Skip to main content
. 2022 Aug 25;22:707. doi: 10.1186/s12879-022-07694-8

Table 3.

Classification performance of binary classifications with selected feature combinations on test sets

Accuracy Sensitivity Specificity PPV NPV
RF
 SPPT/Ctrl (F1) 83.33% 80.00% 85.71% 80.00% 85.71%
 SNPT/Ctrl (F2) 92.86% 85.71% 100% 100% 87.50%
 SPPT/SNPT (F3) 83.33% 80.00% 85.71% 80.00% 85.71%
SVM
 SPPT/Ctrl (F1) 91.67% 80.00% 100% 100% 87.50%
 SNPT/Ctrl (F2) 92.86% 85.71% 100% 100% 87.50%
 SPPT/SNPT (F3) 91.67% 80.00% 100% 100% 87.50%
MLP
 SPPT/Ctrl (F1) 83.33% 60.00% 100% 100% 77.78%
 SNPT/Ctrl (F2) 92.86% 85.71% 100% 100% 87.50%
 SPPT/SNPT (F3) 91.67% 80.00% 100% 100% 87.50%

F1 set: albumin and 9-OxoODE; F2 set: l-Pyroglutamic acid, Enterostatin human and 9-OxoODE; F3 set: Val-Ser, Methoxyacetic acid and Ethyl 3-hydroxybutyrate