Abstract
INTRODUCTION: Repeated measures microbiome studies, including longitudinal and clustered designs, offer valuable insights into the dynamics of microbial communities and their associations with various health outcomes. However, visualizing such multivariate data poses significant challenges, particularly in distinguishing meaningful biological patterns from noise introduced by covariates and the complexities of repeated measures. METHODS: In this study, we propose a framework to enhance the visualization of repeated measures microbiome data using Principal Coordinates Analysis (PCoA) adjusted for covariates through linear mixed models (LMM). Our method adjusts for confounding variables and accounts for the repeated measures structure of the data, enabling clearer identification of microbial community variations across time points or clusters. RESULTS: We demonstrate the utility of our approach through simulated scenarios and real datasets, showing that it effectively mitigates the influence of nuisance covariates and highlights key axes of microbiome variation. DISCUSSION: This refined visualization technique provides a robust tool for researchers to explore and understand microbial community dynamics in repeated measures microbiome studies.