Poster: A novel in silico platform combining data-driven and physics-based models for protein formulation developability assessment

A Novel In Silico Platform Combining Data-Driven and Physics-Based Models for Protein Formulation Developability Assessment

Early identification of formulation risks is essential for accelerating biopharmaceutical development and reducing costly experimental screening. In silico approaches offer an attractive opportunity to predict developability challenges before extensive laboratory work is required.

In this poster, Coriolis Pharma presents an innovative platform that combines data-driven machine learning models with physics-based molecular simulations to assess protein formulation developability. Using only sequence and structural information, the approach evaluates key risk factors such as self-association, aggregation propensity, viscosity, oxidation, and deamidation, while providing guidance for formulation optimization. Through case studies involving monoclonal antibodies and a therapeutic protein, the platform demonstrates its ability to identify potential developability concerns, predict formulation-related risks, and support informed decision-making during candidate selection and formulation development.

The results highlight how integrating artificial intelligence with molecular-level modeling can help de-risk development programs, reduce experimental burden, and enable a more efficient path toward robust biopharmaceutical formulations.

View the poster to learn how advanced in silico tools can support developability assessment and accelerate formulation development for therapeutic proteins.