Skip to main content

Table 4 Performance comparison on the independent MUV dataset

From: Multi-PLI: interpretable multi‐task deep learning model for unifying protein–ligand interaction datasets

Metric

Vina

3D-CNN on DUD-E

3D-CNN on DUD-E/CASR

Single-task on DUD-E

Multi-task

AUC

0.549

0.522

0.499

0.453

0.443

  1. The AUC scores of Vina and 3D-CNN are derived from [10]
  2. Best values are higlhlighted in bold