Skip to main content

Table 4 Evaluation metrics for BBB and Caco-2 permeability prediction for test set with and without CL

From: Prediction of blood–brain barrier and Caco-2 permeability through the Enalos Cloud Platform: combining contrastive learning and atom-attention message passing neural networks

 

BBB Permeability

Caco-2 Permeability

Metrics

Without CL

With CL

Without CL

With CL

ROC-AUC

0.944

0.953

0.896

0.914

Accuracy

0.868

0.885

0.828

0.872

Precision

0.866

0.898

0.801

0.846

Sensitivity

0.939

0.926

0.929

0.949

Specificity

0.740

0.812

0.695

0.771