Identifying the causal relationships of cognitive symptoms for neurodegenerative diseases Lead Investigator: Yang Ni Institution : Texas AM University E-Mail : yni@stat.tamu.edu Proposal ID : 1547 Proposal Description: The goal of this project to improve our mechanistic understanding of the causal relationships among cognitive symptoms, depression, and neuropsychiatric symptoms while adjusting for race, age, and education level. Upon the completion of the project, we expect to have inferred a Bayesian network that describes the causal relationships among cognitive symptoms, depression, and neuropsychiatric symptoms using the UDS longitudinal dataset. With a better understanding of causal relationships of cognitive symptoms (namely, which set of symptoms tend to emerge before others) can potentially assist in early detection of the Alzheimer???s disease. A major challenge in identifying the causal relationships of cognitive symptoms lies in the longitudinal nature of the dataset. Existing Bayesian networks are only applicable to cross-sectional data or time series data for which the measurements are obtained on the same regular time points. We will propose a novel longitudinal Bayesian network approach to model variables that are observed on irregular time points.