A metabolomic analytical approach permits identification of urinary biomarkers for Plasmodium falciparum infection: a case–control study

Table 1 Multivariate analysis and validation of OPLS-DA models of malaria patients and healthy controls

Description	LC-MS
Peak detected
ESI+	6278
ESI−	3466
ESI+ and ESI−	9744
Cross-validation
R²Y	0.993
Q²	0.583
Permutation test
Intercept^a	−0.268
External validation: classification (training/test models)
Sensitivity (%) (true positive rate (TPR) = TP/(TP + FP)	80%
Specificity (%) (true negative rate (TNR) = TN/(TN + FN)	77%
Accuracy (%) = (TP + TN)/(TP + FP + TN + FN)	78%
ROC curve (AUC)^b (TPR vs FPR)	0.83
Classification based on a selected set of biomarkers^c
Sensitivity (%)	91%
Specificity (%)	91%
Accuracy (%)	91%
ROC curve for individual set of metabolites	0.92

^aThe model is considered valid when the regression line of the permuted Q² values intercept at, or below zero
^bAUC: Area under receiver operating characteristic curve, which is the total area under the curve of sensitivity “true positive rate (TPR)” vs 1-specificity “false positive rate (FPR)”, ideal model gives AUC = 1
^cSelected set of biomarkers were 1,3 diacetyl propane, 2-octanedioic acid, N-prolyl histidine, taurine, N-acetylputrescine, N-acetylasparagine, N-acetylspermidine and N-acetylglutamine

ISSN: 1475-2875