3 datasets found

The dataset is used for compound-protein binding affinity prediction, and it contains 218,615 compound-protein pairs.

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

The ChEMBL dataset is a large collection of bioactive molecules, with over 10 million molecules, that can be used for various machine learning tasks, including molecular design.