List your company
ResearchFormerly WizWand

Toxicity Prediction on Tox21 (scaffold)

0.807AUCToxicBlend
0.720680.743090.76550.78791Jun 12, 2018May 6, 2019Mar 29, 2020Feb 20, 2021Jan 14, 2022Dec 8, 2022Nov 2, 2023
Updated 8mo ago

Evaluation results

MethodLinks
ToxicBlendEnsemble=Proposed Blending2018.060.807
XgbFeatures=PLD2018.060.803
Pred. Avg.Ensemble=Simple Averaging2018.060.793
GINGraph-level pre-training=Supervised, Node-level pre-training=EdgePred2019.050.783
GINGraph-level pre-training=Supervised, Node-level pre-training=Context Prediction2019.050.781
GINGraph-level pre-training=Supervised, Node-level pre-training=Attribute Masking2019.050.779
GINGraph-level pre-training=Supervised, Node-level pre-training=Infomax2019.050.778
XgbFeatures=SNG2018.060.777
XgbFeatures=PCFP2018.060.773
NNFeatures=SNG2018.060.77
GINGraph-level pre-training=Supervised, Node-level pre-training=None2019.050.77
APTBase Model=Mole-BERT2023.110.77
Mole-BERTPre-training strategy=Mole-BERT2023.110.768
GINGraph-level pre-training=None, Node-level pre-training=Attribute Masking2019.050.767
APTBase Model=EdgePred2023.110.765
MAMPre-training strategy=MAM2023.110.762
NNFeatures=PCFP2018.060.761
GINGraph-level pre-training=None, Node-level pre-training=EdgePred2019.050.76
EdgePredPre-training strategy=EdgePred2023.110.76
MAMEncoder=vanilla VQ-VAE2023.110.758
GINGraph-level pre-training=None, Node-level pre-training=Context Prediction2019.050.757
GINGraph-level pre-training=None, Node-level pre-training=Infomax2019.050.753
GraphMAEPre-training strategy=GraphMAE2023.110.752
MGSSLPre-training strategy=MGSSL2023.110.752
NNFeatures=PLD2018.060.751
DCGCFeatures=DC2018.060.751
GraphCLPre-training strategy=GraphCL2023.110.751
AttrMaskPre-training strategy=AttrMask2023.110.751
GraphLoGPre-training strategy=GraphLoG2023.110.75
G-ContextualPre-training strategy=G-Contextual2023.110.75
GPT-GNNPre-training strategy=GPT-GNN2023.110.749
AD-GCLPre-training strategy=AD-GCL2023.110.749
GraphMVPPre-training strategy=GraphMVP2023.110.749
TMCLPre-training strategy=TMCL2023.110.749
JOAOPre-training strategy=JOAO2023.110.748
No pretrainPre-training strategy=No pre-train (scratch)2023.110.746
3D InfoMaxPre-training strategy=3D InfoMax2023.110.745
SimGRACEPre-training strategy=SimGRACE2023.110.744
TMCLMode=w/o Ltri2023.110.741
GINGraph-level pre-training=None, Node-level pre-training=None2019.050.74
ContextPredPre-training strategy=ContextPred2023.110.736
G-MotifPre-training strategy=G-Motif2023.110.736
TMCLMode=w/o Lcon2023.110.735
InfoGraphPre-training strategy=InfoGraph2023.110.733
DCNNRFeatures=DC2018.060.73
DCNNFeatures=DC2018.060.724