PROJECT SUMMARY/ ABSTRACT Many specifics of the pathological process of Alzheimer’s disease (AD) remain unknown, such as the precise, functional relationship between tau accumulation and cognitive decline as a function of age, as well as other biomarkers that may modify these relationships. Conventional statistical approaches cannot easily answer questions about the relationship between tau and cognition, due to their dynamic relationship, unknown time lags, and complex measurement error structures. Mathematical modeling techniques—commonly used in infectious disease epidemiology and computational biology—are specialized for the study of complex relationships between biological variables, while incorporating prior knowledge about the relevant physiologic system. The proposed project leverages my quantitative expertise from dissertation research on infectious disease, using data from across the age span of AD onset to elucidate the relationship between tau-PET measures and cognition. As more tau-targeting drugs move through the pipeline, it is important to determine the optimal timing and duration of treatment for trial design and for post-approval clinical guidelines. The ideal timing for tau-targeting therapies may depend on factors such as age, amyloid, or vascular burden. Existing and emerging blood- based biomarkers may offer important information about how tau spreads in the brain and the timing of subsequent atrophy and cognitive decline longitudinally. A growing number of studies now perform tau-PET, and including repeated neuroimaging, making it possible for an improved understanding of the dynamics of tau and cognition in relation to other biomarkers. We propose a biologically motivated, mathematical modeling approach to understand how neuroimaging and other biomarkers can be used to better understand Alzheimer’s disease biology. We plan to fit mechanistic models to data from three cohorts across the age span of AD diagnosis: Alzheimer’s Disease Neuroimagin