Molecular Property Prediction on MoleculeNet Tox21 (scaffold)
85.4ROC-AUCGraphConv + dummy super node
Evaluation Results
| Method | Links | |
|---|---|---|
| GraphConv + dummy super nodeGraph construction=dummy super node2024.09 | 85.4 | |
| KPGT2023.04 | 84.8 | |
| DumplingGNN2024.09 | 82.3 | |
| SVM2023.08 | 81.8 | |
| MolXPT2024.09 | 79.6 | |
| Hu et al.2023.08 | 78.7 | |
| Hu et al.2023.04 | 78.7 | |
| MolCLR2023.04 | 78.7 | |
| L2_SPBackbone=GIN (supervised_contextpred), Strategy=Fine-tuning strategy2022.03 | 78.7 | |
| GTOT-TuningBackbone=GIN (supervised_contextpred), Strategy=GTOT-Tuning2022.03 | 78.6 | |
| StochNormBackbone=GIN (supervised_contextpred), Strategy=Fine-tuning strategy2022.03 | 78.4 | |
| BSSBackbone=GIN (supervised_contextpred), Strategy=Fine-tuning strategy2022.03 | 78.3 | |
| GEM2023.04 | 78.1 | |
| GROVER_largeModel scale=large2024.09 | 78.1 | |
| Uni-Mol2024.09 | 78.1 | |
| Fine-TuningBackbone=GIN (supervised_contextpred), Strategy=Baseline2022.03 | 78.1 | |
| DELTABackbone=GIN (supervised_contextpred), Strategy=Fine-tuning strategy2022.03 | 77.9 | |
| Feature (DELTA w/o ATT)Backbone=GIN (supervised_contextpred), Strategy=Fine-tuning strategy2022.03 | 77.9 | |
| SchNet2023.04 | 77.2 | |
| N-Gram XGBClassifier=XGBoost2024.09 | 77.2 | |
| ChemBFN2024.09 | 77.1 | |
| RF2023.08 | 76.9 | |
| Attentive FP2024.09 | 76.3 | |
| D-MPNN2023.04 | 75.9 | |
| D-MPNN2024.09 | 75.9 | |
| MolCLR2024.09 | 75.9 | |
| PretrainGNN2024.09 | 75.8 | |
| SYN-FUSIONfusion_source=MolCLR2023.08 | 75.1 | |
| GEM2024.09 | 75 | |
| GraphMVP-C2023.04 | 74.4 | |
| N-Gram RFClassifier=Random Forest2024.09 | 74.3 | |
| GIN2023.08 | 74 | |
| GIN2023.04 | 74 | |
| SYN-FUSIONfusion_source=Hu et. al2023.08 | 73.8 | |
| GraphMVP2024.09 | 73.5 | |
| MolCLRbackbone=GIN2023.08 | 72 | |
| GCN2023.08 | 70.9 | |
| GCN2023.04 | 70.9 | |
| MGCN2023.04 | 70.7 | |
| JOAOEvaluation Protocol=Linear Probing, Pre-training Dataset=ZINC15, Backbone=5-layer GIN2024.05 | 70.6 | |
| AttrMaskEvaluation Protocol=Linear Probing, Pre-training Dataset=ZINC15, Backbone=5-layer GIN2024.05 | 69.1 | |
| D-MPNN2023.08 | 68.9 | |
| ContextPredEvaluation Protocol=Linear Probing, Pre-training Dataset=ZINC15, Backbone=5-layer GIN2024.05 | 68.4 | |
| NID_CLEvaluation Protocol=Linear Probing, Pre-training Dataset=ZINC15, Backbone=5-layer GIN2024.05 | 66.3 | |
| SELFormer2023.04 | 65.3 | |
| SimGRACEEvaluation Protocol=Linear Probing, Pre-training Dataset=ZINC15, Backbone=5-layer GIN2024.05 | 64.6 | |
| GraphCLEvaluation Protocol=Linear Probing, Pre-training Dataset=ZINC15, Backbone=5-layer GIN2024.05 | 64.4 | |
| EdgePredEvaluation Protocol=Linear Probing, Pre-training Dataset=ZINC15, Backbone=5-layer GIN2024.05 | 62.7 |