Contribution of genetically-determined blood pressure to ischemic and hemorrhagic stroke
An important proportion of the variation observed in blood pressure is related to common genetic variation encountered in the general population. We are interested in understanding to what extend this genetic component in blood pressure levels influences clinical outcomes, including stroke and coronary heart disease. We are also interested in understanding whether information on this genetic component can help us change the treatment that out patients receive.
Genetic analyses to understand the contribution of circulating lipid levels to cerebrovascular disease
Mounting evidence indicates that the role of circulating lipids levels may be different in hemorrhagic and ischemic stroke. Specifically, very low levels of some lipid fractions – while beneficial for most cardiovascular outcome – may lead to an increase in the risk of spontaneous intracerebral hemorrhage. While this effect is likely small and should probably not change the way lipid-lowering drugs are prescribed, this inverse association between lipid levels and risk of intracerebral hemorrhage points to a novel pathway that could be modulated by future therapeutic applications.
Genetically-determined cardiovascular disease and aging-related conditions
An important portion of our work is focused on testing the general hypothesis that genetically-determined cardiovascular disease – both vascular risk factors and clinical endpoints like stroke – influence cognitive and functional status. On this front, we work within the construct of Yale Claude D. Pepper Older Americans Independence Center, a nationally and internationally recognized authority in Aging research. The center’s mission is to provide intellectual leadership and innovation for aging research that is directed at enhancing the independence of older persons. The center’s unifying theme is the investigation of multifactorial geriatric conditions, encompassing single conditions resulting from multiple contributing factors or affecting multiple outcome domains and multiple conditions occurring simultaneously.
Combining neuroimaging analyses with population genetics
How much we can learn from genetic variation depends on the type, quality and amount of non-genetic information data can be connected to genomic data. In our field, this non-genetic data is called “phenotypic data,” and encompasses a wide array of physiological (height, blood pressure, lipid levels) and pathological (stroke, myocardial infarction and dementia) variables. Along these lines, we are interested in evaluating how genetic variation modifies the acute stages of human disease, with focus on stroke. One key component of this work pertains to finding and characterizing neuroimaging variables that help us quantify the amount of brain in injury during the acute phase of stroke.
Genetic underpinnings of spontaneous intracerebral hemorrhage
Together with several partners from the International Stroke Genetics Consortium to advance our identify genetic risk factors for spontaneous, non-traumatic intracerebral hemorrhage, the most frequent type of hemorrhagic stroke. Our team played a critical role in the completion of the first genome-wide association study of this condition, and is currently working on the second of these studies. This work is one of the main projects of the International Stroke Genetics Consortium Working Group on Intracerebral Hemorrhage.
Machine learning, deep learning and artificial intelligence
These three related approaches are rapidly changing the landscape of all research fields working with Big Data. We are interested in understanding whether these approaches can improve and accelerate the research described above. In particular, we are interested in developing methods based on ML/DL/AI that can help us (1) produce more and more accurate genetic and neuroimaging data, and (2) find patterns when combining these different types of data.