This vignette provides all reproducible codes for our article:

Saritha Kodikara, Susan Ellul, John Carlin and Kim-Anh Lê Cao

Abstract: The microbiome is the collection of all microorganisms co-existing interdependently within an ecosystem. Extremely complex and with interactions between host and environmental factors over time, the microbiome is inherently dynamic. Longitudinal studies are important to capture the microbiome temporal variation and gain mechanistic insights into a microbial systems, but current statistical methods are limited due to the complex and inherent features of the data. We have identified three objectives for longitudinal microbiome statistical analysis (1) to identify microorganisms with differential abundance over time and between sample groups, demographic factors or clinical variables of interest; (2) to identify microorganisms evolving in a similar manner across time; (3) to identify temporal relationships between microorganisms. This review explores the limitations of current methods - that often univariate and not suitable for compositional data, and highlights opportunities for further methodological developments.

Keywords: differential abundance, clustering, networks, compositionality, 16S, shotgun sequencing, relative abundance