Integrate Bioprocess Mechanisms into Modeling, Analytics, and Control Strategies to Advance Biopharmaceutical Manufacturing and Delivery Processes

Abstract

The existing stochastic simulation methodologies tend to ignore ordinary differential equations or (ODE) or partial differential equations or (PDE) mechanistic models that typically represent the scientific understanding of underlying process dynamics and mechanisms. For emerging manufacturing and delivery processes of biopharmaceuticals (such as cell, gene, RNA, protein, peptide therapies and vaccines), this can limit sample efficiency, reliability, and interpretability of critical decision making. It also affects mechanism learning. Therefore, in this tutorial paper, we present the recent studies that integrate bioprocess mechanisms into simulation modeling, risk/sensitivity/predictive analytics, process design and control strategies. They can overcome the key challenges, including high complexity, high uncertainty, and very limited data, and facilitate design and development of productive, robust, and automated end-to-end biopharmaceutical manufacturing and delivery processes.