Development of an Interoperable Module and Novel Mathematical Framework for Modeling pH Dynamics Across Chromatography Modalities

A computational framework that integrates molecular-level characterization of the protein mixture with mechanistic process modeling to more comprehensively represent the coupled interactions among proteins, ligands, buffers, and salt electrolytes.

Industry Need

Chromatographic modeling for biopharmaceutical purification demands a detailed mechanistic understanding of pH dependent interactions among proteins, chromatographic ligands, buffering agents, and electrolytes. Foundational work by Pabst and Carta demonstrated that pH transients and gradients in weak ion exchangers can be effectively modeled by accounting for ligand titration. However, most existing models continue to emphasize the buffering capacity of the stationary phase while overlooking the substantial contributions of protein species. This simplification limits predictive accuracy under high-loading, process-relevant conditions where protein-associated buffering becomes significant. More recently, Hahn et al. extended Protein A chromatography models to include titration of the immobilized ligand, yet their approach excluded the buffering contribution of monoclonal antibodies and assumed a uniform pKa for all carboxyl groups on the Protein A ligand.

Approach

To address the limitations described in the above Need, the team proposes a new computational framework that integrates molecular-level characterization of the protein mixture with mechanistic process modeling to more comprehensively represent the coupled interactions among proteins, ligands, buffers, and salt electrolytes. This generalized framework is designed to be applicable across multiple chromatographic modes—including ion exchange (IEX), mixed-mode, hydroxyapatite, and Protein A systems—and applicable across a wide range of buffer and electrolyte conditions. The overall project will lead to the development and validation of an easy-to-use, well-documented Python-based module for pH-dependent column simulations and model calibrations.

Impacts

By enabling predictive, transferable models, the module will streamline process development and facilitate more efficient technology transfer from early-stage process development to manufacturing.

A new computational framework that integrates molecular-level characterization of protein mixtures with mechanistic process modeling to comprehensively capture the coupled interactions among proteins, ligands, buffers, and salt electrolytes.

Value Statement/Outcomes

This project will enable an interoperable modeling framework that improves pH control in chromatography for drug companies, resulting in quality benefits across downstream purification.

Outputs/Deliverables

A generalized, Python-based framework is designed to support pH-dependent column simulations, including accurate prediction of pH transients, and model calibration across diverse chromatographic modes such as ion exchange, mixed-mode, hydroxyapatite, and Protein A systems.