Advanced computational analysis in metagenomic studies to support precision medicine.

[BACKGROUND] The human microbiome has been linked to host health and is suggested to play a direct role in the onset of certain human diseases, as well as in impacting treatment efficacy. Characterizing the microbiome composition and its interaction with the host is now supported by an established, continuously improving set of bioinformatic and statistical resources that enable reproducible answers to fundamental questions about microbiome sample composition and its association with sample and host information. Extensive evidence highlighted that, in a nondiseased state, the microbiome composition is determined by multiple factors, including the acquisition of microbes at birth, lifestyle, dietary patterns, social interactions, antibiotic use, or probiotic intake, among others. In disease states, the microbiome may alter its composition and, in some cases, present specific biomarkers, as in colorectal cancer. Some microbiome components have also been associated with improved immunotherapy response in clinical oncology, suggesting a potential beneficial role for certain species and supporting the use of the microbiome as an additional therapeutic tool in these scenarios.

[OBJECTIVES] This review summarizes computational approaches for microbiome characterization, highlights key findings on microbiome-disease associations, and provides a perspective on directions and open questions relevant to address in the future.

[SOURCES] We selected scientific studies and reviews, published in peer-reviewed journals, based on their impact in the field and relevance to the topic of this manuscript. Literature selection was conducted by reviewing scientific publications retrieved from major scientific databases, such as PubMed, and by combining with the authors' knowledge of the literature.

[CONTENT] Here we review computational approaches to characterize and model the microbiome's structure in health and disease and discuss multicohort data analysis, integration, and validation methods.

[IMPLICATIONS] Improved microbiome characterization supports precision medicine by informing prevention or treatment, leveraging refined microbiome signature and modulation strategies.