Skip to main content

Table 1 Performance results of distinguishing adaptor proteins with different methods

From: Classification of adaptor proteins using recurrent neural networks and PSSM profiles

Method Cross Validation Independent Test
  Sensitivity Specificity Accuracy AUC MCC Sensitivity Specificity Accuracy AUC MCC
k-NN 0.635 0.750 0.738 0.770 0.254 0.671 0.751 0.743 0.791 0.280
RF 0.185 0.968 0.890 0.837 0.214 0.290 0.923 0.860 0.838 0.216
SVM 0.397 0.934 0.881 0.818 0.332 0.426 0.932 0.881 0.806 0.353
CNN 0.532 0.875 0.841 0.774 0.328 0.548 0.873 0.841 0.783 0.339
RNN 0.812 0.751 0.757 0.853 0.373 0.856 0.798 0.804 0.893 0.446
  1. (k-NN: k=10; RF: num_stimators=500; SVM: c=8.0, g=0.5; CNN: 128 filters; RNN: 512 filters)