Skip to main content
. 2023 Jun 27;24(4):bbad240. doi: 10.1093/bib/bbad240

Figure 4.

Figure 4

Performances of AMP prediction. (A) The numbers of AMPs and non-AMPs in the training datasets with different sequence identity cut-off thresholds. (BC) AUCs and accuracy values of iAMPCN built on the training datasets with different sequence identity cut-off thresholds based on the 10-fold stratified cross-validation test. (D) ROC curves and the corresponding AUC values of various modes using the independent test dataset. (E) AUCs of iAMPCN built on the training datasets with different identity thresholds via the independent test dataset. (F) The accuracy values of identifying short AMPs (5–30 AAs) on the independent test dataset.