Current State and Impact of Implementation of In-Silico Tools in the Biopharmaceutical Industry - Proceedings of the...

The sixth modeling workshop (6MW) was held in June 2025 at Collegeville, Pennsylvania, USA, and supported by the Recovery of Biological Products Conference Series. The goal of the workshop was to assemble modeling practitioners to review and discuss the current state of implementation, progress since the last modeling workshop (5MW) in 2023, gaps and opportunities for development, deployment, and maintenance of models in bioprocess applications. One key observation from the 6MW was the positive impact of modeling on the development of a wide range of therapeutic modalities (proteins, vaccines, peptides) and unit operations (upstream, purification, and formulation), and across time (from discovery via molecular design/developability to commercialization). The scope of 6MW was broader than previous workshops, thus including biophysics and molecular modeling, mechanistic modeling and computational fluid dynamics, plant and sustainability modeling, artificial intelligence (AI) and big data modeling, and Highland Games 2.0. The AI and big data modeling session captured some of the initiatives across the pharmaceutical industry that have flourished with the recent development of new digital and AI technologies. The Highland Games 2.0 session benchmarked specific applications and advances of modeling tools to support developability predictions from sequence, building on the original competition in 2018. One key challenge emphasized at prior modeling workshops was the need for large data sets to support collaborative modeling development within the recovery community. This still remains a gap for the community at large (e.g., a shared reference point for academic/industrial collaborators for benchmarking). Significant progress on developing and curating large data sets within institutions to support in silico molecular and process development was reviewed in two sessions: AI and Big Data and Highland Games 2.0. A potential solution would be the use of specific descriptors in a database that would allow for detailed analyses without sharing proprietary information. The 6MW included two invited papers on building molecular understanding into the manufacturing process and summarizing the ongoing technical challenges, successes and best practices of utilizing federated learning at the National Institute for Innovation in the Manufacturing Biopharmaceuticals (NIIMBL), respectively. Advancement was reported in all of the modeling areas/disciplines discussed in sessions at the 6MW, with the greatest advancements observed in mechanistic modeling applied to mixing and filtration, AI and big data modeling, and mechanistic modeling applied to chromatography. Highlights of the workshop included significant advancements in biophysical/molecular modeling for vaccines and multi-scale applications, and mechanistic models for filtration. Nevertheless, continued research and development for all modeling approaches is still needed and should be pursued.