PROJECT SUMMARY/ABSTRACT In hospitalized patients, the fungus Candida albicans invades the bloodstream and causes hematogenously disseminated candidiasis (HDC). The mortality associated with this infection approaches 40%, even with antifungal current antifungal therapy. Because Candida spp. constitute the third most common cause of nosocomial blood stream infections, HDC is a serious public health challenge of increasing medical and socioeconomic importance. A better understanding of both the fungus and the host is required to develop new approaches to diagnose, prevent, and treat HDC. Characteristics of both the fungus and the host determine which patients develop HDC, and among those, which ones develop persistent candidemia and/or succumb to infection. However, these characteristics are incompletely understood. To address this fundamental gap in knowledge, we propose to analyze an existing biorepository of host and fungal samples selected from a cohort of 300 patients with candidemia. We will compare samples from 35 patients with candidemia who died with those from 35 matched candidemic patients who survived. We will also analyze serial isolates from selected patients with persistent candidemia. Using this biorepository and associated clinical information, we will link fungal genotype/phenotype, and host epigenetics, immune responses, and signaling pathways to patient outcome. Specifically, we will 1) determine C. albicans genomic, epigenetic, transcriptomic, and phenotypic signatures associated with lethality and persistence during candidemia and 2) determine the host response to lethal vs. non-lethal and persisting vs. resolving C. albicans clinical isolates, with a focus on signal transducer and activator of transcription (STAT) molecules. The data generated by the project will be compiled by the Bioinformatics and Data Monitoring (BDM) core and analyzed by the Computation and Predictive Modeling (CPM) core to identify both fungal and host signatures that are associated with persistent and lethal candidemia. These signatures will then be tested by mutational and gene silencing approaches to refine the models and generate new hypotheses that will be tested in an iterative fashion. The results of this comprehensive analysis hold promise to lead to new approaches to predict, prevent, and treat HDC.