Decoding the Microbiome-Metabolome Nexus: A Systematic Benchmark of Integrative Strategies

Abstract

Background: The exponential growth of high-throughput sequencing technologies was an incredible opportunity for researchers to combine different -omics within computational frameworks. In particular metagenomics and metabolomics data have gained an increasing interest due to their implication in many complex diseases. However, currently, no standard seems to emerge for jointly integrating both microbiome and metabolome datasets within statistical models. Results: Thus, in this paper we comprehensively benchmarked fifteen different integrative methods to link microorganisms and metabolites. Methods evaluated in this paper cover most of the researcher's goals such as global associations, data summarization, individual associations and feature selection. Through an extensive simulation study and an application to real gut microbial datasets, we highlighted the best approaches for unraveling complementary biological processes involved between the two omics. We provided general guidelines for practitioners depending on the scientific question and the data at-hand. Conclusion: In summary, we argue that this paper constitutes a promising avenue for establishing research standards when mutually analyzing metagenomics and metabolomics data, while providing foundations for future methodological developments.