a regression-based framework to identify transcriptional regulators (TRs) associated with drug response in cancer.

Key Features:

• Multi-Omics Data Integration: Combines gene expression levels, copy number variations, mutation statuses, cancer type information, and pharmacological profiles to model associations between gene expression and drug response.
• ChIP-seq Integration: Incorporates ChIP-seq datasets to infer TR target genes and characterize regulatory mechanisms underlying transcriptional responses to drugs.
• Regression-Based Modeling: Applies regression models to integrate pharmacogenomic and ChIP-seq data while accounting for low mRNA abundance of TRs, protein modifications, and confounding factors.
• Comparative Performance Assessment: Evaluated using simulated datasets with noise or confounders and real pharmacogenomic data; compared against Pearson correlation analysis, logistic regression models, and gene set enrichment analysis.
• Drug Mechanism Clustering: Derives TR signatures from pan-cancer datasets to cluster drugs by mechanisms of action.
• Validated Predictive Findings: Predicted that loss-of-function of EZH2/PRC2 reduces cancer cell sensitivity to the BRAF inhibitor PLX4720; experimentally validated that pharmacological inhibition of EZH2 increases resistance to PLX4720.

Scientific Applications:

• Drug Response Mechanism Analysis: Identifies TRs involved in drug response pathways to elucidate molecular mechanisms of drug efficacy and resistance in cancer and support therapeutic target discovery.

Methodology:

Integrates pharmacogenomic variables and ChIP-seq-derived TR–target relationships within a regression-based statistical framework to model correlations between TR activity and drug response phenotypes across cancer datasets.