B1347
Title: Pathway analysis over brain structural network with a survival outcome
Authors: Yize Zhao - Yale University (United States) [presenting]
Xinyuan Tian - Yale University (United States)
Fan Li - Yale University (United States)
Denise Esserman - Yale University (United States)
Li Shen - University of Pennsylvania (United States)
Xiwen Zhao - Yale University (United States)
Abstract: Technological advancements in noninvasive imaging techniques provide an unprecedented opportunity to understand how the human nervous system is supported by molecular profiles and how it affects behaviors through the construction of whole brain interconnections on white matter fiber tracts, known as brain structural connectivity. Existing approaches to analyze structural connectivity frequently disaggregate the entire network into a vector of unique edges or summary measures, leading to a substantial loss of information. We propose an integrative Bayesian framework to model the effect pathway between each component and the potential mediating role of brain structural connectivity between genetic exposure and survival outcome. To accommodate the neurobiological architectures of connectivity, including symmetry, hollow and dense interconnections among hub nodes, we develop a structural modeling framework including a symmetric matrix-variate AFT model, and a symmetric matrix response regression to characterize the effect paths. We further impose within-graph sparsity and between-graph shrinkage to identify informative network configurations and eliminate the interference of noisy components. Extensive simulations confirm the superiority of our method compared with existing alternatives. We apply the proposed method to analyze the landmark ADNI study, and obtain neurobiologically plausible insights.
