Integrating Metagenomics and Immunoinformatics to Prioritize Antigens and Immune-Modulating Molecules from Environmental Microbiomes

Environmental microbiomes - especially diverse soil and terrestrial communities - harbor an immense and largely untapped reservoir of microbial genetic diversity. Integrating metagenomic sequencing of these habitats with immunoinformatics offers a new avenue to discover candidate antigens and microbial metabolites that modulate immunity. Metagenomics characterizes the mixed microbial genomes present in soil, water, or microplastic-associated biofilms, enabling recovery of genes and proteins (often via assembly and binning into metagenome-assembled genomes). Immunoinformatics tools can then predict B- and T-cell epitopes or antigenicity of these metagenome-derived proteins, prioritizing those likely to engage host immune receptors. Pipelines such as reverse vaccinology frameworks (e.g., ReVac, VaxiJen, Vaxign) screen genomes using features like surface localization, epitope content, and conservation. New Artificial Intelligence/Machine Learning approaches further refine candidate ranking by integrating multiple predicted features, including MHC-binding profiles and antigenicity. Beyond classical pathogen antigens, environmental microbes can provide innate immune stimuli (e.g., flagellin, LPS) and novel secondary metabolites (e.g., rapamycin-like immunosuppressants), which can be predicted or discovered via metagenomics. We review current sequencing and immunoinformatic workflows - from gene calling to epitope prediction and Machine Learning-based ranking - applied to soil and related microbiomes. We highlight examples of known immunomodulators from soil microbes and discuss how Artificial Intelligence-driven design can accelerate mining of environmental microbiomes for biomedical targets. This integration of metagenomics and immunoinformatics promises a One Health perspective that explicitly links human, animal, and environmental health, expanding antigen discovery beyond gut- or pathogen-centric views, while highlighting computational strategies, limitations, and future directions of this approach.