Increasing FAIRness in agrosystem sciences and plant phenomics
Through the collaborative efforts of the agrosystem and plant sciences communities research data are already available through various data repositories and infrastructures. To foster these developments and increase the value for the communities, enabling FAIR-compliance for scientific datasets is one top priority strategic aim. Due to the heterogeneity of the sub-domains and their requirements, we want to address three challenges with direct relation to specific FAIR principles:
a) To harmonize heterogeneous metadata requirements of agrosystem research infrastructures and to increase interoperability and findability of datasets, the German National Research Data Infrastructure (NFDI) consortium FAIRagro, develops an extension of the common schema.org vocabulary for the agrosystem domain, comparable to the Bioschemas extension, in close cooperation with scientific partners. Hackathon participants will evaluate the current state of the extension, assess whether it is sufficient to describe their own agrosystems data and create crosswalks to other metadata standards such as the ISA model, building cross NFDI bridges to DataPLANT and the ARC as a FAIR Digital Object layer. Based on this informal assessment, new agrosystem types and properties will be proposed for the extension. For these additional types and properties, as well as for the originally proposed ones, participants will link to suitable domain ontologies, which terms will be implemented into the extension for supporting users when annotating their datasets.
b) To ensure that results data of plant phenotyping experiments are compliant with the FAIR recommendations, an appropriate metadata annotation of the experimental results is essential. The community accepted metadata schema for this domain is the MIAPPE standard and the ISA format as corresponding data structure. However, FAIR-compliant annotation of plant phenotyping experiments is currently slowed down by several factors. The manual creation of ISA- and MIAPPE-compliant metadata is time-consuming and error-prone due to the heterogeneous nature of the datasets. Furthermore, scientists performing plant phenotyping experiments are often not experts in the usage of metadata schemas and related ontology concepts. Existing software libraries for programmatic creation of ISA- and MIAPPE-compliant metadata cannot be meaningfully used by most biologists without profound programming skills and in-depth knowledge of ISA and MIAPPE. To overcome these challenges, we will develop an intuitive step-by-step “MIAPPE-Wizard” as a web application that guides users in creating MIAPPE-compliant and ISA-formatted metadata for their plant phenotyping results. The wizard should provide the user with content recommendations and suggestions for ontology terms. It will ensure compliance with ISA- and MIAPPE recommendations and allow users to export and save generated metadata files as ISA-JSON. We will also provide the possibility to generate and publish FAIR Digital Objects (FDOs) by utilizing DataPLANT's ARC infrastructure through an open and reusable REST-API. By developing an easy-to-use web application with interoperability features, we aim to facilitate the adoption of FAIR principles and improve data accessibility and reusability in the plant science communities within de.NBI, ELIXIR, NFDIs, CEPLAS and beyond.
c) FAIDARE (https://urgi.versailles.inra.fr/faidare) is a comprehensive search portal harvesting diverse plant-centric data resources and providing a central access point for the plant research community. It addresses the findability aspect of the FAIR principles and is hosted and maintained by ELIXIR France and established as the European Search Portal for plant related datasets, in collaboration with EMPHASIS. To further increase the usability for the users we will connect additional established resources used by the plant research community. The main targets will be e!DAL-PGP, which is part of the de.NBI/ELIXIR Germany service portfolio and the CorkoakDB, which is provided by ELIXIR Portugal. To achieve this goal, we will use the common schema.org/BioSchemas markup to harvest and index these resources. In addition to these two major infrastructures, we will highly motivate additional resource providers to join us, and assess how this integration can be implemented to add their own system to the federated search portal of FAIDARE. To lower the barrier for connecting new data repositories, we also plan on generating training materials and ‘how-to’ guides to facilitate the future integration of additional data repositories. This will help to encourage the growth and expansion of FAIDARE as a comprehensive search portal for plant data and support capacity building for further resource providers. Therefore we are also looking for people who can support us in creating these materials and have knowledge on how to share and present it in a user appealing and helpful manner. This will enable researchers to access a wider range of plant data in a more efficient way and lead to greater collaboration and innovation within the field. Our vision is to create a comprehensive and user-friendly platform that serves as a one-stop-shop for all plant data needs.
Project Leads: Gabriel Schneider, (ZB MED, FAIRagro), <