PROJECT SUMMARY All pathogens possess genetic diversity that can impact clinically relevant phenotypes such as virulence, susceptibility to drugs, and vaccine efficacy. In the post-genomic era, our ability to catalog microbial genotypes has far outstripped our capacity to profile microbial phenotypes. This limits our ability to build genotype- phenotype maps for traits of interest and hinders the development of broadly effective new antimicrobials, vaccines, and public health interventions. To address this challenge, I developed a molecular barcoding approach that permits parallel fitness phenotyping of hundreds of bacterial clinical isolates in a single in vitro or in vivo experiment. I developed and validated this novel approach in the pathogen Mycobacterium tuberculosis and uncovered strain-specific differences in bacterial fitness during infection and following vaccination in the mouse model. Here, I propose to use this novel tool to interrogate the genetic basis of phenotypic heterogeneity in a related mycobacterial pathogen, Mycobacterium avium (MAC). MAC is an environmental microbe that can cause chronic and treatment-recalcitrant infections and is increasing in incidence. A major challenge in the management of MAC disease is the variability in disease course and treatment outcome, and the bacterial determinants of this variability are unknown. Here, I will leverage my strain barcoding approach and the natural biodiversity of this microbe to elucidate genetic determinants and molecular mechanisms of MAC pathogenicity and antibiotic response. These efforts will inform the development of improved diagnostics and therapeutics for this, and other, chronic bacterial infections. More broadly, this work will provide an intellectual framework and experimental toolkit to uncover the biological basis of heterogeneity in infectious disease phenotypes.