Proposed herein is a systematic media design framework that combines multivariate statistical approaches with in silico analysis of a genome-scale metabolic model of Chinese hamster ovary cell. The framework comprises sequential modules including cell culture and metabolite data collection, multivariate data analysis, in silico modeling and flux prediction, and knowledge-based identification of target media components. Two monoclonal antibody-producing cell lines under two different media conditions were used to demonstrate the applicability of the framework. First, the cell culture and metabolite profiles from all conditions were generated, and then statistically and mechanistically analyzed to explore combinatorial effects of cell line and media on intracellular metabolism. As a result, we found a metabolic bottleneck via a redox imbalance in the TCA cycle in the poorest growth condition, plausibly due to inefficient coenzyme q10-q10h2 recycling. Subsequent in silico simulation allowed us to suggest q10 supplementation to debottleneck the imbalance for the enhanced cellular energy state and TCA cycle activity. Finally, experimental validation was successfully conducted by adding q10 in the media, resulting in increased cell growth. Taken together, the proposed framework rationally identified target nutrients for cell line-specific media design and reformulation, which could greatly improve cell culture performance.