New Horizons in Chemistry: From Data Creation to Data Re-use in NFDI4Chem

Chemistry is rapidly moving toward digitisation. Within Germany's National Research Data Infrastructure (NFDI), the NFDI4Chem consortium aims to digitise key chemical research processes to facilitate the collection, storage, processing, analysis, and publication of research data. Over the past five years, NFDI4Chem has established robust foundations for research data management (RDM) in chemistry by promoting the FAIR (Findable, Accessible, Interoperable, and Reusable) principles, and building a federation of integrated services and infrastructure. The vision of NFDI4Chem is to develop seamless data workflows from data creation to data re-use. Central to data creation is the "Smart Lab," which offers methods, concepts, and software solutions for efficient data capture, documentation, and publication. Its core component, the Electronic Laboratory Notebook (ELN), serves as an essential entry point for researchers, integrating seamlessly with other NFDI4Chem tools directly in the lab. Using the Smart Lab, chemists can load data from instruments into the ELN for analysis and evaluation [1], [2]. Structured and annotated data can subsequently be uploaded into NFDI4Chem's federated repositories, which are recommended by more than 30 chemistry journals. Alternatively, researchers can directly archive and publish data in these repositories, enabling data sharing and reuse by other scientists. Both the Smart Lab and repository federation are built upon a foundation of terminologies, metadata, and standards. Chemists, RDM specialists, and computer scientists collaborate to create Minimum Information for Chemical Investigations (MIChI), which standardise data and metadata across diverse chemical resources. MIChI encompasses guidelines, checklists, data models, and implementation strategies, ultimately reflected in metadata schemas tailored specifically for chemistry [3]. NFDI4Chem services continually incorporate and refine these standards, enhancing service federation iteratively. To generate machine-actionable chemical data, including information about chemical entities, reactions, and experimental conditions, NFDI4Chem employs ontologies. These formally describe the knowledge about chemical entities, their characteristics, reactions, and relevant metadata. NFDI4Chem has curated a comprehensive collection of 39 chemistry-specific ontologies [4], some developed by the consortium, to facilitate semantic annotation, linking, and detailed research data description. The Ontologies4Chem Workshop series brings international ontology creators, users, and developers together, standardising ontology development and curation workflows [5]. The NFDI4Chem Terminology Service (TS) enables searching, accessing, and curating of these ontologies for both human users and software applications like ELNs and repositories. Semantically annotated Data are harvested from repositories and indexed by the NFDI4Chem Search Service, which currently provides access to approximately 140,000 datasets. Beyond infrastructure, NFDI4Chem supports the chemistry community through extensive training and resources. The Knowledge Base offers over 70 articles and videos addressing all facets of chemical research data management. Training courses have reached more than 1,000 chemists, and a dedicated Helpdesk has resolved around 600 community inquiries. Furthermore, NFDI4Chem actively fosters cultural change in the research community by addressing error culture and author guidelines for data publication through initiatives such as the Editors4Chem workshops [6]. In summary, NFDI4Chem delivers comprehensive pipelines for applying standards from initial data creation to eventual reuse. Building upon semantics it is now at the verge of creating fully compliant semantic, machine-actionable data for future reuse by researchers, machine learning algorithms or AI tools.