A multi-dimensional modeling framework integrating metabolomics analysis and process modeling for bioprocess...

Area of Science:

• Biotechnology
• Bioprocess Engineering
• Metabolic Engineering

Background:

• Traditional bioprocess development often overlooks dynamic cell culture trajectories, focusing instead on endpoint results.
• Existing methods like design of experiment (DoE) and response surface methodology have limitations in capturing real-time process dynamics.

Purpose of the Study:

• To develop a multi-dimensional modeling framework integrating deterministic screening design (DSD), time-resolved metabolomics, and process modeling.
• To elucidate how critical process parameters (CPPs) regulate antibody titer and quality attributes during cell culture.
• To enable mechanistic understanding and real-time prediction of bioprocesses for improved optimization and scale-up.

Main Methods:

• Utilized DSD to identify key regulators of antibody titer and quality.
• Employed time-resolved metabolomics to reveal underlying metabolic mechanisms influenced by CPPs (temperature, inoculation density, pH).
• Developed Gaussian process regression (GPR) dynamic models integrating time-series metabolite data and CPPs for predictive modeling.

Main Results:

• DSD identified key CPPs affecting antibody production.
• Metabolomics revealed that temperature and inoculation density impact amino acid and energy metabolism, while pH affects central carbon flux.
• GPR dynamic models accurately predicted viable cell density and antibody titer (test-set R² > 0.90), outperforming linear (PLS) and machine learning (boosted trees, neural networks) models.

Conclusions:

• The integrated framework provides mechanistic insights into bioprocess regulation.
• The developed GPR models offer accurate real-time prediction of bioprocess performance.
• This approach serves as a valuable tool for robust bioprocess optimization and scale-up in upstream biomanufacturing.