Dual-Parameter Profiling of Synergistic and Antagonistic Epigenetic-Targeted Drug Combinations in Cancer: An In Silico Framework for Therapeutic Index Optimization

Background Cancer combination therapies often face challenges due to acquired resistance and unpredictable drug-drug interactions. Epigenetic drugs have shown promise in resensitizing tumors to targeted therapies by reversing epigenetic dysregulation. However, systematic methods to identify synergistic or antagonistic combinations remain limited. Objective To develop a dual-parameter in silico framework for identifying synergistic and antagonistic combinations of epigenetic and targeted therapy drugs in cancer treatment, using SynergyFinder 2.0 and comprehensive drug interaction data. Methods Epigenetic drugs were compiled from the Human Epigenetic Drug Database, and targeted agents were sourced from My Cancer Genome. Drug interaction data were extracted from DrugCombDB and analyzed using SynergyFinder 2.0. Synergy was assessed using four models: Highest Single Agent (HSA), Bliss Independence, Loewe Additivity, and Zero Interaction Potency (ZIP). A ComBination Sensitivity Score (CBSS) was introduced to evaluate therapeutic efficacy. Combinations were classified as synergistic (synergy score > 5 and CBSS > 50%) or antagonistic (synergy score < − 5). Results Among 206 evaluated combinations, 36 were identified as synergistic across all four models, with 23 showing high sensitivity (CBSS > 50%). Notable synergistic pairs included Hydralazine + Gefitinib and Vorinostat + Olaparib. In contrast, 171 combinations were classified as antagonistic, including TRANYLCYPROMINE + IMATINIB, which exhibited mechanistic conflict via Wnt/β-catenin pathway activation. Conclusion This study presents a robust computational framework for dual-parameter profiling of drug combinations, enabling the identification of synergistic pairs with high therapeutic potential and the exclusion of antagonistic combinations.