Federal Project Grant Summary

Trustees of Dartmouth College received an $831,958 Project Grant award on August 15, 2026, from the National Science Foundation (NSF) Division of Biological Infrastructure under the Biological Sciences program (CFDA 47.074). The award funds development of TRANSFORMD, a multiscale machine learning platform designed to integrate molecular dynamics simulations into artificial intelligence (AI) models for protein engineering. The platform addresses a critical gap in current AI approaches by incorporating protein molecular motion and flexibility into mutation effect prediction models, thereby improving accuracy and practical utility for protein engineers. The project extends through July 31, 2030, and is being executed at Dartmouth College's location in Hanover, New Hampshire.

The deliverables include development of software tools that enable researchers without advanced computing expertise to leverage physics-based molecular simulation data for training protein mutation prediction models. A key product is an explainable AI system that elucidates the mechanistic reasons underlying how specific mutations affect protein properties such as catalytic rates, binding affinities, and fold stability. The platform is designed to serve dual purposes: enhancing predictive accuracy for biotechnology applications ranging from chemical manufacturing to gene therapy, and providing educational value for teaching students protein engineering principles. By bridging molecular simulation science with accessible machine learning tools, the award supports both fundamental research advancement and broader dissemination of protein engineering methodologies.