Glycosylation, one of the most common post-translational modifications (PTMs) for proteins, occurs via a complex set of reactions by which sugars are added to and removed from proteins. These glycans can impact folding stability, intracellular and extracellular targeting, and mediation of binding to host cells. In therapies, the glycoform distribution can impact the immunogenicity, circulatory half-life, pharmacokinetics, and effector functions. To ensure consistent product quality and glycoform distribution, an understanding of factors that affect the glycoforms is required. These factors range from protein structure around the glycosylation site to various genetic mutations, to host enzyme expression levels to culture conditions. The models for this range from data-driven to mechanistic. Data-driven models can be constructed without needing mechanistic information, but their lack of a mechanistic basis makes development of a broadly applicable control strategy more challenging. Additionally, knowledge gained from data-driven models are not necessarily additive, which makes utilization of different observations in the literature difficult.
This project will develop software for glycosylation modeling of fed-batch Chinese Hamster Ovary (CHO) cell culture. The project will develop
(1) A next-generation hybrid model of N- and O-linked glycosylation,
(2) A well-defined workflow for constructing and testing a mechanistic-machine learning model with parameter estimation, uncertainty quantification, and machine learning for adapting the model across cell lines and culture conditions,
(3) User-friendly software implementing the models and the associated workflow, which is computationally efficient enough to be incorporated into optimization-based real-time control,
(4) An adaptive model predictive control formulation with associated scenario and robustness analysis and real-time software implementation, and
(5) User manuals, training materials, and training in the methodologies and software.
Reduce process development time and cost of Antibody products.
Savings of 1 month per product for major deviation investigation (predictions are faster than experiments). Reduced deviations and investigations with improved process understanding and control strategy. Accelerated filing by 1 month if on critical path.
Savings of 60 experiments (3 FTE mo) per product for process design during Phase I to licensure. Spanning process development, characterization, technology transfer and investigations
Reduce process development time and cost of Antibody products
Publication BD5 - GlycoPy In-review Quasi-Steady-State Approach for Efficient Multiscale Simulation of mAb Glycosylation in CHO Cell Culture Yingjie Ma, Jing Guo, and Richard Braatz AIChE Journal
Publication BD5 - GlycoPy Draft Adaptive Nonlinear Model Predictive Control of the Glycosylation Process of Monoclonal Antibody in CHO Cell Culture Yingjie Ma, Jing Guo, and Richard Braatz AIChE Journal
Publication BD5 - GlycoPy Planned "A General Chemical and Bioprocess Modeling, Optimization and Optimal Control Package in Python" Yingjie Ma, Jing Guo, and Richard Braatz AIChE Journal
Publication BD5 - GlycoPy Planned Parameter Estimation and Model-based Analysis of the Monoclonal Antibody Glycosylation Jing Guo, Yingjie Ma and Richard Braatz Biotechnol. Bioeng
Publication BD5 - GlycoPy Planned Model-based Design of Experiments for the Glycosylation Process of Monoclonal Antibodies in CHO Cell Culture Yingjie Ma, Jing Guo, and Richard Braatz Biotechnol. Bioeng
Presentation BD5 - GlycoPy Presentation complete Glycopy: A Multiscale Model-Based Simulation, Optimization, and Optimal Control Python Package for Monoclonal Antibody Glycosylation in Cell Culture Yingjie Ma, Jing Guo, and Richard Braatz 2023 AIChE Annual Meeting
Presentation BD5 - GlycoPy Presentation complete Model-based Design of Experiments for Antibody Glycosylation in CHO Cell Culture Jing Guo, Yingjie Ma and Richard Braatz 2023 AIChE Annual Meeting
Presentation BD5 - GlycoPy Presentation complete Nonlinear Model Predictive Control of the Monoclonal Antibody Glycosylation in Cell Culture Yingjie Ma, Jing Guo, and Richard Braatz 2024 AIChE Annual Meeting
Presentation BD5 - GlycoPy Presentation complete Model-based Analysis of the Monoclonal Antibody Glycosylation Across CHO Cell Lines Jing Guo, Yingjie Ma and Richard Braatz 2024 AIChE Annual Meeting
Presentation BD5 - GlycoPy Presentation complete GlycoPy: A Python Package for Multiscale Model-Based Simulation, Optimization and Optimal Control of mAb Glycosylation in Cell Culture Jing Guo, Yingjie Ma and Richard Braatz 2024 NIIMBL National Meeting
Publication BD5 - O-linked Glycosylation In-review Enhanced O-glycosylation Site Prediction Using Explainable Machine Learning Technique with Spatial Local Environment Seokyoung Hong, Krishna Gopal Chattaraj, Jing Guo, Bernhardt L. Trout, Richard D. Braatz Bioinformatics
Publication BD5 - Machine Learning In-review Linear and Neural Network Models for Predicting N-glycosylation in Chinese Hamster Ovary Cells Based on B4GALT Levels Pedro Seber and Richard D. Braatz Computers and Chemical Engineering https://doi.org/10.1101/2023.04.13.536762
Ma, Y. Guo, J. Maloney, A. Braatz, R. (2024). Quasi-Steady-State Approach for Efficient Multiscale Simulation and Optimization of mAb Glycosylation in CHO Cell Culture, AIChE, 2409.00281. https://doi.org/10.48550/arXiv.2409.00281
Ma, Y., Guo, J., & Braatz, R. D. (2026). GlycoPy: An equation-oriented and object-oriented software for hierarchical modeling, optimization, and control in Python. arXiv. https://doi.org/10.48550/arXiv.2601.01413
Ma, Y., Guo, J., Dubs, A. B., Ganko, K., & Braatz, R. D. (2026). Adaptive nonlinear model predictive control of monoclonal antibody glycosylation in CHO cell culture.Control Engineering Practice, 169, 106731. https://doi.org/10.1016/j.conengprac.2025.106731
Seber, P. & Braatz, R. D. (2024). Recurrent neural network-based prediction of O-GlcNAcylation sites in mammalian proteins. Computers & Chemical Engineering. https://doi.org/10.1016/j.compchemeng.2024.108818
Seber, P., & Braatz, R. D. (2025). Linear and neural network models for predicting N-glycosylation in Chinese Hamster Ovary cells based on B4GALT levels. Computers & Chemical Engineering, 194, 108937. https://doi.org/10.1016/j.compchemeng.2024.108937
Seokyoung Hong, Krishna Gopal Chattaraj, Jing Guo, Bernhardt L Trout, Richard D Braatz, Enhanced O-glycosylation site prediction using explainable machine learning technique with spatial local environment, Bioinformatics, Volume 41, Issue 2, February 2025, btaf034, https://doi.org/10.1093/bioinformatics/btaf034
Guo, J., Presenter, GlycoPy: A Python Package for Multiscale Model-Based Simulation, Optimization and Optimal Control of mAb Glycosylation in Cell Culture, NIIMBL National Meeting, Washington DC, June 26, 2024
Guo, J., Presenter, Model-based Analysis of the Monoclonal Antibody Glycosylation Across CHO Cell Lines, AIChE Annual Meeting, San Diego CA, October 30, 2024.
Guo, J., Presenter, Model-Based Design of Experiments for Antibody Glycosylation in CHO Cell Culture, AIChE Annual Meeting, Orlando FL, November 6, 2023.
Ma, Y., Presenter, Glycopy: A Multiscale Model-Based Simulation, Optimization, and Optimal Control Python Package for Monoclonal Antibody Glycosylation in Cell Culture, AIChE Annual Meeting, Orlando FL, November 6, 2023. https://aiche.confex.com/aiche/2023/prelim.cgi/Paper/665617
Ma, Y., Presenter, Nonlinear Model Predictive Control of the Monoclonal Antibody Glycosylation in Cell Culture, AIChE Annual Meeting, San Diego CA, October 30, 2024.
Login to the NIIMBL member portal to access additional project information, including presentations, progress updates, reports, and more.
Not yet a member? Learn more about which level of NIIMBL membership is right for you and your organization.
Massachusetts Institute of Technology
AstraZeneca
Boehringer Ingelheim
Bristol-Myers Squibb
Cytiva
EMD Millipore Corporation
Federal Stakeholder: National Institute of Standards and Technology
Genentech, Inc.
Merck Sharp & Dohme LLC
NIIMBL
Pfizer, Inc.
Sanofi
University of Delaware
