Table 3 Performance evaluation on classification datasets from TDC

From: MolGraph-xLSTM as a graph-based dual-level xLSTM framework for enhanced molecular representation and interpretability

 

HIA

Pgp

Bioavailability

CYP2D6-I

CYP3A4-I

CYP2C9-I

hERG

AMES

DILI

 

AUROC

AUPRC

AUROC

AUPRC

AUROC

AUPRC

AUROC

AUPRC

AUROC

AUPRC

AUROC

AUPRC

AUROC

AUPRC

AUROC

AUPRC

AUROC

AUPRC

FPGNN

0.958 ± 0.012

0.985 ± 0.006

0.930 ± 0.007

0.936 ± 0.008

0.666 ± 0.035

0.875 ± 0.012

0.863 ± 0.005

0.641 ± 0.020

0.869 ± 0.004

0.840 ± 0.007

0.867 ± 0.002

0.720 ± 0.005

0.846 ± 0.029

0.929 ± 0.021

0.836 ± 0.006

0.878 ± 0.009

0.899 ± 0.017

0.896 ± 0.025

DeeperGCN

0.965 ± 0.017

0.990 ± 0.005

0.884 ± 0.009

0.900 ± 0.011

0.579 ± 0.113

0.811 ± 0.070

0.850 ± 0.005

0.613 ± 0.009

0.883 ± 0.004

0.852 ± 0.005

0.871 ± 0.002

0.741 ± 0.010

0.734 ± 0.050

0.869 ± 0.037

0.818 ± 0.010

0.861 ± 0.010

0.864 ± 0.026

0.865 ± 0.018

DMPNN

0.976 ± 0.004

0.993 ± 0.001

0.889 ± 0.005

0.904 ± 0.008

0.617 ± 0.050

0.849 ± 0.022

0.872 ± 0.004

0.644 ± 0.010

0.887 ± 0.003

0.858 ± 0.004

0.878 ± 0.004

0.750 ± 0.006

0.741 ± 0.024

0.857 ± 0.025

0.838 ± 0.012

0.875 ± 0.013

0.883 ± 0.021

0.872 ± 0.032

HiGNN

0.974 ± 0.007

0.991 ± 0.003

0.882 ± 0.022

0.885 ± 0.022

0.620 ± 0.070

0.850 ± 0.042

0.867 ± 0.007

0.633 ± 0.013

0.886 ± 0.006

0.853 ± 0.009

0.878 ± 0.009

0.741 ± 0.020

0.807 ± 0.027

0.918 ± 0.010

0.822 ± 0.014

0.856 ± 0.018

0.892 ± 0.011

0.870 ± 0.028

TransFoxMol

0.951 ± 0.036

0.983 ± 0.016

0.875 ± 0.011

0.890 ± 0.013

0.619 ± 0.019

0.840 ± 0.023

0.859 ± 0.004

0.603 ± 0.014

0.854 ± 0.009

0.827 ± 0.011

0.867 ± 0.004

0.721 ± 0.013

0.799 ± 0.017

0.907 ± 0.010

0.816 ± 0.011

0.860 ± 0.009

0.896 ± 0.024

0.896 ± 0.013

BiLSTM

0.893 ± 0.021

0.957 ± 0.012

0.890 ± 0.033

0.881 ± 0.035

0.605 ± 0.059

0.824 ± 0.021

0.834 ± 0.005

0.592 ± 0.012

0.829 ± 0.008

0.779 ± 0.014

0.852 ± 0.004

0.698 ± 0.009

0.843 ± 0.022

0.933 ± 0.014

0.747 ± 0.015

0.790 ± 0.015

0.826 ± 0.030

0.829 ± 0.020

AutoML

0.880 ± 0.012

0.933 ± 0.006

0.821 ± 0.007

0.769 ± 0.011

0.509 ± 0.026

0.761 ± 0.010

0.762 ± 0.011

0.444 ± 0.015

0.773 ± 0.004

0.652 ± 0.007

0.776 ± 0.010

0.558 ± 0.002

0.682 ± 0.021

0.814 ± 0.010

0.756 ± 0.008

0.760 ± 0.007

0.766 ± 0.017

0.710 ± 0.019

MolGraph-xLSTM (Ours)

0.977 ± 0.010

0.992 ± 0.003

0.908 ± 0.011

0.919 ± 0.009

0.684 ± 0.118

0.872 ± 0.057

0.872 ± 0.002

0.659 ± 0.007

0.890 ± 0.005

0.863 ± 0.007

0.881 ± 0.006

0.757 ± 0.006

0.846 ± 0.017

0.933 ± 0.006

0.832 ± 0.011

0.870 ± 0.010

0.904 ± 0.006

0.888 ± 0.006

  1. Best results are shown in bold.