Skip to main content

Table 2 Predictive performance of sequence features for the four benchmark dataset

From: Prediction of HIV-1 protease cleavage site using a combination of sequence, structural, and physicochemical features

Features DT LR ANN
Acc.(%) AUC Acc.(%) AUC Acc.(%) AUC
746 Dataset
 AAC 83.7 0.897 86.4 0.938 81.0 0.935
 DipC 75.6 0.793 86.4 0.865 91.9 0.974
 PseAAC 78.3 0.787 86.4 0.938 81.0 0.885
 Seq_All 78.3 0.831 86.4 0.847 91.9 0.979 *
1625 Dataset
 AAC 91.4 0.908 84.1 0.904 91.4 0.952
 DipC 92.6 0.861 96.3 0.972 98.7 0.987
 PseAAC 90.2 0.822 87.8 0.921 87.8 0.945
 Seq_All 92.6 0.882 96.3 0.958 98.7 0.984
Schilling Dataset
 AAC 87.7 0.664 86.5 0.856 88.9 0.858
 DipC 87.7 0.526 87.1 0.806 89.5 0.790
 PseAAC 87.1 0.500 86.5 0.864 88.3 0.858
 Seq_All 87.7 0.611 87.7 0.802 87.1 0.821
Impens Dataset
 AAC 85.1 0.500 80.8 0.857 89.3 0.886
 DipC 85.1 0.500 82.9 0.579 93.6 0.893
 PseAAC 87.2 0.721 78.7 0.814 87.2 0.868
 Seq_All 87.2 0.802 85.1 0.696 89.3 0.875
  1. *The best accuracy and AUC in each dataset are underlined