New MoleculeNet Datasets

[lilleswing]

Two years have passed since the publication of MoleculeNet. Since then many strides in supervised learning for molecules have been made. Are all of the datasets from the original paper still relevant to the challenges projects are facing today? Are there new datasets that should be added to the benchmark?

[rbharath]

We should add enamine as a dataset: https://enamine.net/library-synthesis/real-compounds/real-compound-libraries

[rbharath]

We should consider adding the crystallography open database. Following up on discussion from deepchem/deepchem#425

[rbharath]

We should consider adding the cambridge structural database. Following up on discussion from https://www.ccdc.cam.ac.uk/solutions/csd-system/components/csd/.

Following up on discussion from deepchem/deepchem#426.

[rbharath]

Following up on the discussion from deepchem/deepchem#867.

We should try to add some more assay binding data.