September 23, 2019
17:30
Cliniques Universitaires Saint-Luc, Service de Cardiologie, Salle de réunion -2 route 589
Speaker: Julie Hussin, PhD
IVADO Assistant Professor
Université de Montréal
Abstract:
Building predictive models of complex human disease that are useful in a medical setting has proved extremely challenging. This is particularly true for cardiovascular diseases, for which extensive research has focused on the discovery by genome-wide association studies (GWAS) of thousands of inherited disease-associated SNPs. On one hand, these discoveries provided overwhelming evidence for the role of genetics in CVD predisposition. On the other hand, genetic data do not show the full picture, as these disorders occur through a complex web of interactions between molecular phenotypes (genes, metabolites, micro-organisms) and environmental factors. This quest for full molecular profiling has now led to an explosion of ‘omics’ technologies producing datasets at all biological levels, ultimately modulated by individuals’ genetic makeup. Therefore, our research activities are focused on combining genomics with multi-omics data to understand cardiovascular health, with applications for unraveling molecular processes behind myocardial infarction, pulmonary hypertension and heart failure, along with co-occurring conditions such as diabetes, auto-immune disorders and reproductive challenges. To reach these goals, my research group collaborates closely with clinicians and fundamental researchers at the Montreal Heart Institute to generate and analyze omics data according to the most pressing research questions in cardiovascular research. In this talk, I will present the research questions currently tackled by my group, all of which are relevant to precision cardiovascular medicine, along with the data available to us to advance this field. I will also describe the different computational challenges and the novel techniques employed and developed to answer these questions, which includes artificial intelligence methods developed in collaboration with the Montreal Institute for Learning Algorithms. These projects hold the potential to improve cardiovascular disease risk-estimating algorithms and predictive models of human health.