Skip to main content

Table 1 The detailed performance for the four protein families on both Dataset1 and Dataset2 by 10-fold cross validation

From: Predicting drug-target interactions from drug structure and protein sequence using novel convolutional neural networks

Type   Acc Sen Pre F1 AUC
Enzymes Dataset1 0.943 0.927 0.903 0.915 0.985
  Dataset2 0.920 0.881 0.880 0.881 0.973
Ion channels Dataset1 0.919 0.894 0.867 0.881 0.970
  Dataset2 0.900 0.948 0.792 0.863 0.949
GPCRs Dataset1 0.884 0.818 0.831 0.824 0.945
  Dataset2 0.920 0.899 0.866 0.882 0.968
Nuclear receptors Dataset1 0.884 0.872 0.798 0.833 0.936
  Dataset2 0.907 0.891 0.841 0.865 0.966