Figure 2:

Panels A, B, C, and D display the comparative performance of IPEV and HTP (KNN, SVC, LR, and QDA) with 5-fold cross-validation across groups A, B, C, and D, respectively. Sn = TP/(TP + FN), Sp = TN/(TN + FP), ACC = (TP + TN)/(TP + TN + FP + FN), Precision = TP/(TP + FP), F1-score = 2 × Precision × Recall/(Precision + Recall), where TP, TN, FP, and FN respectively represent true positive, true negative, false positive, and false negative. As the method with the best performance in HTP, KNN is selected for comparison. The mean and standard deviation of 5-fold cross-validation are computed to elaborate on performance evaluation. Due to a lack of reconstruction between the train and validation sets, the performance of HTP (KNN) is overestimated. (In this article, prokaryotic viruses are treated as positive samples.)