Chembl Training Data - Masterthesis - Mario Surlemont

This here is a dataset that I compiled to train and validate a machine learning model in the context of my Masterthesis. It contains a csv file that holds (lipophile) molecules from the chembl database as SMILES strings. Additionally it contains processed files that store a certain pharmacophoric representation of the molecules. The pharmacophores are either stored as pointclouds (in the 3D-space) or corresponding graphs. It also contains files that represent the same pharmacophoric representation for the interaction pharmacophore of a few lipophile targets of the DUD-E dataset and their acttives and decoys as well.