Skip to main content

Table 4 The performances of individual feature-based models on balanced Human dataset

From: A genetic algorithm-based weighted ensemble method for predicting transposon-derived piRNAs

Index Feature AUC ACC SN SP
F1 1-Spectrum Profile 0.754 0.690 0.731 0.649
F2 2-Spectrum Profile 0.841 0.756 0.780 0.732
F3 3-Spectrum Profile 0.839 0.750 0.747 0.754
F4 4-Spectrum Profile 0.829 0.740 0.732 0.748
F5 5-Spectrum Profile 0.802 0.718 0.681 0.755
F6 (3,1)-Mismatch Profile 0.862 0.772 0.819 0.725
F7 (4,1)-Mismatch Profile 0.854 0.761 0.788 0.734
F8 (5,1)-Mismatch Profile 0.842 0.750 0.754 0.747
F9 (3,1)-Subsequence Profile 0.850 0.767 0.809 0.725
F10 (4,1)-Subsequence Profile 0.866 0.782 0.821 0.743
F11 (5,1)-Subsequence Profile 0.875 0.791 0.829 0.754
F12 1-RevcKmer 0.746 0.699 0.889 0.509
F13 2-RevcKmer 0.803 0.724 0.774 0.673
F14 3-RevcKmer 0.818 0.732 0.765 0.698
F15 4-RevcKmer 0.808 0.718 0.717 0.718
F16 5-RevcKmer 0.791 0.702 0.658 0.746
F17 PCPseDNC 0.836 0.757 0.776 0.738
F18 PCPseTNC 0.849 0.765 0.787 0.742
F19 SCPseDNC 0.833 0.754 0.770 0.739
F20 SCPseTNC 0.832 0.751 0.777 0.725
F21 Sparse Profile 0.904 0.819 0.815 0.824
F22 PSSM 0.880 0.807 0.815 0.799
F23 LSSTE 0.688 0.631 0.664 0.598