6th Modeling Workshop (6MW)

Call for Abstracts

Conference Logistic & Submission Process

Conference location: Collegeville, PA

Conference date: June 16-18th 2025 (2.5 days)

Abstract due date: Friday, January 24th, 2025 (due date will not be extended)

Communication of submission acceptance: Friday, February 14th, 2025

How to submit an abstract: Use the following link https://forms.office.com/r/5iPS2k638L. You will be required to enter 1. your title, 2. the authors with their affiliation and the presenting author underlined, 3. the preferred format (oral presentation or poster), 4. the session(s) where you want to present, 5. your abstract. If more than one session applies to your work, list the sessions by order of preference. More information will be e-mailed to you about the conference upon acceptance (preferred accommodation, best travel routes, site access...). If you encounter some difficulties uploading your abstract contact andre.c.dumetz@gsk.com with the following title “6MW2025 Abstract submission”.

6MW2025 Organization Committee: Arne Staby (Novo Nordisk), David Roush (Roush Biopharma Panacea), Felix Wittkopp (Roche), John Welsh (Rivanna Bioprocess Solutions), Jessica Lyall (Genentech), Deenesh Kavi Babi (NNE), Stephen Hunt (Merck), Robert Todd, Julie Robinson (Merck), Xuankuo Xu (BMS), James Reilly (Regeneron), André Dumetz (GSK), Sayantan Chattoraj (GSK), Sameer Talwar (GSK), Siddharth Parimal (GSK), Chris Gerberich (GSK)

Description of the Sessions

Session 1: Biophysics and Molecular Modeling

Biophysical modeling of biologics at the molecular level has become increasingly common with the advancement of software packages that streamline the modeling workflow. In silico screening of protein properties can be leveraged in a high impact manner to identify liabilities and rank order candidates for development. This molecular level understanding also guides process development, where biophysical and molecular modeling can identify ligands and mobile phase conditions for chromatographic purification as well as formulation conditions despite the semi-quantitative nature of the predictions.

In this session we invite abstracts focused on biophysics and molecular modeling tools and applications across the development pipeline. Submissions featuring hybrid or multi-scale models to support candidate selection, process development, and formulation are welcome. Submissions that focus on model development and advances in current modeling technologies (e.g. parameterization limitations) are also requested. We particularly invite submissions that focus on the key questions to come out of the 5MW workshop1, such as prediction of impurity

clearance/impurity interactions, applications in the cell/gene therapy space, addressing limitations of “big data set” availability, and implementation for late stage development, characterization, or regulatory activities.

Keywords: biophysical properties, molecular modeling, developability, downstream process design, formulation

Session 2: Mechanistic modeling & Computational Flow Dynamics (CFD)

Mechanistic modeling refers to the process of explaining a system's behavior in terms of the underlying physical, chemical, and/or biological process associated with the system (e.g. first principles). These models are often used to capture the cause-and-effect relationships within a system of interest. This session will address industrial challenges and use cases of applied mechanistic (including CFD) models in the biopharmaceutical industry and the regulatory challenges associated with the use of these models. How can mechanistic models improve process development - including, but not limited to, selection and sequencing of unit operations and evaluation of the operating space? How are mechanistic models best applied at the interface of development and manufacturing towards process design, especially for scaling up and process characterization or regulatory activities? We invite abstracts that highlight novel approaches or advancements in these areas for mechanistic models of any unit operation including, but not limited to upstream, harvest, chromatography, and filtration.

Keywords: model validation, unit operation design, parameter estimation, scale-down model qualification, design, and operating space

Session 3: Plant & Sustainability Modeling

Plant modeling (or flowsheet and system modeling) uses interconnected unit operation models to study, analyze and evaluate the full decision space (e.g. technology selection, technology impact etc.) of a given process. Leveraging knowledge gained and output from plant modeling, process sustainability can additionally be evaluated for selection of the best (optimal) flowsheet topology.

The top 4 challenges identified at the 2023 5MW1 are described as follows. First, data quality and availability which is based on historical and real‐time collected data maybe inconsistent (time-step wise) or incomplete (missing) thereby leading to unreliable models. Second, lack of variation in large‐scale manufacturing data leads to challenges within model parameter estimation and calibration used in creation of the flowsheet models. Third, complexity of manufacturing system which consists of numerous flowsheets and inputs making it a challenge to both identify model complexity and system representation. Finally, integration with other manufacturing systems for real-time simulation-optimization through connection of plant models (virtual) with manufacturing systems (physical) is complex.

Submissions addressing one or more of the identified challenges focusing on general flowsheet modeling topics and/or methods are welcomed. Examples of general flowsheet modeling topics are facility design/retrofit, technology selection, flowsheet optimization and logistics, sustainability etc. and examples of methods are standardized methodologies, workflows etc., developed and deployed within your organization as a standard workflow. Specifically, the submissions can describe case studies, sustainability, technology transfer or uncertainty quantification.

Case studies can highlight key learnings, key remaining open questions and describing scalable and non-scale approaches, sustainability can highlight how flowsheet models can be used for solvent selection, minimizing waste, process intensification opportunities, etc. Technology transfer can highlight how process/tech transfer from CMC-to-MSAT-to-Full scale has improved using models and uncertainty quantification can highlight how understanding uncertainty can lead to minimization of batch-to-batch variation within manufacturing.

Keywords: flowsheet simulation, process synthesis, technology selection, process capacity, plant-wide control

Session 4: Artificial Intelligence (AI) & Big Data Modeling

The field of artificial Intelligence (AI) and big data has been booming in recent years. Its effects are most obvious in industries such as retail, transportation, finance, and medical imaging. However, beyond the most discussed applications that have reached the mainstream media, modeling coupled with AI and big data has the potential to impact candidate selection, process development, and commercial manufacturing. The fundamental mission of CMC biotech scientists remains to develop robust, cost-efficient, and scalable processes to manufacture drug for clinical trials and supply patients post approval, and both AI and big data modeling has the potential to simplify and facilitate this mission.

This session aims to create a forum to discuss current applications of AI and big data modeling, present ongoing work with its associated challenges, and discuss the long-term vision of a development and manufacturing environment where AI and big data modeling is omnipresent.

Submissions presenting case studies, technical issues, and novel approaches where AI and big data modeling have been used and made a significant impact are encouraged. This session is intended to cover the full lifecycle of a biopharmaceutical from candidate selection to commercial manufacturing. Work that employs AI/big data techniques such as data mining, feature extraction and selection, machine learning, optimization, and generative AI is welcomed, along with demonstration of how these techniques helped to improve process development and control and/or reduce development times. Any regulatory aspect/experience and interactions with the regulatory agencies are particularly encouraged.

Keywords: artificial intelligence (AI), machine learning, generative AI, modeling, big data, candidate selection, process development, tech transfer, process validation, life cycle

Session 5: Highland Games 2.0

In 2018, the Recovery community participated in the Highland Games where teams competed to predict biophysical and drug properties of six monoclonal antibodies from amino acid sequences alone. Now, 7 years later, we invite the participating teams (and newcomers) to engage in a Highland Games reboot to demonstrate how in silico models have evolved by re-predicting the same challenge and presenting the results and model improvements at 6MW. Please contact the organizing committee if you are interested in participating.

References

1Wittkopp, F., Welsh, J., Todd, R., Staby, A., Roush, D., Lyall, J., Karkov, S., Hunt, S., Griesbach, J., Bertran, M.-O., & Babi, D. (2024). Current state of implementation of in silico tools in the biopharmaceutical industry—Proceedings of the 5th modeling workshop. Biotechnology and Bioengineering, 121, 2952–2973. https://doi.org/10.1002/bit.28768