Wengong Jin

FlashBind improves drug discovery by offering faster and accurate protein-ligand interaction predictions compared to Boltz-2, with real-world validation in antibiotic screening.

This paper introduces Boltz2ESI, a framework predicting enzyme-substrate interactions using a foundation model, outperforming existing methods by capturing active-site plasticity and evolutionary context.

spRefine denoises and imputes spatial transcriptomics data using a reference-free deep learning framework, improving data integration and revealing new aging-related biological insights.

A masked-expression Transformer trained on 983 paired H&E and spatial transcriptomics slides predicts 38,984 genes from histology without fine-tuning, leading HEST-Bench and improving survival and mutation prediction.

DSMBind predicts binding energy and designs nanobodies without experimental data, achieving performance comparable to supervised models.