Accuracy of salivary multi-marker models in the diagnosis of TB disease.

Receiver operator characteristics (ROC) curve showing the accuracy of the optimal five-marker biosignature (IL-1β, IL-23, ECM1, HCC1, fibrinogen) in the diagnosis of TB disease in all study participants, regardless of HIV infection status (A), frequency of analytes in top general discriminant analysis (GDA) models that most accurately classified individuals as TB patients or ORD irrespective of HIV status (B), ROC curve showing the accuracy of the most accurate eight-marker biosignature (granzyme A+PAI-1+GDF-15+SAA+IL-21+ENA-78+IL-12(p40) and IL-13) in the diagnosis of TB disease after exclusion of HIV positive individuals (C), and frequency of analytes in the top GDA models that most accurately classified study participants as TB or ORD in the absence of HIV infection (D). The bar graphs (B and D) indicate the frequency of analytes in the most accurate GDA models.