Project Summary Alzheimer’s disease (AD) is the most common form of dementia in the United States. Microstructural abnormalities in white matter (WM) are often reported in AD and are associated with neurodegeneration. Even though neuroimaging studies identified disruption of WM integrity in AD several key questions remain unanswered including which brain regions have the strongest WM changes in early stage of AD and what biological processes underlie WM abnormality during disease progression. To address these critical questions, we will systematically interrogate diffusion tensor imaging (DTI), genetic, gene expression and clinical information using Alzheimer’s Disease Neuroimaging Initiative (ADNI) data. Our goal in this proposal is to perform network and comprehensive genetic analysis using neuroimaging parameters and genomics data to better understand basic mechanisms and biological pathways underlying WM changes during disease progression. We hypothesize that key genetic drivers play significant role in loss of WM integrity and connectivity in AD-related brain regions. In our first aim we will identify significant DTI-features in regions relevant to AD-pathology related endophenotypes using cross-sectional and longitudinal data. Through correlating regional DTI features with clinical and cognitive traits such as CSF tau/p-tau and abeta levels, episodic memory scores and amyloidosis (PET), we will quantify regional white matter abnormality and identified most vulnerable brain regions. In the second aim we will perform co-diffusion- expression network analysis of the combined DTI and gene expression data. To accomplish this aim multiscale gene co-expression network analysis will be performed on the gene expression data to identify co-expressed gene modules that were further correlated with DTI features. Lastly, we will perform deep biological mining on the co-diffusion-expression networks to identify potential upstream regulators and downstream indicators for white matter microstructure and connectivity abnormality in AD. We will integrate DTI and multi–omics data to understand the underlying biological mechanisms through identification of common/rare and structural variants, epigenetic changes or any level of blood metabolite changes related to WM structure during disease progression. In summary, integrative framework in this proposal will allow us the identification of genetic risk factors and key biological pathways underlying disruption of WM changes in early stage of AD. The proposed research in this fellowship will provide me with advance training in computational genetics, complex data processing, and neuroimaging. The training plan will build on my understanding of systems biology of WM changes in AD.