Cancer cells have been observed to undergo a metabolic reprogramming in which glycolytic fluxes are upregulated whereas oxidative phosphorylation is downregulated. This metabolic phenotype is known as theWarburg effect and it is observed even under high oxygen conditions. To study the mechanism behind this metabolic reprogramming: - We formulate a pipeline for consistently reducing data-driven phenotypic models from a human large genome scale model (GEM). - We reduced a model (PetitHuman) from the GEM Recon 2 [1] that is representative of the central carbon metabolism of mammalian cells. - For this pipeline we combine fluxomics and metabolomics data to simulate a normal mammalian cells metabolic phenotype during exponential growth phase and the metabolic phenotype of a cancer cell. - We then analyze the feasible flux ranges of the two simulated metabolic phenotypes. - Same pipeline is used to reduce consistent data-driven phenotypic networks to study variability across different tumor tissues.