My lab emphasizes developing Mutexa, a physics-informed artificial intelligence (AI) platform for “intelligent” protein engineering, enabling researchers to identify function-enhancing mutants while uncovering the molecular insights behind unpredictable experimental outcomes. Protein engineering, despite over decades of progress, remains reliant on labor- and resource-intensive experimental screening, which delivers only “what you screen for” and offers little insight into the structure-function relationships underlying mutation effects. While AI is widely recognized for its potential to accelerate protein engineering, I question the feasibility of achieving generalizable predictive models through AI alone. I envision the future of “intelligent protein engineering” lies in combining AI with physics-based modeling, where molecular-level protein dynamics, electrostatics, and other microscopic features illuminate sequence-function relationships underlying the mutation effects, enabling the creation of generalizable models for computational protein engineering with high fidelity.

In this talk, I will present the technical foundations of Mutexa and its application in two protein engineering challenges: one for designing industrial bidomain enzymes that maintain high activity at lower temperatures (known as cold-adapted enzymes), and the other for predicting the structures of lasso peptides, a class of ribosomally synthesized and post-translationally modified peptides, as antibiotics. These applications showcase Mutexa’s unique ability to drive the discovery of functional proteins beyond traditional screening-based approaches, offering solutions for sustainable biomanufacturing and antimicrobial development.