Smooth Time-Varying Gaussian Graphical Models to Study Disease Progression

• Activity Number: 354 - Multivariate Analysis and Graphical Models
• Type: Contributed
• Date/Time: Wednesday, August 5, 2020 : 10:00 AM to 2:00 PM
• Sponsor: Section on Statistical Learning and Data Science
• Author(s): Erin Mcdonnell* and Shanghong Xie and Karen Marder and Yuanjia Wang
• Companies: Columbia University and Columbia University and Columbia University and Columbia University
• Keywords: Gaussian graphical models; Graphical lasso; Time-varying networks; Huntington's disease; Brain imaging; Precision medicine

Abstract

We seek to understand how the complex interrelationships between clinical symptoms and brain imaging biomarkers change over time leading up to diagnosis of a disease in subjects with a known risk of disease. We propose time-dependent Gaussian graphical models that ensure temporal and structural smoothness across time-specific networks to examine the trajectories of interactions between markers aligned at the disease onset. Specifically, we anchor subjects relative to the time of disease diagnosis and estimate networks at each time point of interest using all available data, applying kernel weights to borrow information across observations that are close to the time of interest. An additional penalty term in the form of a Laplacian matrix is introduced to inform the current graph structure with the structure at the previous time. We conduct extensive simulation studies to compare our approach with the regular graphical lasso. We then apply our method to data from PREDICT-HD, a large prospective observational study of pre-manifest Huntington's disease (HD) patients, to identify symptom and imaging network changes that precede clinical diagnosis of HD.