The UW Medicine Institute for Protein Design (IPD) today released RFD4-Proteina, a new open-source AI model designed to create proteins with specific functions. It is the first AI protein model the University of Washington has co-developed with Nvidia, and follows a major computational boost for the institute.
Earlier this month, IPD announced a substantial grant of computing power from the foundation of Nvidia CEO Jensen Huang and his wife, Lori Huang. The institute began accessing Nvidia GPUs through cloud provider CoreWeave in June, a grant expected to support its work for a year.
David Baker, director of IPD and a 2024 Nobel laureate in chemistry, said in a statement: “RFD4-Proteina will open the door to new solutions in medicine, sustainability, materials and biotechnology.” He previously told GeekWire that the foundation’s grant has been “transformational” for the institute’s work.
The model’s name nods to IPD’s earlier RFdiffusion tool and references Nvidia’s Proteina-Complexa computational models. According to the team, RFD4-Proteina is “a considerable improvement over previous methods,” running faster and requiring significantly less computational power. It enables researchers to design previously unimaginable proteins, with potential applications as drugs for challenging diseases or as tools for managing environmental waste such as toxic chemicals and plastics.
Seth Woodbury, a UW bioengineering predoctoral candidate who used RFD4-Proteina in his enzyme research, said: “In just a few days we designed a pesticide-degrading enzyme that performs many fold better than my previous best, which took months to design.” Woodbury published a paper in Nature on enzyme research last year.
The work was led by UW doctoral students Nathaniel Corley, Maxwell Kazman and Rafi Brent, along with Kieran Didi, a visiting student from the University of Oxford. RFD4-Proteina is available through Berkeley Software Distribution.