Towards more standards for sample descriptions: The FAIR WISH project

Abstract

In many scientific disciplines, physical samples represent the origin of research results. They record unique events in history, support new hypotheses, and are often not reproducible. At the same time, samples are essential for reproducing and verifying research results and deriving new results by analysing existing samples with new methodology. Consequently, the inclusion of sample metadata in the digital data curation processes is an important step to provide the full provenance of research results. The largest challenge is the lack of standardisation and the large variety of sample types and individuals involved: Most samples are collected by individual researchers or small groups that may have internal agreements for sample descriptions, but these might only be used for one expedition or within a small community, and rarely reach beyond institutional boundaries. The International Generic Sample Number (IGSN, www.igsn.org) is a globally unique, resolving, and persistent identifier (PID) for physical samples with a dedicated metadata schema supporting discovery functionality in the internet. IGSNs allow data and publications to be linked directly to the samples from which they originate and provide contextual information about a particular sample on the internet. The aim of the project FAIR WISH (FAIR Workflows to establish IGSN for Samples in the Helmholtz Association), funded by the Helmholtz Metadata Collaboration (HMC) was to work towards more standardisation of rich sample descriptions. Project outcomes include (i) standardised, rich and discipline-specific IGSN metadata schemes for different physical sample types within the Earth and Environmental sciences (EaE), (ii) workflows to generate machine-readable IGSN metadata from different states of digitisation and (iii) the FAIR Samples Template. The FAIR SAMPLES Template enables metadata collection and batch upload of samples at various sample hierarchies (parent, children at different hierarchy levels) at once. The ability to fill the FAIR SAMPLES Template by individual researchers or research teams or to create scripts to fill it out directly from databases for a wide range of sample types makes the template flexible with a wide applicability. The structured metadata, captured with the FAIR SAMPLES Template and converted into XML files, already represents an important step for the standardisation of rich sample descriptions and their provision in machine-readable form. Standardised workflows for metadata documentation and compliance with international metadata standards address the challenges associated with reproducibility of samples and their insufficient documentation. The developments within the FAIR WISH project provide a foundation for a more collaborative and integrated scientific enterprise. Future efforts in this area can build on this framework to further improve the accessibility and interoperability of sample data and advance the collective understanding of Earth's environmental processes.