Abstract It is well established that departure from healthy and tightly-controlled metabolism is a hallmark of cancer and hence, studying these metabolic changes can help us gain useful therapeutic insights. Several mathematical methodologies have been developed to quantify these changes to better understand metabolic adaptations in tumor cells. However, they (i) do not consider different genetic profiles and protein signatures leading to diversity among cancer patients and (ii) do not mimic biological interactions between metabolism and cancer phenotypes (such as chemo-resistance, infinite growth potential and invasive capacities). To successfully deconstruct metabolic and phenotypic links in tumorigenesis, we have developed a novel mathematical framework. First, to achieve patient-specificity we have devised a technique for reconstructing metabolic models that integrate high-throughput transcriptomic and proteomic data using a biased random-walk algorithm. The NetWalker software effectively implements this technique to score interactions between genes and proteins based on their expression levels and connectivity with each other and their metabolic targets. These scores are then used to identify contextually important sub-networks of the central carbon metabolism. Our reconstruction of the OVCAR-3 metabolic network independently identifies metabolic features in low-invasive ovarian cancers as highlighted in previously published literature and recent discoveries in our lab. These personalized models are then used to estimate phenotypically-consistent intracellular metabolic fluxes, with our novel metabolic flux analysis technique. Our approach is designed on the rationale that cancer cells regulate metabolism to maintain multiple phenotypic functions by optimally using the limited nutrients available in their environments. We quantitatively show via correlation studies and principal component analyses on NCI-60 and CCLE cell-lines’ gene-expression data that certain phenotypes are inherently competitive. This competition is manifested in metabolic functions, which is effectively modeled using our 13C-Multiobjective Metabolic Flux Analysis (13C-MOMFA) tool. It utilizes genome-scale metabolic models by integrating high-throughput omics data, along with measurements from metabolic assays and 13-Carbon (13C) stable isotope-labeled tracer studies, to quantify fluxes in cancer cells. Our approach is the first to connect cancer metabolism to their malignant phenotype - a crucial step to discover the metabolic underpinnings of cancer pathology. It also captures the heterogeneity in cancer cells and patients, hence providing a useful framework for targeted therapy. Citation Format: Abhinav Achreja, Lifeng Yang, Tyler Moss, Vasudha Sehgal, Juan Marini, Prahlad T. Ram, Deepak Nagrath. Design of phenotype-driven flux analysis approach for personalized metabolic models of cancer patients. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 1204. doi:10.1158/1538-7445.AM2015-1204