Microbiome science needs AI competence, not just AI literacy.

An AI-competence framework for microbiome science is proposed, emphasizing five capabilities and “biological closure” to ensure predictive performance maps to meaningful biology.

• Why it matters: AI literacy and responsible tool use may not prevent biologically misleading conclusions caused by issues like data leakage, missing metadata, ecological confounding, poor model choice, and unverified generative-AI outputs.
• What they did: The authors outline five connected competence areas—data/platform fluency, machine-learning competence, advanced model competence, large-language-model and agent competence, and biological closure—then recommend assessing competence via reproducible artifacts.
• The result: Shared datasets, computational platforms, and benchmark challenges are suggested to broaden access and establish common standards for evaluating AI competence in microbiome research.

AI summary of the abstract · may contain errors · not medical advice · read the paper.