Molecular Property Prediction on MoleculeNet BBBP (scaffold)
96.4ROC AUCDumplingGNN
Evaluation Results
| Method | Links | |
|---|---|---|
| DumplingGNN2024.09 | 96.4 | |
| KPGT2023.04 | 90.8 | |
| SELFormer2023.04 | 90.2 | |
| Attentive FP2024.09 | 87.5 | |
| MGCN2023.04 | 85 | |
| SchNet2023.04 | 84.8 | |
| ChemBFN2024.09 | 80.5 | |
| DMP(TF+GNN)Model category=Specialist2023.11 | 77.8 | |
| MPTr+CIPEBackbone=MPTr, Positional encoding=CIPE2026.06 | 76.6 | |
| MolBERT2023.04 | 76.2 | |
| SYN-FUSIONfusion_source=Hu et. al2023.08 | 75.5 | |
| GraSPNet2026.02 | 74.4 | |
| SYN-FUSIONfusion_source=MolCLR2023.08 | 74.2 | |
| MolCLRbackbone=GIN2023.08 | 73.9 | |
| MolCLR2023.04 | 73.6 | |
| APTBase Model=Mole-BERT2023.11 | 73.1 | |
| SVM2023.08 | 72.9 | |
| MolFMModel category=Specialist2023.11 | 72.9 | |
| Uni-MolModel category=Specialist2023.11 | 72.9 | |
| MolXPT2024.09 | 72.9 | |
| ChemBERTa-22023.04 | 72.8 | |
| SimSGT2026.02 | 72.8 | |
| BGRL2026.02 | 72.5 | |
| GraphLogPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 72.5 | |
| Hi-GMAE-GPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 72.5 | |
| GraphLoG2026.06 | 72.5 | |
| GraphMVP-C2023.04 | 72.4 | |
| GEM2023.04 | 72.4 | |
| GraphMVP-CModel category=Specialist2023.11 | 72.4 | |
| Instruct-GSModel category=LLM Based Generalist2023.11 | 72.4 | |
| MolCLR2024.09 | 72.4 | |
| Uni-Mol2024.09 | 72.4 | |
| GEM2024.09 | 72.2 | |
| CP2026.02 | 72.1 | |
| GraphMAEPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 72 | |
| GraphMAE2026.06 | 72 | |
| Mole-BERT2023.11 | 71.9 | |
| Mole-BERTPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 71.9 | |
| GCN2023.08 | 71.8 | |
| GCN2023.04 | 71.8 | |
| JOAO2026.02 | 71.8 | |
| GraphMAE2026.02 | 71.7 | |
| MPTr+RWSEBackbone=MPTr, Positional encoding=RWSE2026.06 | 71.7 | |
| GraphMAE22026.02 | 71.6 | |
| GraphMAE2Pre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 71.6 | |
| GTOT-TuningBackbone=GIN (supervised_contextpred), Strategy=GTOT-Tuning2022.03 | 71.5 | |
| RF2023.08 | 71.4 | |
| GraphCL2026.02 | 71.4 | |
| MPTr+CycleSEBackbone=MPTr, Positional encoding=CycleSE2026.06 | 71.3 | |
| D-MPNN2023.08 | 71.2 | |
| SimGRACE2023.11 | 71.2 | |
| GraphMAE2023.11 | 71.2 | |
| SimGRACEPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 71.2 | |
| RGCLPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 71.2 | |
| D-MPNN2023.04 | 71 | |
| D-MPNN2024.09 | 71 | |
| Hi-GMAE-FPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 71 | |
| G2PT2026.06 | 71 | |
| MPTrBackbone=MPTr2026.06 | 70.9 | |
| Hu et al.2023.08 | 70.8 | |
| Hu et al.2023.04 | 70.8 | |
| MolCA(1D)Model category=LLM Based Generalist, Input modality=1D2023.11 | 70.8 | |
| GraphMVP2023.11 | 70.8 | |
| Mole-BERT2026.02 | 70.8 | |
| GraphMVPPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 70.8 | |
| AD-GCL2023.11 | 70.7 | |
| APTBase Model=EdgePred2023.11 | 70.7 | |
| ContextPred2023.11 | 70.6 | |
| KV-PLMModel category=Specialist2023.11 | 70.5 | |
| MoMuModel category=Specialist2023.11 | 70.5 | |
| ADGCL2026.02 | 70.5 | |
| Hi-GMAE-GPre-training status=None, Pre-training dataset=N/A, Evaluation protocol=Supervised Learning, Split type=Scaffold-split2024.05 | 70.4 | |
| UnifiedMolPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 70.4 | |
| JOAOPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 70.2 | |
| MolCA(1D + 2D)Model category=LLM Based Generalist, Input modality=1D + 2D2023.11 | 70 | |
| BSSBackbone=GIN (supervised_contextpred), Strategy=Fine-tuning strategy2022.03 | 70 | |
| G-Contextual2023.11 | 69.9 | |
| ChemBERTa v2Model category=Specialist2023.11 | 69.8 | |
| StochNormBackbone=GIN (supervised_contextpred), Strategy=Fine-tuning strategy2022.03 | 69.8 | |
| Chem-GMNetPre-trained=true, Pre-training Corpus=10M SMILES, Pre-training Task=MLM-10M2026.05 | 69.8 | |
| GraphCLModel category=Specialist2023.11 | 69.7 | |
| N-Gram RFClassifier=Random Forest2024.09 | 69.7 | |
| MGSSL2026.02 | 69.7 | |
| GraphCLPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 69.7 | |
| AD-GCLPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 69.7 | |
| GraphCLPre-training strategy=GraphCL2026.06 | 69.7 | |
| ChemBERTa-2Pre-trained=true, Pre-training Corpus=10M SMILES, Pre-training Task=MLM-10M2026.05 | 69.6 | |
| GraphMVP2024.09 | 69.5 | |
| Infomax2026.02 | 69.2 | |
| GraphGPSPre-training status=None, Pre-training dataset=N/A, Evaluation protocol=Supervised Learning, Split type=Scaffold-split2024.05 | 69.2 | |
| PretrainGNN2024.09 | 69.1 | |
| 3D InfoMax2023.11 | 69.1 | |
| GINGraph-level pre-training=None, Node-level pre-training=Infomax2019.05 | 68.8 | |
| MGSSL2023.11 | 68.8 | |
| InfomaxPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 68.8 | |
| SSL pre-trainingPre-training strategy=SSL2026.06 | 68.8 | |
| GROVER_largeModel scale=large2024.09 | 68.7 | |
| GINGraph-level pre-training=Supervised, Node-level pre-training=Context Prediction2019.05 | 68.7 | |
| Fine-TuningBackbone=GIN (supervised_contextpred), Strategy=Baseline2022.03 | 68.7 | |
| InfoGraph2023.11 | 68.7 |