Bioprocess biomarker identification and diagnosis for industrial mAb production based on metabolic profiling and multivariate data analysis.

Monoclonal antibody (mAb) production is a complex bioprocess influenced by various cellular and metabolic factors. Understanding these interactions is critical for optimizing manufacturing and improving yields. In this study, we proposed a diagnostic and identification strategy using quantitative proton nuclear magnetic resonance (H qNMR) technology-based pharmaceutical process-omics to analyze bioprocess variability and unveil significant metabolites affecting cell growth and yield during industrial mAb manufacturing. First, batch level model (BLM) and orthogonal partial least squares-discriminant analysis (OPLS-DA) identified glucose and lactate as primary contributors to culture run variability. Maintaining an optimal glucose set point was crucial for high-yield runs. Second, a partial least squares (PLS) regression model was established, which revealed viable cell density (VCD), along with glutamine, maltose, tyrosine, citrate, methionine, and lactate, as critical variables impacting mAb yield. Finally, hierarchical clustering analysis (HCA) highlighted one-carbon metabolism metabolites, such as choline, pyroglutamate, and formate, as closely associated with VCD. These findings provide a foundation for future bioprocess optimization through cell line engineering and media formulation adjustments, ultimately enhancing mAb production efficiency.