Xing Che

This paper identifies a novel non-canonical RT-loop degron in VAV1 and demonstrates how conformational restriction and computational modeling enhance molecular glue potency for proteasomal degradation.

Feynman-Kac steering with an ipTM reward at inference time lets Chai-1r recover three of eight molecular-glue ternary complexes below 3 Å without retraining.

YDS-GlueFold improves ternary complex prediction by enhancing AlphaFold 3 with guided diffusion, achieving high accuracy on novel protein pairs not seen during training.

FragLlama predicts next molecular fragments for drug design, generating molecules with desired properties and exploring novel chemical spaces.