Property Prediction & ADMET
ML models for molecular property prediction, ADMET profiling, and chemistry language models.
Tutorials
Transformers for Chemistry
Jupyter
LLMs
Transformers
DeepChem Tutorials
Jupyter
GNNs
Molecular property prediction
Tools
Property Prediction & QSAR
DeepChem
Comprehensive ML library for chemistry and life sciences
Chemprop
Message passing neural networks for molecular property prediction
DGL-LifeSci
GNN library for life science applications
scikit-mol
RDKit and scikit-learn bridge
MolFeat
Molecular featurization and embeddings
olorenchemengine
Unified API for diverse models with uncertainty quantification
ROBERT
Ensemble of automated ML protocols
chemml
ML and informatics for chemical and materials data
DScribe
Descriptor library with various fingerprinting techniques
MORFEUS
Fast calculations of molecular features from 3D structures
ADMET & Drug Discovery
TDC
Therapeutics Data Commons — ML benchmarks for drug discovery
oddt
Open Drug Discovery Toolkit
DescriptaStorus
Descriptor computation for machine learning
TorchDrug
GNN-based library
DeepPurpose
Deep learning for compound and protein modeling
SwissADME
Drug-likeness and PK properties
ADMET-AI
ADMET prediction based on TDC data
ADMETlab 2.0
PK and toxicity predictions
Language Models
ChemBERT
Chemical reaction extraction (Base)
ChemDFM
Chemistry dialogue foundation model (13B)
ChemLLM
Chemical large language model (7B)
MolT5
Molecule-language translation (60M–770M)
MoleculeSTM
Multi-modal structure-text model
Datasets
Property Prediction
AquaSolDB
Curated aqueous solubility
ESol
Water solubility for organic small molecules
Lipophilicity
Octanol/water distribution coefficients
Photoswitch Dataset
405 curated photoswitch molecules