Kunyang Sun

LinkLlama improves linker design by combining natural language prompts with 3D spatial awareness, achieving higher chemical validity than existing methods.

CAFE improves co-folding models for discovering allosteric and cryptic binding sites by using orthosteric blockers to guide fragment exploration.

The paper investigates how competing factors in biomolecular condensates modulate the rate of poly(A) RNA deadenylation by CNOT7, revealing concentration-dependent increases and diffusion-related decreases in enzymatic efficiency.

This paper introduces KinConfBench, a benchmark of 2,225 human kinase chains to evaluate cofolding models' ability to recover conformational states, finding they struggle with structural diversity and ligand-induced motions.

DrugPlayGround benchmarks large language models for drug discovery tasks like drug characteristics and interactions, assessing their performance against traditional methods.

AlphaFlex generates atomistic conformational ensembles for human proteins with intrinsically disordered regions, integrating AlphaFold models and implicit bilayers for transmembrane proteins.