Jason Zhang, an assistant professor of bioengineering at the UCLA Samueli School of Engineering, is using generative AI to design proteins that have never existed in nature. Zhang, a member of the California NanoSystems Institute, previously worked as a postdoctoral researcher in David Baker's lab at the University of Washington — the lab whose computational protein engineering work earned Baker the 2024 Nobel Prize in Chemistry.
Zhang's lab develops AI models trained on experimental data to generate new protein designs. The process begins with a biological target, the AI generates candidate amino acid sequences, those sequences are reverse-translated into DNA and expressed in the wet lab. His team can test 20,000 designed proteins in a single experiment to see whether they bind to their target.
One major focus is on disordered proteins — proteins that lack a fixed 3D shape and are implicated in neurodegenerative diseases, diabetes, and certain cancers. Because these proteins have no stable pocket for small-molecule drugs to bind to, they have traditionally been considered undruggable. Zhang's approach flips the problem: his AI creates new protein folds that have a pocket designed to capture the disordered protein, forcing it into an ordered state.
The lab is also pursuing what Zhang calls the virtual cell — a digital twin of human cells that could predict how various drugs affect immune cells or cancer cells. Additionally, Zhang collaborates with UCLA medical researchers on CAR-T cell therapies targeting rare liver cancers.
Zhang's work is part of a broader movement in AI-enabled protein design that builds on advances from institutions like the University of Washington and DeepMind's AlphaFold.
