MoleculeNet
ESOL
Structure
1,128Graphs
13Nodes
14Edges
1Classes
9Node Feature Dim
3Edge Feature Dim
Coverage
186Results
1Variants
26Papers
- BiScale-GTR: Fragment-Aware Graph Transformers for Multi-Scale Molecular Representation Learning BiScale-GTR: Fragment-Aware Graph Transformers for Multi-Scale Molecular Representation Learning
- Hierarchical Molecular Representation Learning via Fragment-Based Self-Supervised Embedding Prediction Hierarchical Molecular Representation Learning via Fragment-Based Self-Supervised Embedding Prediction
- MolGA: Molecular Graph Adaptation with Pre-trained 2D Graph Encoder MolGA: Molecular Graph Adaptation with Pre-trained 2D Graph Encoder
- ICLR 2025: NExT-Mol: 3D Diffusion Meets 1D Language Modeling for 3D Molecule Generation NExT-Mol: 3D Diffusion Meets 1D Language Modeling for 3D Molecule Generation ICLR 2025
- AAAI 2024: Pre-training Graph Neural Networks on Molecules by Using Subgraph-Conditioned Graph Information Bottleneck Pre-training Graph Neural Networks on Molecules by Using Subgraph-Conditioned Graph Information Bottleneck AAAI 2024
- ICML 2024: UniCorn: A Unified Contrastive Learning Approach for Multi-view Molecular Representation Learning UniCorn: A Unified Contrastive Learning Approach for Multi-view Molecular Representation Learning ICML 2024
- ICML 2024: Structure-Aware E(3)-Invariant Molecular Conformer Aggregation Networks Structure-Aware E(3)-Invariant Molecular Conformer Aggregation Networks ICML 2024
- NeurIPS 2023: SE(3)-Invariant Multiparameter Persistent Homology for Chiral-Sensitive Molecular Property Prediction SE(3)-Invariant Multiparameter Persistent Homology for Chiral-Sensitive Molecular Property Prediction NeurIPS 2023
- ICML 2023: Graph Neural Networks with a Distribution of Parametrized Graphs Graph Neural Networks with a Distribution of Parametrized Graphs ICML 2023
- Atom-Motif Contrastive Transformer for Molecular Property Prediction Atom-Motif Contrastive Transformer for Molecular Property Prediction
- Deep Prompt Tuning for Graph Transformers Deep Prompt Tuning for Graph Transformers
- AAAI 2023: 3D Denoisers are Good 2D Teachers: Molecular Pretraining via Denoising and Cross-Modal Distillation 3D Denoisers are Good 2D Teachers: Molecular Pretraining via Denoising and Cross-Modal Distillation AAAI 2023
- Motif-aware Attribute Masking for Molecular Graph Pre-training Motif-aware Attribute Masking for Molecular Graph Pre-training
- KDD 2023: On Structural Expressive Power of Graph Transformers On Structural Expressive Power of Graph Transformers KDD 2023
- 3D Molecular Geometry Analysis with 2D Graphs 3D Molecular Geometry Analysis with 2D Graphs
- Improving Molecular Pretraining with Complementary Featurizations Improving Molecular Pretraining with Complementary Featurizations
- Improving Molecular Contrastive Learning via Faulty Negative Mitigation and Decomposed Fragment Contrast Improving Molecular Contrastive Learning via Faulty Negative Mitigation and Decomposed Fragment Contrast
- OOD-GNN: Out-of-Distribution Generalized Graph Neural Network OOD-GNN: Out-of-Distribution Generalized Graph Neural Network
- AAAI 2021: Molecular Contrastive Learning with Chemical Element Knowledge Graph Molecular Contrastive Learning with Chemical Element Knowledge Graph AAAI 2021
- ICML 2021: 3D Infomax improves GNNs for Molecular Property Prediction 3D Infomax improves GNNs for Molecular Property Prediction ICML 2021
- ICLR 2021: 16 Pre-training Molecular Graph Representation with 3D Geometry 16 Pre-training Molecular Graph Representation with 3D Geometry ICLR 2021
- AAAI 2021: GeomGCL: Geometric Graph Contrastive Learning for Molecular Property Prediction GeomGCL: Geometric Graph Contrastive Learning for Molecular Property Prediction AAAI 2021
- ChemRL-GEM: Geometry Enhanced Molecular Representation Learning for Property Prediction ChemRL-GEM: Geometry Enhanced Molecular Representation Learning for Property Prediction
- Optimal Transport Graph Neural Networks Optimal Transport Graph Neural Networks
- ICLR 2020: Memory-Based Graph Networks Memory-Based Graph Networks ICLR 2020
- MoleculeNet MoleculeNet
Graph regression Aqueous solubility regression (log mol/L) on 1,128 molecules. SMILES and labels are fixed; molecular graph featurization is implementation-defined. Random and scaffold splits are not comparable. Evaluated by RMSE.
Dataset details
Task
Aqueous solubility regression (log mol/L) on 1,128 molecules. The primary catalog metric is RMSE.
Data
Aqueous solubility regression (log mol/L) on 1,128 molecules.
Features
MoleculeNet distributes SMILES and targets, not one mandatory graph featurization. A graph implementation normally derives atom and bond fields with RDKit or DeepChem; results depend on featurizer, salt handling, stereochemistry, and invalid-molecule policy.
Splits and evaluation
The catalog records these protocols or variants: Scaffold split. Evaluation uses RMSE (lower is better). Exact masks or folds must come from the cited release.
Quirks and pitfalls
Small samples, imbalance, missing assay labels, duplicate compounds, and scaffold leakage are common. Missing labels are not negatives. Always report split algorithm, featurizer, metric aggregation, and dataset version; random and scaffold results are not comparable.
Sources
SOTA Progression
new SOTA
matches SOTA at that time
other proposed result
other baseline result
Leaderboard
| # | Model | RMSE ↓ | Paper |
|---|---|---|---|
| 1 | Graphormer-SPIS | 0.4840
±
0.0050 | |
| 2 | Graphormer-SPD | 0.4920
±
0.0040 | |
| 3 | conan-FGW | 0.5140
±
0.0190 | |
| 4 | MemGNN | 0.5400
±
0.0100 | |
| 5 | GAT | 0.5520
±
0.0070 | |
| 6 | GeomMPNN | 0.5550 | |
| 7 | GCN | 0.5640
±
0.0150 | |
| 8 | LiGhT FT | 0.5650
±
0.0410 | |
| 9 | GeomGCL | 0.5750 | |
| 10 | AttentiveFP | 0.5780 | |
| 11 | LiGhT DeepGPT | 0.5790
±
0.0380 | |
| 12 | MoleculeNet | 0.5800 | |
| 13 | MPNN | 0.5800
±
0.0300 | |
| 14 | KCL | 0.5820 | |
| 15 | CoMPT | 0.5890 | |
| 16 | conan | 0.5910
±
0.0250 | |
| 17 | ProtoW-Dot | 0.5940
±
0.0310 | |
| 18 | GraphSAGE | 0.6010
±
0.0080 | |
| 19 | ProtoW-L2 | 0.6050
±
0.0290 | |
| 20 | Weave | 0.6100
±
0.0700 | |
| 21 | ProtoS-L2 | 0.6110
±
0.0340 | |
| 22 | Graphormer-id | 0.6120
±
0.0020 | |
| 23 | SE(3)-I MPPH | 0.6120
±
0.0830 | |
| 24 | MAT | 0.6240 | |
| 25 | DimeNet++ | 0.6330 | |
| 26 | SchNet-emb | 0.6350
±
0.0570 | |
| 27 | D-MPNN | 0.6350
±
0.0270 | |
| 28 | Graphormer-Neighbor | 0.6390
±
0.0340 | |
| 29 | DMPNN | 0.6470 | |
| 30 | GAT | 0.6540
±
0.0280 | |
| 31 | GIN | 0.6650
±
0.0260 | |
| 32 | SchNet-scalar | 0.6720
±
0.0270 | |
| 33 | D-MPNN+TopK Pool | 0.6730
±
0.0870 | |
| 34 | D-MPNN+SAG Pool | 0.6740
±
0.0340 | |
| 35 | ChemProp3D | 0.6810
±
0.0230 | |
| 36 | GraphGPS DeepGPT | 0.6850
±
0.1300 | |
| 37 | HMGNN | 0.7010 | |
| 38 | UniMol | 0.7070 | |
| 39 | MoLlama | 0.7400
±
0.0100 | |
| 40 | FPGNN | 0.7470 | |
| 41 | MPG | 0.7580 | |
| 42 | GraphGPS FT | 0.7720
±
0.1020 | |
| 43 | KANO | 0.7720 | |
| 44 | MolPROP | 0.7770
±
0.0200 | |
| 45 | UniMol | 0.7880
±
0.0290 | |
| 46 | ChemRL-GEM | 0.7980
±
0.0290 | |
| 47 | Geometry+Graph | 0.7980
±
0.0290 | |
| 48 | Graph | 0.8150
±
0.0250 | |
| 49 | AMCT | 0.8150
±
0.0100 | |
| 50 | S-CGIB | 0.8160
±
0.0190 |
SOTA performance
set a new SOTA at the time
matched the SOTA at the time