Table 1 Databases and software for DL atomistic design (‘k’, ‘mil’ = thousand, million).
From: Recent advances and applications of deep learning methods in materials science
Databases | |||
|---|---|---|---|
DB name | Datasize | Link | Ref. |
JARVIS-DFT | 56k | ||
JARVIS-FF | 2.5k | ||
MP | 144k | ||
OQMD | 816k | ||
AFLOW | 3.5mil | ||
QM9 | 134k | ||
ANI | 20mil | ||
MD17 | 1mil | ||
Tox21 | 760k | ||
CCCBDB | 2069 | ||
HOPV15 | 350 | ||
C2DB | 4000 | ||
FreeSolv | 504 | ||
NOMAD | 11mil | ||
OPTIMADE | 18mil | ||
Open catalyst | |||
project | 1.2mil | ||
MatBench | 200k | ||
MCloud | 22mil | ||
CoreMOF | 163k | ||
QMOF | 22k | ||
PDB | 183k | ||
PDBBind | 23k | ||
MOAD | 39k | ||
Software packages | |||
|---|---|---|---|
Model name | Applications | Link | Ref. |
ALIGNN | Mol, Sol | ||
SchNetPack | Mol, Sol | ||
CGCNN | Sol | ||
MEGNet | Mol, Sol | ||
DimeNet | Mol | ||
MPNN | Mol | ||
MatDeepLearn | Sol | ||
GATGCNN | Sol | ||
ANI | Mol | ||
Amp | Sol | ||
TensorMol | Mol | ||
TorchMD | Mol | ||
PROPhet | Sol | ||
DeepMD | Mol | ||
ænet | Sol | ||
E3NN | Mol | ||
Neural | |||
fingerprint | Mol | ||
DeepChemSt. | Mol | ||
MoleculeNet | Mol, Sol | ||
dgl-lifesci | Prot | ||
gnina | Prot | ||