(468b) Hybrid Model Development for Parameter Estimation and Process Optimization of Hydrophobic Interaction Chromatography

The biopharmaceutical industry has attracted considerable attention, as can be witnessed by the increasing market demands and approvals of biologics ^1-3. During the manufacturing of biological drugs, the formation of protein aggregates during production and separation processes remains a primary concern ⁴. To resolve this concern, hydrophobic interaction chromatography (HIC) is widely used in downstream polishing steps for the separation of targeted monomeric forms of protein therapeutics from the dimeric and/or multimeric species. Despite being commonly employed as an efficient purification strategy, the mechanism for HIC adsorption is quite complex, depending on various process parameters, like pH, salt concentration, and adsorbent ligand hydrophobicity. Inspired by the quality by design (QbD) initiative ⁵, mechanistic modeling has become an important tool for process characterization, but it is highly dependent on the understanding of the underlying phenomena. Due to limited understanding of the HIC adsorption mechanism, there is an absence of a reliable mechanistic model to capture the interaction between proteins and ligands under varying salt ions inside HIC ^{6, 7}. Therefore, it is essential to develop a model with high accuracy and suitable hypotheses to simulate the HIC process ⁸. Hybrid modeling strategy is a promising alternative to accurately describe the binding mechanism and reduce the model development effort as this approach can exploit the available information about the process and represent the missing knowledge by a data-driven component ^{9, 10}.

In this work, a mechanistic model is first calibrated using an equilibrium dispersive model for hydrodynamics and modified isotherm derived by Wang et al. ¹¹ for adsorption, followed by the model validation with experimental data. Due to the limited understanding of the underlying adsorption mechanism, a hybrid model is proposed by combining a simpler multi-component Langmuir isotherm (MCL) with a neural network (NN). Different methods to integrate the MCL with NN are investigated to find the appropriate hybrid model structure. During parameter estimation, a regularization strategy is incorporated to avoid overfitting. Additionally, the effect of different NN structures and regularization rates is comprehensively examined to obtain the hybrid model with the best performance. To ensure the generalizability of the developed hybrid model, an in-silico dataset is generated using the mechanistic model to test the extrapolation capability of the hybrid model. Finally, process optimization is conducted to find the optimal operating conditions under product quality constraints, and the optimal results obtained from the mechanistic and hybrid models are compared thoroughly.

References

1. Gargalo, C.; Heras, S.; Jones, M.; Udugama, I.; Mansouri, S.; Krühne, U.; Gernaey, K., Towards the development of digital twins for the bio-manufacturing industry. Digital Twins 2020, 1-34.
2. Ding, C.; Ardeshna, H.; Gillespie, C.; Ierapetritou, M., Process design of a fully integrated continuous biopharmaceutical process using economic and ecological impact assessment. Biotechnology and Bioengineering 2022, 119 (12), 3567-3583.
3. Ding, C.; Ierapetritou, M., A novel framework of surrogate-based feasibility analysis for establishing design space of twin-column continuous chromatography. International Journal of Pharmaceutics 2021, 609, 121161.
4. Wasalathanthri, D. P.; Rehmann, M. S.; Song, Y.; Gu, Y.; Mi, L.; Shao, C.; Chemmalil, L.; Lee, J.; Ghose, S.; Borys, M. C.; Ding, J.; Li, Z. J., Technology outlook for real-time quality attribute and process parameter monitoring in biopharmaceutical development-A review. Biotechnol. Bioeng. 2020, 117 (10), 3182-3198.
5. Shekhawat, L. K.; Chandak, M.; Rathore, A. S., Mechanistic modeling of hydrophobic interaction chromatography for monoclonal antibody purification: process optimization in the quality by design paradigm. Journal of Chemical Technology & Biotechnology 2017, 92 (10), 2527-2537.
6. Lietta, E.; Pieri, A.; Cardillo, A. G.; Vanni, M.; Pisano, R.; Barresi, A. A., An Experimental and Modeling Combined Approach in Preparative Hydrophobic Interaction Chromatography. Processes 2022, 10 (5).
7. Andris, S.; Hubbuch, J., Modeling of hydrophobic interaction chromatography for the separation of antibody-drug conjugates and its application towards quality by design. J Biotechnol 2020, 317, 48-58.
8. Narayanan, H.; Seidler, T.; Luna, M. F.; Sokolov, M.; Morbidelli, M.; Butté, A., Hybrid Models for the simulation and prediction of chromatographic processes for protein capture. Journal of Chromatography A 2021, 1650, 462248.
9. Chen, Y.; Ierapetritou, M., A framework of hybrid model development with identification of plant‐model mismatch. AIChE Journal 2020, 66 (10), e16996.
10. Wang, G.; Briskot, T.; Hahn, T.; Baumann, P.; Hubbuch, J., Root cause investigation of deviations in protein chromatography based on mechanistic models and artificial neural networks. J Chromatogr A 2017, 1515, 146-153.
11. Wang, G.; Hahn, T.; Hubbuch, J., Water on hydrophobic surfaces: Mechanistic modeling of hydrophobic interaction chromatography. J Chromatogr A 2016, 1465, 71-8.