Globally, there are an estimated 10 million cases of tuberculosis (TB) each year, resulting in 1.6 million deaths. TB is the leading cause of death from a single infectious agent worldwide. TB is highly infectious and The Global Plan to Stop TB states that “… without new medicines, diagnostics and effective vaccines, we will not achieve the steep reductions in incidence and mortality that we need…” Of the 1.6 million deaths, 300,000 were HIV-infected individuals, a population where current diagnostics often fail. TB diagnostics utilized in developing countries, where TB is most prevalent, depend on clinical screening algorithms and sputum microscopy which are limited by low sensitivity and specificity. TB detection remains a significant diagnostic challenge. Current diagnostic methods are most often based on detection of biomarkers in a sputum sample which is difficult to obtain in children and HIV-infected individuals. The much more easily obtained urine sample has been suggested as an alternative patient sample for diagnosis. A urine lateral flow test for TB is available based on the TB surface biomarker LAM, but it has low sensitivity. PCR-based approaches, which potentially are much more sensitive and specific, have been proposed for detecting nucleic acid biomarkers in urine but remain controversial. A particular challenge for TB DNA biomarker testing in urine is DNA degradation into fragments small enough to pass through the kidney into the urine. The characteristics of fragmentation in urine presents three main challenges: 1) fragments are present in very low concentrations, 2) fragments present are too short for extraction by available methods, and 3) fragments of biomarkers are too short for amplification by traditional PCR methods. We describe innovative methods to overcome these challenges to 1) achieve efficient extraction and concentration of fragmented IS6110 from large volumes of urine using high gradient magnetic separation, and 2) a method of achieving amplicon reconstruction from short IS6110 fragment to make full-length IS6110 amplicons and enable PCR detection. Limited, but promising, preliminary data suggest this approach can be applied to existing PCR reactions for the TB biomarker IS6110. Aim 1 studies are proposed to characterize the fragment extraction and concentration approach. Aim 2 examines how the amplicon reconstruction approach performs and how this method can be extended to identify drug resistance. This proposal aims to develop these two technologies before testing these designs in prospective patient samples in a subsequent R01 with our collaborators in South Africa.