Honglin Li

This paper introduces Mac-Diff, a diffusion model that generates diverse protein conformations by aligning sequence and geometry views, showing improved recovery of conformational distributions and meta-stable states.

The paper introduces IDKG, a knowledge graph integrating multimodal data to advance infectious disease research, demonstrating its utility in drug repurposing for hMPV infection.

The paper develops ConMeter, a metabolite prediction model, and uses it to design an orally bioavailable RSK4 inhibitor with improved bioavailability and anti-tumor efficacy against esophageal cancer.

CHIEF improves protein design by combining five models, achieving higher sequence recovery and lower perplexity, and successfully designing functional malate dehydrogenase.

AnnoPRO improves protein function annotation by using image-like representations and multimodal deep learning, showing better performance for rare protein families.