DrugBank

The dataset used in the paper is a sparse user-item interaction matrix, a protein-protein similarity matrix, and a drug-drug similarity matrix.

BibTex: