Deep integration of multi-omics and patient health data to clarify risk factors of enterococcal infective endocarditis - Project Summary Enterococcus faecalis exhibits remarkable hardiness, intrinsic resistance, and a proclivity to exchange resistance driving mortality rates as high as 30% during systemic infections. Progression to infective endocarditis (IE) caused by adherence and proliferation of bacterial cells on heart tissues presents a poor prognosis with few treatment options. The relative contributions of clonal lineages, bacterial genetic factors, or host risk-factors for IE development in humans are still unclear. Animal model and in vitro studies have identified presumed virulence factors associated with distinct stages of IE disease. Also, studies of patient risk factors for IE have suggested the presence of a prosthetic heart valve or having a community acquired infection are important. However, no study has tracked and assessed all these bacterial and host factors concurrently, which is critical for making optimal healthcare decisions. The proposed work intends to fill this gap by integrating bacterial genomics, proteomics, and patient health data from a comprehensive collection of 1,189 enterococcal bacteremia cases from an urban region of Canada. Such a unique dataset collected over 15 years will both validate and reveal bacterial genetic factors important for human IE infections. It will also clarify the importance of microbial genetics in the context of patient risk factors. Aim 1 will curate and assess a range of patient health and demographic variables from all cases of enterococcal bacteremia to detect patient- level characteristics associating with IE. Additionally, these characteristics will be mapped to phylogenetic relationships to link patient factors with clonal spread of bacterial lineages. In Aim 2, genetic variants associated with IE will be detected via a microbial genome wide association study. Using a linear mixed model incorporating corrections for population structure and covariates, a comprehensive assessment of significant variants will be accomplished. Finally, in Aim 3, the expression profiles of ~1,300 proteins per isolate will be used to compare IE vs. non-IE cases to both validate the importance of previously identified virulence factors and find new associations unobtainable from genomics alone. Following this, the isolates with the highest and lowest IE associations will be studied in blood conditions, facilitating a more representative view of protein expression during IE. Successful completion of this work will provide critical links between bacterial variation and IE, which will be applicable to improved diagnostic and treatment solutions. The proposed work will be done at the highly collaborative and innovative Broad Institute under the guidance of Dr. Ashlee Earl and Dr. Michael Gilmore (Harvard Medical School) and in close partnership with clinical and data scientist collaborators. Under a clear training plan, the postdoctoral fellow is perfectly placed to network with superb researchers, publish high-impact publications, and build professional skills. The fellow will also have access to extensive computational and molecular resources and talented personnel that will improve their technical skills.
