2

Developed a novel deep graph learning method for drug-target interaction prediction. Demonstrated an application of our method to repurpose existing drugs against the COVID-19 virus.

Developed a methodology and related algorithms to automatically induce a hierarchical structure of disease modules from proximity relations in the interactome network, and to align, label, and systematically compare this structure with a manually defined disease ontology.

Developed a deep graph learning-based model for predicting disease-gene associations. The model jointly learns functional and connectivity patterns using a novel technique based on attributed random walks.