High-throughput sequencing technologies are widely used to study microbial ecology across species and habitats in order to understand the impacts of microbial communities on host health, metabolism, and the environment. Due to the dynamic nature of microbial communities, longitudinal microbiome analyses play an essential role in these types of investigations. Key questions in microbiome studies aim at identifying specific microbial taxa, enterotypes, genes, or metabolites associated with specific outcomes, as well as potential factors that influence microbial communities. However, the characteristics of microbiome data, such as sparsity and skewedness, combined with the nature of data collection, reflected often as uneven sampling or missing data, make commonly employed statistical approaches to handle repeated measures in longitudinal studies inadequate. Therefore, many researchers have begun to investigate methods that could improve incorporating these features when studying clinical, host, metabolic, or environmental associations with longitudinal microbiome data. In addition to the inferential aspect, it is also becoming apparent that visualization of high dimensional data in a way which is both intelligible and comprehensive is another difficult challenge that microbiome researchers face. Visualization is crucial in both the analysis and understanding of metagenomic data. Researchers must create clear graphic representations that give biological insight without being overly complicated. Thus, this Research Topic seeks to both review and provide novels approaches that are being developed to integrate microbiome data and complex metadata into meaningful mathematical, statistical and computational models. We believe this topic is fundamental to understanding the importance of microbial communities and provides a useful reference for other investigators approaching the field.