Multi-omics-driven kinetic modeling reveals metabolic vulnerabilities and differential drug-response dynamics in ovarian cancer

Cancer cells reprogram metabolism to fuel aggressive growth and resist therapies, yet mapping these metabolic changes remains a major challenge. While constraint-based genome-scale models describe steady states, kinetic models that resolve time-dependent fluxes and metabolite concentrations in human cancers are rare. Here, we integrate multi-omics data and physicochemical constraints to construct large-scale kinetic models of ovarian cancer metabolism that capture how tumor cells adapt their nutrient use and energy production. Comparing ovarian cancer cells with and without BRCA1 loss revealed distinct metabolic strategies driven by transcriptional regulation. The models reproduce hallmark phenotypes, correctly predict known metabolic drug targets, and identify previously uncharacterized vulnerabilities in nucleotide and lipid synthesis. Simulations of drug treatments mirror clinical responses and consistently reveal a ceramide-linked stress signature common to many chemotherapies. Using the same framework, we show that BRCA1 loss redirects metabolic pathway activity through enzyme activity changes linked to transcription factors interacting with BRCA1 , suggesting regulatory routes for network-level rewiring. By capturing dynamic fluxes and concentrations, these models bridge molecular insight with therapeutic action, guiding biomarker discovery and dosing strategies. This open-access resource provides a mechanistic foundation for testing metabolic interventions, deepening our understanding of cancer metabolism across tumor types, and advancing the promise of precision oncology.