microbetag: simplifying microbial network interpretation through annotation, enrichment tests and metabolic complementarity analysis

Microbial co-occurrence network inference is often hindered by low accuracy and tool dependency. Building on the reverse-ecology paradigm, we applied data integration and metabolic modeling methods to address these challenges.

We introduce microbetag (https://hariszaf.github.io/microbetag/), a comprehensive software ecosystem designed to enhance network annotation. Nodes (taxa) are enriched with phenotypic traits, based on genome and literature knowledge, while edges represent two types of metabolic complementarities, highlighting potential cross-feeding relationships. Pathway complementarities are inferred from genome annotations, based on a donor species' ability to complete a missing step in a beneficiary's KEGG module. Seed complementarities, on the other hand, are derived from the metabolites a species requires exogenously, as determined by its metabolic model, and the donor's ability to produce these metabolites.

microbetag’s online version relies on microbetagDB, a database of 34,608 high-quality genomes with detailed annotations. It effectively identified known metabolic interactions on previously published data. A preprocessing module allows the analysis of large data sets, while a stand-alone version allows users to apply microbetag to custom reference genomes/bins/MAGs. Ultimately, MGG, a CytoscapeApp we developed, offers a streamlined, user-friendly interface for retrieving and visualizing microbetag-annotated networks.

Combined with network clustering and enrichment analysis, microbetag serves as a robust hypothesis-generating tool.