Abstract

The microbiome is a complex ecosystem that plays a vital role in our overall health and well-being. To unravel microbial mysteries, it is important to understand the temporal variation of the microbes. Advancements in cost-effective sequencing technologies and simplified sample (faecal) collection methods have resulted in an exponential surge of longitudinal microbiome studies. These studies allow us to gain a deeper understanding of the microbiome than ever before. However, the data generated through these studies pose significant analytical challenges, hindering our ability to extract meaningful insights due to its many layers of complexities, including being compositional, high-dimensional, sparse, over-dispersed, and multivariate. This presentation will first give a brief overview of three main analytical objectives in longitudinal microbiome studies: (1) differential abundance over time and between sample groups, demographic factors or clinical variables of interest; (2) clustering of microorganisms evolving concomitantly across time and (3) network modelling to identify temporal relationships between microorganisms. Objective (3) is of particular interest as it allows us to quantify changes in microbial interactions in different sample groups of interest, but methods to solve this problem are lacking. In our new approach, we infer association networks between microbial taxa that change over time and between groups. By building several networks across time and groups, we aim to identify differences in the subgraphs or communities. These differences provide insights into the underlying biological mechanisms that drive the microbial response to different conditions. I will illustrate our new approach on both simulated and real data.