PROJECT ABSTRACT Tuberculosis (TB) drug resistance is a major threat to global health, with only about a third of the half million drug-resistant TB patients enrolled into treatment, highlighting a critical gap for diagnosing drug resistant TB. To address this challenge, investigators from Princeton, Rutgers, Fudan, and Harvard University have convened clinical, technologic, laboratory and epidemiologic expertise to develop and evaluate a new platform to address critical unmet TB drug susceptibility testing needs. To improve accuracy and comprehensiveness within rapid, simplified workflows for TB drug susceptibility testing, we will leverage a highly-multiplexed CRISPR platform called CARMEN. Our project benefits from optimization of CARMEN over the last 4 years for viral, bacterial, and single nucleotide polymorphism-based variant and drug resistance detection, machine-learning based design strategies, and ongoing advancements in simplified workflows. The project further benefits from a large bank of MDR-TB strains and sputum samples for assay optimization and validation and a team with complementary expertise in molecular TB drug resistance, technology development and validation, and clinical implementation. This proposal will test the hypothesis that a TB DST based on the CARMEN platform will meet or exceed minimal WHO performance and operational targets for high-priority first and second line drugs. Specific aims are to develop a comprehensive CRISPR-based TB drug susceptibility test to achieve accurate genotypic and phenotypic predictions for: (1) resistance to high-priority first-line drugs, (2) resistance to high-priority second line drugs, and 3) to further simplify the manual workflow, speed, and instrumentation to exceed operational criteria for use in peripheral laboratories.