PhD & Postdoc

Liu Yang

Artificial Intelligence · Tsinghua University

Liu Yang is a PhD candidate in Artificial Intelligence at Tsinghua University, supervised by Prof. Jun Zhu. His thesis develops conditional diffusion probabilistic models for protein structure prediction and de novo protein sequence design, achieving state-of-the-art performance on CASP15 targets and generating functional enzyme sequences validated by wet-lab collaborators. Fifteen publications in Nature Machine Intelligence, NeurIPS and ICML place him among the most cited PhD-stage contributors in generative protein modelling. Liu was named on Forbes China 30 Under 30 in Science and Technology and completed a research internship at DeepMind's protein structure team. He has received acquisition offers from two biotech companies developing AI-designed therapeutics. His diffusion model generated a novel lipase that outperforms the wild-type enzyme 3.5-fold at 70 degrees Celsius.

Thesis Topic

Conditional diffusion models for protein structure prediction and de novo enzyme design

Transition Signals