This study investigates the application of closed-loop evolutionary multiobjective optimization (EMO) to the discovery of antibody purification processes that represent the best trade-off in terms of manufacturing costs, robustness, and likelihood of meeting purity specifications. The EMO algorithm optimized these objectives as a function of critical decisions relating to the (chromatography) sequence, equipment sizing strategy and the operating conditions adopted for each chromatography step, subject to multiple demand and process-related (resin requirement) constraints. Batch-to-batch variations in product titres, dynamic binding capacities, product yields, and removal capabilities for HCPs and aggregates are accounted for by the optimizer using Monte Carlo simulations. The optimization procedure is applied to an industrially-relevant case study focused on a single-product antibody manufacturing facility subject to different upstream and downstream train ratios. To facilitate decision making, methods for visualizing the trade-offs exhibited by the optimal purification processes found by the EMO algorithm are provided.

Multiobjective evolutionary optimization for antibody puri?cation process design

245th National Meeting (2013)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.