Open Microbiome Data: A FAIR² Collection for AI-Ready and Reusable Data

Scientific knowledge is fundamentally built on data; yet, for too long, microbiome datasets have remained siloed in fragmented repositories, poorly annotated, and inconsistently formatted. As microbiome science becomes increasingly data-intensive, with analyses driven by machine learning applied to metagenomic, metatranscriptomic, and metabolomic profiles, merely depositing raw sequences in archives is no longer sufficient. To enable responsible and trustworthy reuse, data must be well-structured, machine-actionable, ethically aligned, and rich in contextual metadata. Addressing this challenge, Frontiers has introduced a new article type: the FAIR² Data Article, a citable, peer-reviewed publication format that elevates datasets to become first-class research outputs.A FAIR² Data Article presents a dataset as the primary focus of publication, separate from traditional hypothesis-driven narratives. It allows detailed documentation of how the data was generated, processed, and structured, providing the transparency and provenance needed to evaluate its quality and suitability for reuse. Each article is accompanied by a FAIR² Data Portal that includes interactive visualizations, machine-actionable metadata, an executable Jupyter notebook, and a podcast-style audio summary to enhance accessibility. An embedded AI assistant guides readers through the dataset, helping researchers engage with the data in a meaningful way.This new standard builds on the FAIR principles — Findable, Accessible, Interoperable, and Reusable — by adding three key dimensions: AI-readiness, ethical alignment, and contextual richness. These additions ensure that datasets are not only usable by machines, but also responsibly managed and fully interpretable by diverse scientific audiences. As outlined in the FAIR² press release, this approach reflects Frontiers’ continued commitment to open science, reproducibility, and scientific integrity in the age of AI.Across the microbiome field, researchers face persistent challenges in preparing datasets for reuse — ranging from inconsistent adherence to metadata standards such as MIxS and MIMARKS, missing sample and processing context, to unclear data licensing and incomplete pipeline provenance. These barriers hinder meta-analyses and cross-cohort comparisons, slow scientific progress, and limit the translational and clinical impact of microbiome research. By offering a structured, peer-reviewed venue for publishing high-quality datasets, FAIR² Data Articles aim to unlock the full value of microbiome data and make it more accessible, reusable, and trustworthy.To mark the launch of this new article type in Frontiers in Microbiomes, this Research Topic welcomes both:• FAIR² Data Articles that showcase high-quality, reusable, AI-ready datasets in the fields of host-microbiome interactions and microbiome omics and bioinformatics, and• Scholarly perspectives on the development, implementation, and impact of FAIR data in microbiome research and the role of the new FAIR² Data Article.We encourage researchers, data scientists, bioinformaticians, clinicians, and open science advocates to contribute to this multidisciplinary collection. Whether you are submitting a FAIR² Data Article or offering a critical perspective on the FAIR² framework and its role in advancing microbiome science, your work will help shape the future of open and reproducible microbiome research. Topic editors: Robert G Beiko, Susana Fuentes Source: https://www.frontiersin.org/research-topics/82892/open-microbiome-data-a-fair2-collection-for-ai-ready-and-reusable-data