Abstract:
|
New biomedical technologies allow us to cost-effectively collect enormous quantities of diverse data. These data hold unprecedented potential to illuminate the complex workings of biological systems as well as causes and potential treatments for diseases. However, integrating diverse types of big-biomedical data such as neuroimaging, genomics, transcriptomics, and epigenomics measured on the same cohort of subjects poses a major statistical challenge. We discuss a number of innovations for exploratory analysis of diverse data via network analysis and supervised marker selection. Our methods are illustrated via case studies on the Cancer Genome Atlas and the Religious Orders Study/Rush Memory and Aging Project where statistical data integration yields new markers for cancer and Alzheimer's disease, respectively.
|