Microbiome systems encompass diverse ecological niches and host associations, with their scale and complexity challenging traditional analytical frameworks. Advances in artificial intelligence are transforming microbiome research by enabling improved integrative analyses of microbial genomes, community structure, and functional potential. In this Review, we outline how these developments create opportunities for microbiome research to move beyond descriptive analyses towards predictive modelling and hypothesis generation, including emerging insights into microbial function, host-microbe interactions, and ecosystem-level organisation. As the field grapples with challenges in model interpretability, generalisability, and causal inference, emerging strategies, such as multiomics and multicohort integration, provide promising avenues to deepen biological understanding. Addressing ethical considerations, including data privacy, algorithmic bias, and equitable access, will be crucial for translating artificial intelligence-driven microbiome discoveries into robust and inclusive clinical applications.
Yu et al. (Fri,) studied this question.