Methods for high-dimensional data in HIV/CVD research

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): Recent technological advances have yielded vast quantities of molecular and cellular data, affording unprecedented opportunities to understand complex disease etiologies and to inform clinical management strategies. However, in order to derive information from these rich stores of data we need to develop sound and appropriate analytic techniques. This need is especially relevant in studies at the intersection of human immunodeficiency virus (HIV) and cardiovascular disease (CVD), which are characterized by an elaborate set of interactions among viral and host factors. These factors include viral and host genetic profiles, as well as markers of caloric metabolism, immune activation and inflammation, which work together to determine response to therapy and overall disease progression. A comprehensive assessment of these markers presents several analytical challenges owing to the large number of potentially informative variables and the largely uncharacterized relationship among them. We propose a multi-faceted strategy that focuses on the development and application of integrative statistical approaches. Such approaches will allow us to explore and characterize novel hypotheses relating to the complex relationships among multiple genetic, environmental, demographic, and clinical factors and measures of disease progression. Specifically, this continuation application focuses on advancing and applying statistical methods in two settings: first, we consider population-based genetic association studies of innate-immunity, adipokine, drug metabolism and drug transport genes and markers of immune reconstitution, inflammation and risk of CVD in HIV-infected individuals; and second, we address investigations of metabolic and immunologic profiles that associate with immune recovery, inflammation and risk of CVD. The Specific Aims of the proposed research are to develop and evaluate: (1) Latent class and mixture modeling paradigms for (a) discovering and characterizing multi-locus genotype-trait associations and (b) evaluating unobservable haplotype-trait associations in candidate-gene investigations; and (2) Hierarchical mixture models and machine learning approaches for (a) monitoring quantitative biomarkers in resource-limited settings and (b) characterizing high- dimensional predictors of immune reconstitution and inflammation. IMPACT: This research will lead to the creation of appropriate and carefully evaluated analytic tools to derive information from the rich array of molecular and cellular data now available. Ultimately, this research will advance our ability to translate molecular and cellular level data for clinical decision making, serving at the cornerstone of personalized medicine. PUBLIC HEALTH RELEVANCE: The newly available array of data on genetic polymorphisms and cellular level immune factors promises unprecedented opportunities to elucidate complex disease etiology and inform clinical management strategies. Using human immunodeficiency virus (HIV) and cardiovascular disease (CVD) as our model systems, we propose to develop, evaluate and apply new analytic approaches for high-dimensional data. Ultimately, these methods will allow us to derive information from the vast quantities of molecular and cellular data for personalized, clinical decisions and thus serve as a central component of translational medicine.