This paper provides a systematic analysis of the emerging legal profiles in the artificial intelligence ecosystem, structured along three interdependent conceptual axes. Firstly, it examines the multi-level regulatory framework taking shape in the European Union. The EU Regulation 2024/1689 is critically explored in its risk-based approach, with particular attention to the categories of high-risk AI systems. The synergies and tensions with the legal framework governing data circulation in the Union are analyzed, with a profound influence in terms of compliance on data-driven technologies. This includes the EU Regulation 2016/679, which directly addresses crucial issues such as automated decisions and profiling; the EU Regulation 2022/868 on data altruism mechanisms and the reuse of public data; the EU Regulation 2023/2854, which introduces rules related to accessing data generated, among others, by IoT and industrial devices, as well as the EU Regulation 2025/327. Secondly, the complex issue of AI liability is evaluated, in the dual dimension of and safety. Particular attention will be paid to the now dismissed EU regulatory proposals. A specific focus will deal with the possible applicability of the product liability discipline. Finally, several sectoral criticalities are identified through case studies in healthcare and transport domains, evaluating for each the difficult balance between potential benefits and risks from algorithmic biases and systemic discrimination. The methodology combines dogmatic analysis, legal comparison, and concrete case studies, contributing to the debate on harmonization between technological innovation, protection of fundamental rights and sustainable legal governance models in the AI era.

The European Nucleotide Archive (ENA; https://www.ebi.ac.uk/ena), hosted at the European Molecular Biology Laboratory's European Bioinformatics Institute (EMBL-EBI), remains a global, open-access platform for the submission, archiving, dissemination, and reuse of nucleotide sequence data. In 2025, ENA continues to advance its mission of fostering FAIR (findable, accessible, interoperable, reusable) data principles through innovations in interoperability, scalability, and global engagement, providing infrastructure for a rapidly growing volume of data across diverse domains. This article highlights the key developments in 2025, including the progress of the technical transformation, enhanced support for large-scale biodiversity projects, and the implementation of the International Nucleotide Sequence Database Collaboration Global Participation Initiative. We also discuss infrastructure enhancements to handle exponential data growth and improve user experiences and data discovery.

The respiratory microbiome plays an important role in maintaining human health. Despite the rapid growth of literature and publicly accessible data on the respiratory microbiome, a large-scale, well-curated database is still lacking. Here, we introduced ResMicroDb, a comprehensive database and analysis platform for the human respiratory microbiome. ResMicroDb contains 106 464 samples from 514 projects, spanning 10 sample sites, 72 sample types, and 146 phenotypes. Notably, it includes ~7-fold more respiratory samples than existing multi-body-site resources. To improve the reusability and accessibility of data, a standardized bioinformatics pipeline was employed to generate taxonomic profiles, and 32 metadata fields were manually curated. ResMicroDb also provides 11 908 microbe-disease associations, identified from 132 case-control studies, to deepen the understanding of microbiome-disease relationships. Additionally, ResMicroDb offers three tools for in-depth analysis: "Microbiome Composition" for visualizing taxonomic profiles of user-selected samples; "Sample Similarity Search" for inferring the characteristics of new samples by comparing them to the database based on similarity; and "Cross-study Analysis" for identifying common and specific microbial characteristics across cohorts, phenotypes and sample sites. ResMicroDb serves as a versatile and valuable resource for advancing a broad spectrum of respiratory microbiome research and clinical relevance. ResMicroDb is freely accessed at https://resmicrodb.cncb.ac.cn.

Plant-pollinator interactions play a pivotal role in ecosystem functioning and sustainable agriculture. However, plant-polinator datasets are scattered across various networks, in country-specific initiatives, and stored in isolated silos, making them difficult to access by scientists and decision-makers. By promoting the adoption of Findable, Acessible, Interoperable, and Reusable (FAIR) data standards (Wilkinson et al. 2016) across multiple initiatives worldwide, we are working to transform the fragmented nature of these datasets and make data on plant-pollinator interactions widely available. As the biodiversity community advances towards FAIR data, machine-actionable metadata has emerged as a critical enabler for scalable data assessment, discovery, and reuse. However, while FAIR principles emphasize machine-readability, many datasets are still evaluated manually or lack structured metadata entirely, limiting their integration into global platforms. This study shares practical insights from the WorldFAIR Agricultural Biodiversity Case Study, in which we operationalized machine-actionable FAIR metadata for the review of plant-pollinator interaction datasets (Drucker et al. 2024). We developed a semi-automated workflow to assist in evaluating datasets against the FAIR principles using tools from the Global Biotic Interactions initiative (GloBI, Poelen et al. 2014). The GloBI bots "Nomer" and "Elton" can read structured metadata from standard vocabularies such as Darwin Core (DwC), Ecological Metadata Language (EML), and the Plant-Pollinator Interactions (PPI) vocabulary. Nomer focuses on taxonomic alignment with several taxonomic catalogues, such as GBIF Backbone and Catalogue of Life. Elton extracts species interactions from datasets of various structures and formats, including DwC-Archives. By relying on machine-readable metadata, the bots were able to flag inconsistencies, suggest improvements, and generate repeatable reports across pilot projects in Argentina, Brazil, the African continent, Kenya, Colombia, East Africa, Central Asia, and the USA (for example, Elton et al. 2025). This helped researchers assess dataset interoperability without needing full access to the raw data, a crucial feature given legal and institutional access constraints. To make the data review report readable for researchers, GloBI's bots produce a document resembling a data publication, complete with a title, authors, publication date, abstract, introduction, and other relevant sections. Our results underscore the transformative role of machine-actionable metadata in biodiversity data governance. Automating FAIR assessments not only increases transparency and repeatability but also accelerates the integration of datasets into platforms like GloBI. While human expertise remains essential, tools like Nomer and Elton demonstrate that FAIR assessment can evolve beyond checklists to become dynamic, scalable, and integrated into the data lifecycle. Resources and code are openly available in the online repositories Zenodo*3 and GitHub*2, and are summarized in the GloBI platform*1. To help alleviate the burden of manually reviewing data as part of scientific publication review, we propose deploying domain-specific, automated data review processes that enable researchers to better understand how to make their data easier to review and reuse. Recognizing that publishing reusable, integrated data remains mostly a manual process, we recommend that plant-pollinator and species interaction datasets be registered with one or more infrastructures (e.g., GloBI, GBIF) to benefit from the domain-specific data review services they offer. We suggest that data publishers continue to collaborate on building, maintaining, and improving similar infrastructures to assess and increase the quality and FAIRness of published scientific data. Through the adoption of standards such as Ecological Metadata Language, Darwin Core, Plant-Pollinator Interactions Vocabulary, and Relation Ontology, we aim to enhance the understanding of how plant-pollinator interactions contribute to sustaining life on Earth while ensuring that data is easily findable, accessible, interoperable, and reusable for further research and analysis (FAIR).

Creating findable, accessible, interoperable, and reusable (FAIR) data and metadata is essential for researchers. Effective Research Data Management (RDM) is crucial for achieving this, as recognized by funding organizations. Both data FAIRness and RDM rely on well-structured and documented data, including organized storage in a clear folder structure. Establishing a suitable folder structure early on is beneficial, yet junior scientists often struggle due to uncertainty about their needs. Ready-made templates can help establish an initial structure and foster good RDM habits. While templates exist for small-scale projects, a comprehensive folder structure tailored to doctoral research is lacking. This paper presents a folder structure template (FST) designed for PhD candidates in life and natural sciences, encompassing locations for code, results, figures, manuscripts, background information, and administrative paperwork. The FST offers a practical solution for digital data organization, along with best practice RDM recommendations and metadata recording templates. By implementing a thoughtful structure early on and maintaining it consistently, researchers can manage their data more efficiently, leading to improved FAIR data outcomes and faster publication.

Reusing clinical trial data to consolidate and advance medical knowledge.

The growing complexity and heterogeneity of research data in materials science and engineering (MSE) demand structured and interoperable solutions for effective data management and reuse. To address this challenge, this article introduces the National Research Data Infrastructure (NFDI)‐MatWerk Ontology (MWO), as a semantic foundation to standardize metadata, link distributed datasets, and support digital research data management (RDM) in MSE. MWO addresses the need for the structured, standardized, and semantically rich representation of key entities, processes, and resources involved in the generation, sharing, and reuse of MSE research data. Aligned with the Basic Formal Ontology (BFO) , MWO develops as a modular extension of the NFDIcore ontology, and reuses Platform MaterialDigital core ontology (PMDco) MWO offers broad semantic coverage, modeling elements such as researchers, organizations, projects, software, workflows, datasets, metadata schemas, instruments, events, and services. It supports modular, scalable development through ontology design patterns (ODPs) and is maintained via the Ontology Development Kit (ODK) following best practices from the Open Biomedical Ontologies (OBO) Foundry. MWO also serves as the foundation for the MSE Knowledge Graph (MS‐KG), which integrates semantically interlinked research data across the NFDI‐MatWerk consortium and wider MSE community.

Abstract As the availability of historical biodiversity data continues to grow, ensuring its usability through adherence to FAIR principles (Findable, Accessible, Interoperable, and Reusable) has become increasingly essential. This study focuses on solving key challenges in interpreting biodiversity data from historical texts, particularly in identifying and aligning common species names with their modern scientific counterparts. We address five main challenges: spelling variations, the invention of new terms, semantic shifts between broad and narrow naming conventions, and the renaming or reclassification of historical terms. To tackle these issues, we tested a range of large language models (LLMs) (GPT‑4, LLaMA3-405B, Mistral-8B, and Qwen3-30B-A3B) for their ability to resolve these challenges and support terminology alignment. The initial entity detection was performed using GPT-4o, which achieved a 92% success rate in detecting historical common names and correctly identified 98% of scientific terms on a test dataset. Comparative evaluation of the ability to match historical common names with modern equivalents revealed that GPT-4o consistently delivered the most accurate and nuanced outputs across four of the five challenges, demonstrating strong contextual understanding. The results highlight the potential of advanced LLMs to not only identify entities but also to interpret historical naming conventions, thereby enhancing the reusability and interoperability of biodiversity data in line with FAIR principles.

Non-human eukaryotic genome assemblies and annotations are frequently dispersed across repositories, supplementary materials, or remain unpublished. Their access often requires technical expertise and large downloads, hindering usability. The Swedish Reference Genome Portal*1 (SRGP) offers a web platform to easily find, explore, and share genomic data from non-human eukaryotic species studied in Sweden. The SRGP gathers assemblies and annotations into species-specific pages with an interactive JBrowse2 genome browser, metadata, and download links (Fig. 1). This web interface provides user-friendly visualisations aimed at broad user communities, particularly in biodiversity and evolution. Unlike global resources such as Ensembl or the UCSC Genome Browser, which focus on model organisms, the SRGP showcases species studied in Swedish institutions. It fosters community-driven data sharing and broadens access to genomic annotations that are rarely published (e.g., from population genomics), while adhering to Findable, Accessible, Interoperable and Reusable (FAIR) principles. In doing so, it bridges global infrastructures with local research needs, enabling both specialists and non-specialists to make better use of genomic data. The technical architecture of the SRGP (Fig. 2) is designed to make the service easy to use and reuse by others: Simple and secure : Built with Hugo, a static site generator that requires no databases or heavy servers, enabling lightweight maintenance. Lightweight and fast : Composed of prebuilt pages and files for quick loading and easy scalability. Open and reusable : Fully open-source, allowing others to adapt and extend the code. Easy to update : New species or datasets can be added by editing configuration files and rebuilding the site. Portable : Packaged with Docker for easy deployment anywhere, including Kubernetes clusters. User-friendly : Integrates JBrowse2 for interactive exploration of genomic data. Simple and secure : Built with Hugo, a static site generator that requires no databases or heavy servers, enabling lightweight maintenance. Lightweight and fast : Composed of prebuilt pages and files for quick loading and easy scalability. Open and reusable : Fully open-source, allowing others to adapt and extend the code. Easy to update : New species or datasets can be added by editing configuration files and rebuilding the site. Portable : Packaged with Docker for easy deployment anywhere, including Kubernetes clusters. User-friendly : Integrates JBrowse2 for interactive exploration of genomic data. The SRGP is designed for simple deployment and maintenance. Genome data and annotations are processed with automated scripts in Docker containers to ensure consistency and reproducibility. The website is built with Hugo, which generates fast, lightweight pages requiring no database. Both data and pages are bundled into Docker images and deployed on a Kubernetes cluster, with ArgoCD handling updates automatically. This setup provides reliable operation on standard infrastructure, easy scaling for new datasets, and secure, low-maintenance access to genomic resources (Fig. 2). In summary, the SRGP makes previously hidden genomic data, especially from local or specialised communities, easy to find and explore through a single, user-friendly browser. It lowers access barriers while remaining simple, fast, and scalable. By unifying scattered resources, it provides a reusable model for other countries, institutions, or research groups. This presentation is available in (presentación disponible en español en) Fuentes‐Pardo et al. (2025).

GMrepo (Gut Microbiome Data Repository) is a curated and consistently annotated database of human gut metagenomes, designed to improve data reusability and enable cross-project and cross-disease comparisons. In this latest release, GMrepo v3 has been expanded to 890 projects and 118 965 runs/samples, including 87 048 16S rRNA and 31 917 metagenomic datasets. The number of annotated diseases has increased from 133 to 302, allowing more comprehensive disease-related microbiome analyses. We systematically identified microbial markers between phenotype pairs (e.g. healthy versus diseased) at the project level and compared them across datasets to detect reproducible signatures. As of this release, GMrepo v3 includes 1299 marker taxa (726 species and 573 genera) associated with 167 phenotype pairs, derived from 275 carefully curated projects. To assess marker stability, we developed the Marker Consistency Index (MCI), which summarizes the prevalence and directional consistency of markers across studies. Among 400 markers showing altered abundances in ≥10 projects, 143 were consistently enriched in healthy controls (MCI > 75%), while 85 were enriched in diseases (MCI < 25%). A marker-centric interface enables users to explore marker behavior across diseases. The GMrepo v3 database is freely accessible at https://gmrepo.humangut.info.

INTRODUCTIONUnderstanding and maximizing complex health data is crucial for accelerating discovery, translational research, funding priorities, and improving data management. Rapid, cost‐effective progress can be made by repurposing datasets. This work explores the dementia cohort landscape, identifies cohorts relevant to dementia translation, and highlights areas to strengthen health cohort infrastructure.METHODPubMed was searched for publications utilizing dementia‐related cohorts (1970–2024), supplemented by international dementia data platforms. A template aligned with the C‐Surv data model was used to summarize administrative details and the presence of measurements across 17 themes.RESULTSFrom 4596 publications and 11 data platforms, 883 cohorts were identified (558 population and 325 clinical). Of these, 74% indicated data availability for future research, though metadata reporting varied. Cohort metadata are accessible via the landscape tool.DISCUSSIONThis work reveals extensive global dementia‐related data for repurposing and identifies priority areas for improvement, including metadata transparency, data accessibility, and locations to prioritize for future research.HighlightsA total of 883 cohorts were identified globally (1970 to 2024): 558 population and 325 clinicalThe Global South is substantially underrepresentedSeventy‐four percent of cohorts offer data access, but protocols and metadata quality vary widelyOnly 45% of cohorts were discoverable via existing data platformsThe online landscape tool enables strategic discovery and reuse of dementia data

Insects represent the most diverse animal group and play essential roles in ecosystems, agriculture, and human health. The rapid accumulation of high-quality genomes and diverse omics datasets provides unprecedented opportunities to advance research in insect biology and evolution, yet also poses challenges in data integration, accessibility, and reuse. To address these demands, we developed InsectBase 3.0 (www.insect-genome.com), an upgraded platform that integrates extensive insect multi-omics data and provides user-friendly tools for furthor analysis. InsectBase 3.0 covers 3020 species across 24 insect orders, including 1651 chromosome-level assemblies and 61 353 curated transcriptomes. In addition, the new version incorporates a large-scale protein structure dataset (474 300 predicted models and 381 experimental entries), genome resequencing and variant resources, curated transposable element libraries, and 3D morphological reconstructions. To organize these heterogeneous resources, InsectBase 3.0 introduces knowledge graphs centered on species and genes, enabling users to navigate cross-dataset relationships more intuitively. In addition, the web framework has been modernized to improve efficiency, interactivity, and visualization. Together, these updates establish InsectBase 3.0 as the most comprehensive open-access insect omics platform to date, providing a valuable resource for evolutionary, functional, and applied research in insects.

The exponential growth of neuroscientific data necessitates platforms for data management and multidisciplinary collaboration. In this paper, we introduce Pennsieve, an open-source, cloud-based scientific data management platform that supports findable, accessible, interoperable, and reusable (FAIR) data sharing. It has integrated tools for data visualization, processing, and peer-reviewed data publishing that promote collaborative research and high-quality datasets optimized for downstream analysis, both in the cloud and on-premises. Pennsieve welcomes data submissions from individual investigators and small labs through entire consortia. It already serves more than 80 research groups worldwide and forms the core for several large-scale, interinstitutional projects and major government neuroscience research programs. Pennsieve stores over 125 TB of scientific data, with 35 TB of data publicly available in more than 350 high-impact datasets. By facilitating scientific data management, discovery, and analysis, Pennsieve fosters a robust and collaborative research ecosystem for neuroscience and beyond.

Abstract. Palaeoflood records are natural evidence of past flood events (typically found in landforms, sediments, or vegetation). Over the last 25 years, several palaeoflood record databases have been implemented. However, many of these databases are outdated, lack accessible or comprehensive palaeohydrological information, and present challenges in terms of data accessibility and reuse, particularly for non-research communities (e.g. planning administrations or flood risk managers). This work introduces PaleoRiada, the first open database that compiles published palaeoflood records from Spain. PaleoRiada stores typological, hydrological, temporal, and spatial data collected from approximately 126 publications (including journal articles, scientific reports, and book chapters). This database has been implemented using a simple Relational Database Management System (RDMS), integrated into a web platform, and is freely accessible at https://doi.org/10.5281/zenodo.17391823 (Sandoval-Rincón et al., 2025). The PaleoRiada database contains 299 palaeoflood records (both geological and biological) dated between 2014 CE and 97 000 BP and distributed across both Atlantic (164) and Mediterranean (135) catchments. PaleoRiada includes 157 records with specific discharge values ranging from 0.02 to 320 m3km-2s-1. These records are associated with a variety of river systems, including wide alluvial plains (25), Mediterranean ephemeral streams (17), mountain torrents (36), and confined valley rivers (79). Additionally, they encompass overbank flood events (102), flash floods (48), dam failures (1), and hyperconcentrated flow events (6). The relationship between PaleoRiada and the Spanish Flood-prone Mapping Project (SNCZI) indicates that approximately 80 % of the PaleoRiada records are not included in the flood-prone areas defined by SNCZI. Therefore, several records can be consulted to prioritise or propose new areas for preliminary flood risk assessment. Accessibility and simplified data query and entry in PaleoRiada can facilitate the application of palaeoflood data in land planning and flood risk management.

Understanding the status and trends of Essential Ocean Variables (EOV) and Essential Biodiversity Variables (EBV) is crucial for informing policy-makers and the public about sustainable management of marine biodiversity. Marine image data hold significant potential in this context, offering a permanent, information-rich and non-extractive record of marine environments at the time of capture. Quantitative image-based measurements such as species abundance and distribution have proven to be highly effective for engaging diverse stakeholders. The exchange and reuse by experienced fisheries scientists and marine ecologists of nine image-based datasets (including images, metadata and annotations) collected through various protocols revealed a substantial disconnect between initial expectations and actual practical usability, particularly in terms of understanding and reusing data. Two key issues were highlighted. First, the link between the datasets and their potential applications in deriving EOV/EBV was often inadequately described or absent. Second, despite both initial and ongoing efforts to document the data, new users continued to face challenges in understanding underlying properties and contextual features of datasets. We suggest these findings are likely to characterize many, if not most, historical image-based datasets. While standards promoting the FAIR principles (Findable, Accessible, Interoperable, and Reusable) for image-based data are emerging, our focus here is on the specific features of documentation that enable or facilitate reuse of data for the purpose of deriving EOV/EBV. From this perspective, we provide a set of recommendations for documenting both images and their associated annotations, aimed at supporting broader applications of in situ image data in marine conservation and ecology.

Reverse BSA-QTLseq: a new genotype-driven bioinformatics approach for simultaneous trait mapping.

Abstract. Marine glycerol dialkyl glycerol tetraethers (GDGTs) are used in various proxies (such as TEX86) to reconstruct past ocean temperatures. Over 20 years of improvements in GDGT sample processing, analytical techniques, data interpretation and our understanding of proxy functioning have led to the collective development of a set of best practices in all these areas. Further, the importance of Open Science in research has increased the emphasis on the systematic documentation of data generation, reporting and archiving processes for optimal reusability of data. In this paper, we provide protocols and best practices for obtaining, interpreting and presenting GDGT data (with a focus on marine GDGTs), from sampling to data archiving. The purpose of this paper is to optimize inter-laboratory comparability of GDGT data, and to ensure published data follows modern open access principles.

Introduction: The use of publicly available datasets and large-scale registries, such as the Society of Thoracic Surgeons (STS) National Databases and the Medicare database, has revolutionized accessibility to large sample data across research centers. However, there are few existing data governance protocols for managing the risk of Type I and Type II across the expanding portfolio of studies requesting from the same registry. Research Questions We investigate how repeated and uncoordinated reuse of datasets increases the riskiness of Type I/II errors due to dependent risks, undermining the reliability of research findings. We also examine strategies to manage this risk by differentiating between actively managed and passively managed databases. Methods: We adopt a decision-theoretic perspective to analyze how reuse of datasets can result in a dependence structure between tests that increases the disutility of the portfolio of Type I/II errors as measured by the actuarial notion of stop loss order. Results: Figure 1 shows the distribution of Type I errors for a two-sample t-test comparing the means of seven treatment groups against a common control group with data reuse of the control (Design 1) vs. without reuse of the control (Design 2) in 10,000 simulations of the global null. While the FWER of Design 1 was lower than that of Design 2, the error distribution of Design 1 is strictly preferable in stop loss order. We further demonstrate how subsampling strategies and portfolio optimization techniques can be deployed in a variety of contexts to mitigate the effects of data reuse. Conclusion: We are the first to propose a novel quantitative framework for reducing false positives and false negatives across multiple requests of the same database, providing a foundation for database managers to implement error control policies. Existing measures of error control, such as per-comparison error rates, false discovery rates, and familywise error rates fail to address the complex error structures that arise from multiple, overlapping studies. As reliance on large clinical datasets grows, especially those with high usage, robust error management strategies are crucial. Active dataset management offers a way to maintain the validity of conclusions from large registries. We advocate for increased funding of small, well-powered studies and the development of guidelines and software for dataset managers to effectively allocate inferential resources.

Monolithic 3D (M3D) ICs represent a significant advancement in VLSI technology, and unfortunately introduce a new dimension to the traditional, already NP-complete, 2D placement problem. A promising approach is to convert 2D placement into 3D placement using tier partitioning algorithms. Existing tier partitioning algorithms do not leverage historical data, such as prior placement configurations and performance metrics, that could improve efficiency and quality of placement. Utilizing historical data can help refine decision-making processes, reduce computational overhead, and enhance the overall effectiveness of tier partitioning strategies. To allow reuse of historical data, we propose a novel approach for M3D placement based on models and Reinforcement Learning (RL) that we name Dynamic Hierarchical Cluster Assignment with RL (DHCARL), which simplifies the search process by forming dynamic hierarchical clusters to preserve 2D placement when assigning clusters to 3D tiers. This approach uses a small set of parameters that enable effective reuse of historical data. With a Machine Learning (ML) model, low-code AutoML libraries help evaluate and select an RL policy. We propose novel Placement-aware graph networks (PAGN) to capture design connectivity and placement patterns in placement representation. We tested DHCARL on a diverse set of benchmarks, ranging from multiplexers (8- to 128-bit MuxShifter) to large-scale designs, including PicoRV32A, TATE-Pairing, and JPEG Encoder, implemented in 2-tiered and 4-tiered Monolithic 3D technology. Experimental results demonstrate that the proposed Reinforcement Learning-based Dynamic Hierarchical Assignment (DHCARL) approach outperforms state-of-the-art methods like TP-GNN and ART-3D, achieving a 6.3% lower placement cost on TATE-Pairing and 5.9% on PicoRV32A with similar runtime as TP-GNN.

Definitions to data flow: Operationalizing MIABIS in HL7 FHIR.