Inferring the Metabolic Objectives of Mammalian Cells via Inverse Modeling of Fluxomics and Metabolomics

Metabolism reflects evolutionary priorities that govern how cells allocate resources. In mammalian cells, metabolic objectives are layered and context-dependent, making it difficult to pinpoint the priorities that underlie observed phenotypes. Here, we introduce ObjFind-M, an inverse optimization framework that infers reaction-level metabolic objectives in mammalian cells directly from fluxomic and metabolomic data. Using Chinese hamster ovary (CHO) cells as a data-rich mammalian cell system, ObjFind-M consistently identifies mitochondrial ATP synthase as the central metabolic driver, supported by key TCA cycle and electron transport chain nodes. Priorities adjust with cellular state, favoring glycolysis-TCA coupling in the growth phase and shifting toward oxidative phosphorylation and redox balance when proliferative activity slows. High recombinant protein producing CHO cells emphasize citrate shuttling and beta-oxidation, linking energy supply with biosynthetic capacity for protein secretion. Benchmarking against conventional objectives demonstrates that maximizing ATP production most accurately reproduces experimental fluxes. By quantifying metabolic objectives directly from data rather than assuming them a priori, ObjFind-M provides a framework for identifying reaction-level strategies that shape cellular decision making.