Pengwei Hu

V3Cell predicts organoid responses to chemical perturbations using non-invasive bright-field microscopy, creating virtual 3D cell models that mimic real organoids in structure and developmental dynamics.

This paper introduces a geometric deep learning framework for drug repositioning that leverages heterogeneous information networks to predict drug-disease associations.

The paper proposes iGRLDTI, a graph representation learning method for drug-target interaction prediction, achieving better performance than existing models on benchmark datasets.