About
I am a Ph.D. student at Shanghai Jiao Tong University, working at the intersection of artificial intelligence, enzyme engineering, and computational drug discovery.
My research develops structure-aware learning systems and scientific agents that connect molecular representations with practical problems in biocatalysis, biosynthesis, and therapeutic discovery.
I focus on building models that are useful beyond benchmark performance: systems that can retrieve relevant enzymes, generate reaction-conditioned protein designs, reason over biosynthetic pathways, and support iterative scientific decision-making. A recurring theme in my work is to make molecular AI more grounded in structure, mechanism, and executable discovery workflows.
Across my recent work, I am interested in systems that reason across protein structures, reactions, and molecular transformations, turning those connections into practical workflows for biocatalysis, biosynthesis, and therapeutic target discovery.
Before beginning my Ph.D., I earned an M.S. from the Hong Kong University of Science and Technology and a B.A.Sc. from the Beijing Institute of Technology.
Outside academia, I spent nearly three years at Galixir, developing computational pipelines for virtual screening, molecular generation, and retrosynthesis.
Interests
- Enzyme foundation models
- Reaction-conditioned enzyme design
- Generative protein design
- Biosynthesis and pathway discovery
- Scientific LLM agents
- Molecular discovery workflows
Education
Shanghai Jiao Tong University
Ph.D. Student
Hong Kong University of Science and Technology
M.S., Biomolecular Engineering & Health Informatics
Beijing Institute of Technology
B.A.Sc., Computer Science & Technology
Selected Publications
- Nature Catalysis
A structure-aware enzyme retrieval model that connects geometric protein representations with evolutionary signals for enzyme discovery.
- bioRxiv
A self-evolving virtual disease biologist for automating therapeutic target discovery and hypothesis generation.
- arXiv
A flow-matching approach for generating reaction-specific catalytic pockets with co-evolutionary dynamics.