Molecular Property Prediction on MoleculeNet SIDER (scaffold)
0.849ROC-AUCMPTr+CIPE
Evaluation Results
| Method | Links | |
|---|---|---|
| MPTr+CIPEBackbone=MPTr, Positional encoding=CIPE2026.06 | 0.849 | |
| MPTr+HKSEBackbone=MPTr, Positional encoding=HKSE2026.06 | 0.843 | |
| MPTr+CycleSEBackbone=MPTr, Positional encoding=CycleSE2026.06 | 0.842 | |
| MPTr+RWSEBackbone=MPTr, Positional encoding=RWSE2026.06 | 0.839 | |
| MPTr+LapPEBackbone=MPTr, Positional encoding=LapPE2026.06 | 0.838 | |
| MPTrBackbone=MPTr2026.06 | 0.835 | |
| SELFormer2023.04 | 0.745 | |
| DumplingGNN2024.09 | 0.74 | |
| GraphConv + dummy super nodeGraph construction=dummy super node2024.09 | 0.735 | |
| SYN-FUSIONfusion_source=Hu et. al2023.08 | 0.699 | |
| RF2023.08 | 0.684 | |
| SVM2023.08 | 0.682 | |
| GEM2023.04 | 0.672 | |
| GROVER_largeModel scale=large2024.09 | 0.672 | |
| MolXPT2024.09 | 0.672 | |
| N-Gram RFClassifier=Random Forest2024.09 | 0.668 | |
| GraphMVP2024.09 | 0.659 | |
| Uni-Mol2024.09 | 0.659 | |
| N-Gram XGBClassifier=XGBoost2024.09 | 0.655 | |
| PretrainGNN2024.09 | 0.654 | |
| Hu et al.2023.08 | 0.652 | |
| Hu et al.2023.04 | 0.652 | |
| MolCLR2023.04 | 0.652 | |
| SYN-FUSIONfusion_source=MolCLR2023.08 | 0.65 | |
| MolCLRbackbone=GIN2023.08 | 0.649 | |
| KPGT2023.04 | 0.649 | |
| GraphMVP-C2023.04 | 0.639 | |
| GINGraph-level pre-training=Supervised, Node-level pre-training=Attribute Masking2019.05 | 0.639 | |
| L2_SPBackbone=GIN (supervised_contextpred), Strategy=Fine-tuning strategy2022.03 | 0.638 | |
| GINGraph-level pre-training=Supervised, Node-level pre-training=EdgePred2019.05 | 0.633 | |
| GTOT-TuningBackbone=GIN (supervised_contextpred), Strategy=GTOT-Tuning2022.03 | 0.633 | |
| D-MPNN2023.08 | 0.632 | |
| Feature (DELTA w/o ATT)Backbone=GIN (supervised_contextpred), Strategy=Fine-tuning strategy2022.03 | 0.63 | |
| DELTABackbone=GIN (supervised_contextpred), Strategy=Fine-tuning strategy2022.03 | 0.629 | |
| BSSBackbone=GIN (supervised_contextpred), Strategy=Fine-tuning strategy2022.03 | 0.628 | |
| Mole-BERT2023.11 | 0.628 | |
| Mole-BERTPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 0.628 | |
| GINGraph-level pre-training=Supervised, Node-level pre-training=Context Prediction2019.05 | 0.627 | |
| Fine-TuningBackbone=GIN (supervised_contextpred), Strategy=Baseline2022.03 | 0.627 | |
| GraSPNet2026.02 | 0.625 | |
| StochNormBackbone=GIN (supervised_contextpred), Strategy=Fine-tuning strategy2022.03 | 0.622 | |
| UnifiedMolPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 0.622 | |
| GINGraph-level pre-training=Supervised, Node-level pre-training=None2019.05 | 0.621 | |
| Hi-GMAE-FPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 0.62 | |
| G2PT2026.06 | 0.619 | |
| MGSSL2023.11 | 0.616 | |
| AD-GCL2023.11 | 0.615 | |
| MGSSL2026.02 | 0.615 | |
| AD-GCLPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 0.615 | |
| MAM2023.11 | 0.614 | |
| APTBase Model=EdgePred2023.11 | 0.614 | |
| GraphLogPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 0.612 | |
| RGCLPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 0.612 | |
| GraphLoG2026.06 | 0.612 | |
| GPSEaugmented2026.06 | 0.611 | |
| GINGraph-level pre-training=None, Node-level pre-training=Attribute Masking2019.05 | 0.61 | |
| G-Motif2023.11 | 0.61 | |
| AttrMaskingPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 0.61 | |
| SSL pre-trainingPre-training strategy=SSL2026.06 | 0.61 | |
| GINGraph-level pre-training=None, Node-level pre-training=Context Prediction2019.05 | 0.609 | |
| GINGraph-level pre-training=Supervised, Node-level pre-training=Infomax2019.05 | 0.609 | |
| GATPre-training status=None, Pre-training dataset=N/A, Evaluation protocol=Supervised Learning, Split type=Scaffold-split2024.05 | 0.609 | |
| ContextPredPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 0.609 | |
| APTBase Model=Mole-BERT2023.11 | 0.608 | |
| JOAO2026.02 | 0.608 | |
| MAMEncoder=vanilla VQ-VAE2023.11 | 0.607 | |
| Attentive FP2024.09 | 0.606 | |
| SimSGT2026.02 | 0.606 | |
| AttentiveFP2026.06 | 0.606 | |
| GraphMAE2023.11 | 0.605 | |
| AttrMask2023.11 | 0.605 | |
| SAGPoolPre-training status=None, Pre-training dataset=N/A, Evaluation protocol=Supervised Learning, Split type=Scaffold-split2024.05 | 0.605 | |
| GraphCLPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 0.605 | |
| Hi-GMAE-GPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 0.605 | |
| GraphCLPre-training strategy=GraphCL2026.06 | 0.605 | |
| GINGraph-level pre-training=None, Node-level pre-training=EdgePred2019.05 | 0.604 | |
| JOAO2023.11 | 0.604 | |
| EdgePred2023.11 | 0.604 | |
| BGRL2026.02 | 0.604 | |
| GraphMAEPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 0.603 | |
| GraphMAE2026.06 | 0.603 | |
| SimGRACE2023.11 | 0.602 | |
| GraphMVP2023.11 | 0.602 | |
| CP2026.02 | 0.602 | |
| SimGRACEPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 0.602 | |
| GraphMVPPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 0.602 | |
| GraphGPS2026.06 | 0.602 | |
| NID_CLEvaluation Protocol=Linear Probing, Pre-training Dataset=ZINC15, Backbone=5-layer GIN2024.05 | 0.601 | |
| Infomax2026.02 | 0.601 | |
| GraphMAE2026.02 | 0.6 | |
| GCNPre-training status=None, Pre-training dataset=N/A, Evaluation protocol=Supervised Learning, Split type=Scaffold-split2024.05 | 0.6 | |
| JOAOPre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 0.6 | |
| GraphCL2023.11 | 0.598 | |
| GraphLOG2026.02 | 0.598 | |
| ContextPred2023.11 | 0.597 | |
| GraphLoG2023.11 | 0.596 | |
| TMCL2023.11 | 0.596 | |
| Attr mask2026.02 | 0.596 | |
| GraphMAE22026.02 | 0.596 | |
| GraphMAE2Pre-training status=Pre-trained, Pre-training dataset=ZINC15, Evaluation protocol=Fine-tuning, Split type=Scaffold-split2024.05 | 0.596 |