This application addresses the Broad Challenge Area 05 (Comparative Effectiveness Research) and specific Challenge Topic 05-AG-103 (Imaging and Fluid Biomarkers for Early Diagnosis and Progression of Aging- related Diseases and Conditions including Neurodegenerative Diseases). In collaboration with the Memory and Aging Center at the University of California San Francisco (UCSF-MAC), we have been collecting - as part of ongoing studies - DNA, RNA, and gene expression data from peripheral blood in a large cohort of patients with degenerative dementia and controls. This series is extremely well characterized, with a large body of clinical, neuropsychological, and imaging data. Our approach is based on the hypothesis that the genetic component associated with degenerative dementia is reflected in peripheral tissues, such as peripheral blood, and can be captured by gene expression studies to be used as an intermediate phenotype for genetic association studies. Our group and others at UCLA are developing novel data analysis methods (iWGCNA), which provide a powerful framework for integrating genetic marker data, gene expression data, and complex phenotypes. Responding to the Challenge Areas and specific Challenge Topics identified by the NIH we propose to: 1) obtain genotyping data on a large series of patients with degenerative dementia and controls, where gene expression and longitudinal imaging and neuropsychological data are already available, and develop a biomarker set and classifier;2) to build a classifier based on imaging data collected on the same subjects;3) to compare the genetic/genomic and the imaging classifiers, and build a composite classifier and assess whether performance using the two different sources of information together is better than either alone. We expect that a composite, multi-modal biomarker set will perform better than either method alone, and will be the first multi-modal classifier in dementia, based solely on these biomarkers. The overarching goal of this work is to identify surrogate biomarkers in order to build a diagnostic classifier that would be invaluable in early and accurate diagnosis, as well as for sample stratification for clinical trials. This project leverages our collaboration with the UCSF-MAC by adding a new dimension to the phenotypic and genetic data obtained via other long-term funding mechanisms in one of the largest and best-characterized AD/FTLD series in the world. Since most of the phenotype, gene expression, and imaging data are already available, only a relatively marginal effort and cost is required to extend the genetic characterization of this cohort with genotyping data, providing a unique opportunity and niche for this Challenge Grant mechanism. The work proposed here is ready to begin, and will result in immediate job creation and retention consistent with the goals of the Recovery Act.
Identification of biomarkers enriching diagnostic, prognostic, and therapeutic capabilities is an important goal in dementia. We propose to build a molecular classifier based on peripheral blood samples and imaging data from demented patients. This would be a valuable tool for biomarker identification, improved patients classification, therapy evaluation, and to further our understanding of disease pathophysiology.
Showing the most recent 10 out of 17 publications
