.png){kind=icon}
.png){kind=icon}
Return
Optimization strategies for monoclonal antibody production: advances in simulation and artificial intelligence in bioprocessing
.png){kind=icon}
OA
AI
K
M
DOI：10.1080/19420862.2026.2706886.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The rapid growth of monoclonal antibody (mAb) therapies has increased the need for efficient, scalable, and affordable manufacturing processes. However, mAb production remains complex because of nonlinear upstream cell culture behavior, expensive downstream purification, especially Protein-A chromatography, and plant-level bottlenecks that can increase cost, cycle time, and manuacturing uncertainty. This review examines recent developments in mAb manufacturing with focus on process simulation, mathematical optimization, and artificial intelligence/machine learning (AI/ML) across upstream processing (USP), downstream processing (DSP), and integrated plant-level operation. In USP, media optimization, dynamic feeding, high-density cultures, and continuous perfusion bioreactors are discussed in relation to productivity and critical quality attributes (CQAs). In DSP, alternative and intensified purification strategies are reviewed with a focus on recovery, impurity clearance, scalability, cost, and technology maturity. AI/ML applications are also discussed from early-stage development and cell-line screening to upstream control, CQA prediction, chromatography optimization, and downstream decision support. Despite these advancements, challenges such as data heterogeneity, limited standardized datasets, model transferability, and regulatory constraints remain important barriers to implementation. Overall, this review uniquely connects simulation and AI/ML approaches to practical optimization across the full mAb manufacturing workflow, including design, scheduling, debottlenecking, purification, monitoring, and quality prediction. The combination of process simulation, continuous bioprocessing, and AI/ML-based decision support may enable more flexible, reliable, and cost-effective mAb manufacturing. However, these benefits depend on validation through robust models, process-specific case studies, and technoeconomic analysis.
Keywords:
Artificial intelligence
downstream processing
machine learning
monoclonal antibodies
process simulation
Protein-A alternatives
upstream processing
