List your company
ResearchFormerly WizWand

Molecular Property Prediction on Tox21

90.7ROC AUCMolSight
47.5458.74569.9581.155May 22, 2022Jan 27, 2023Oct 5, 2023Jun 11, 2024Feb 17, 2025Oct 25, 2025Jul 3, 2026
Updated 4mo ago

Evaluation results

MethodLinks
MolSightInput Modality=2D image, Training Curriculum=S62026.0590.7
MvMRL2026.0384.5
MPNN2026.0384.4
N-Gram2026.0384.2
GROVER2026.0384.2
DMMRL2026.0384.2
SGGRL2026.0383.7
GCN2026.0382.9
Weave2026.0382
FP-GNN2026.0381.5
Transformer2026.0381.3
GraSeq2026.0381
RNN2026.0380.6
ChemATP2025.1280.23
GROVERModel Type=Graph-based2026.0680
UniMol2025.1279.6
MolCLRCMPNNModel Type=Graph-based2026.0678.4
MOLEBLENDInput Modality=Molecular graph2026.0577.8
MOLINTERACTInput Modality=Molecular graph2026.0577.3
Uni-MolInput Modality=Molecular graph, uses ground-truth 3D conformers=true2026.0577
MoleculeSDEInput Modality=Molecular graph2026.0576.8
AttrMaskingPre-training dataset=ZINC15, Protocol=Fine-tuning2022.0576.7
AttrMaskingModel=AttrMasking2026.0476.7
MolCLRGINModel Type=Graph-based2026.0676.7
SchNetModel Type=Graph-based2026.0676.6
AD-GCLModel=AD-GCL2026.0476.54
CHCLModel=CHCL2026.0476.06
EdgePredModel=EdgePred2026.0476
GraphMVP2025.1275.9
GraphormerType=Graph-based / Full Supervision2026.0775.9
ContextPredPre-training dataset=ZINC15, Protocol=Fine-tuning2022.0575.7
GraphLoGPre-training dataset=ZINC15, Protocol=Fine-tuning2022.0575.7
ContextPredModel=ContextPred2026.0475.7
AutoGCLModel=AutoGCL2026.0475.69
MoMu2025.1275.6
GraphMAEPre-training dataset=ZINC15, Protocol=Fine-tuning2022.0575.5
GATType=Graph-based / Full Supervision2026.0775.4
InfomaxPre-training dataset=ZINC15, Protocol=Fine-tuning2022.0575.3
InfomaxModel=Infomax2026.0475.3
JOAOPre-training dataset=ZINC15, Protocol=Fine-tuning2022.0575
JOAOModel=JOAO2026.0474.98
GMCLModel=GMCL2026.0474.94
GCNType=Graph-based / Full Supervision2026.0774.9
LLM4SD2025.1274.7
InstructMol2025.1274.7
GraphMVPInput Modality=Molecular graph2026.0574.5
3D InfoMaxInput Modality=Molecular graph2026.0574.5
No-pretrainPre-training dataset=None, Protocol=Supervised2022.0574.3
No PretrainModel=No Pretrain2026.0474
GINType=Graph-based / Full Supervision2026.0774
GraphCLPre-training dataset=ZINC15, Protocol=Fine-tuning2022.0573.9
GraphCLModel=GraphCL2026.0473.87
MolCLRGCNModel Type=Graph-based2026.0673.5
MGCNModel Type=Graph-based2026.0670.7
ChemDFM14BStrategy=+MOLE-RAG2026.0668.8
GPT-4oStrategy=+MOLE-RAG2026.0667
MolRAG2025.1266.42
Qwen34B-InstStrategy=+MOLE-RAG2026.0666.2
GIMLETType=Pre-training2026.0761.2
CoTPrompting Strategy=Chain-of-Thought2025.1259.58
GPT-4o-miniStrategy=+MOLE-RAG2026.0659.3
MolBasicShot=2-shot2026.0757.9
MoMuType=Pre-training2026.0757.6
MolBasicShot=4-shot2026.0757.3
DAPrompting Strategy=Direct Answer2025.1256.85
MolRAGShot=4-shot2026.0756.6
MolBasicShot=0-shot2026.0756.3
MolBasicShot=1-shot2026.0756.1
ChemDFM14BStrategy=Baseline2026.0655
MolRAGShot=2-shot2026.0755
Mistral7B-InstStrategy=+MOLE-RAG2026.0654.8
MolRAGShot=1-shot2026.0754.5
MolRAGShot=0-shot2026.0753.4
Mistral7B-InstStrategy=Baseline2026.0653.1
GPT-4oStrategy=Baseline2026.0652.7
Llama3.2-3BStrategy=+MOLE-RAG2026.0652.5
GPT-4o-miniStrategy=Baseline2026.0652.3
Llama3.2-3BStrategy=Baseline2026.0650
Qwen34B-InstStrategy=Baseline2026.0649.6
Galactica-1.3BType=Pre-training2026.0749.5
KVPLMType=Pre-training2026.0749.2