configs:
- config_name: split
data_files: data.csv
- config_name: labels
data_files: labels/*.csv
license: mit
Overview
We present MoleculeCLA: a large-scale dataset consisting of approximately 140,000 small molecules derived from computational ligand-target binding analysis, providing nine properties that cover chemical, physical, and biological aspects.
| Aspect | Glide Property (Abbreviation) | Description | Molecular Characteristics |
|---|---|---|---|
| Chemical | glide_lipo (lipo) | Hydrophobicity | Atom type, number |
| glide_hbond (hbond) | Hydrogen bond formation propensity | Atom type, number | |
| Physical | glide_evdw (evdw) | Van der Waals energy | Size and polarizability |
| glide_ecoul (ecoul) | Coulomb energy | Ionic state | |
| glide_esite (esite) | Polar thermodynamic contribution | Polarity | |
| glide_erotb (erotb) | Rotatable bond constraint energy | Rotational flexibility | |
| glide_einternal (einternal) | Internal torsional energy | Rotational flexibility | |
| Biological | docking_score (docking) | Docking score | Binding affinity |
| glide_emodel (emodel) | Model energy | Binding affinity |
Data Format
The 'data.csv' file contains information on scaffold splitting for training, testing, and validation sets, along with the SMILES representations of molecules and their corresponding molecular IDs for identification.
The 'labels/*.csv' file contains data on molecular properties derived from binding analysis, along with their corresponding molecule IDs, Each file name corresponds to a specific protein target name.
The 'diversity_molecule_set.pkl' file contains the 3D coordinates of molecules, necessary for 3D-based molecular representation learning methods.