A Consensus Clustering Model for Integrating Multi-Source Longitudinal Data

• Activity Number: 473 - Advances in Bayesian Methods and Mixture Modeling for Health Data
• Type: Topic Contributed
• Date/Time: Wednesday, August 10, 2022 : 2:00 PM to 3:50 PM
• Sponsor: SSC (Statistical Society of Canada)
• Author(s): Zihang Lu*
• Companies: Queen's University
• Keywords: Clustering; Data integration; Joint modeling; Mixture and latent variable modeling; Longitudinal data; Multi-source data

Abstract

In medical research, identifying clinically relevant patient subgroups is an important step toward personalized medicine. Integrative clustering is often used to identify patient subgroups by joint modeling datasets collected from multiple sources describing the same set of individuals. Simultaneously clustering multi-source data allows more information to be incorporated into the clustering process and could generate deeper biological insight regarding disease heterogeneity. Motivated by a birth cohort study, we propose a consensus clustering model for integrating multi-source longitudinal data with distinct data structures and variable types. An efficient Gibbs sampling algorithm is developed to estimate the model parameters and the number of clusters. Analysis of real and simulated data will be presented and discussed.