PROJECT SUMMARY The vast majority of child tuberculosis (TB) deaths occur in children <5 years old, highlighting the importance of identifying young children at high risk of developing TB and initiating preventive treatment. However, current tests for Mycobacterium tuberculosis (Mtb) infection have poor predictive value for TB progression. To make progress toward biomarker-targeted TB preventive therapy in young children, there is an urgent need to identify novel host markers that reflect the unique pathogenesis of childhood TB, can be detected earlier in the course of disease progression and can be more easily translated to a point-of-care assay. The overall objective of the proposed project is to identify biomarker signatures among young children that meet the World Health Organization (WHO)-recommended minimum accuracy targets for a test of TB progression. We hypothesize that a subset of host biomarkers of childhood TB disease that reflect unsuccessful immune control of Mtb infection will have the best performance for early prediction of TB disease progression in young children. To examine our hypothesis, we propose to leverage 1) biorepositories that provide access to banked samples from children with presumptive TB and healthy children with and without TB exposure and Mtb infection (Uganda, South Africa, the Gambia); 2) biorepositories that provide serial samples from young children followed for 12 months or more for incident TB disease (Uganda, South Africa, Vietnam); and 3) state-of-the-art platforms and bioinformatic pipelines for multi-omics biomarker discovery. In Aim 1, we will measure and compare biomarker levels (host cell-free RNA, proteins, metabolites and antibodies to Mtb antigens) in symptomatic children with presumptive TB and healthy children to identify candidate host biomarkers that differentiate childhood TB disease, differentiate Mtb infection, overlap between TB disease and Mtb infection, and segregate Mtb infection into multiple sub-groups. In Aim 2, we will use pathway analysis, in vitro human models and in vivo mouse models to prioritize candidate host biomarkers that are functionally linked to immune control of Mtb infection. In Aim 3, we will derive biosignatures consisting of the prioritized candidate host biomarker(s) and evaluate their accuracy for predicting TB progression overall and among children living with HIV in independent training and test sets. Completion of these aims will result in identification of promising biosignatures that can be further validated in large-scale field studies and translated into point-of-care tests for predicting progression of childhood TB.