Dan Teng

MG2Act predicts molecular glue degradation by modeling sequential interactions between E3 ligase and substrate, outperforming baselines and identifying new degraders.

This study shows that resistance to different WRN inhibitors in MSI-H cancers creates orthogonal vulnerabilities, enabling alternating therapy and mutation-tolerant inhibitor design.

This paper introduces MolGenBench, a benchmark for evaluating molecular generative models in drug discovery, highlighting gaps between current models and real-world pharmaceutical workflows.

PBCNet2.0 predicts protein-ligand binding affinity using a cartesian tensor-based Siamese Neural Network, achieving high accuracy and efficiency compared to physics-based simulations.