Multi-Modal Data Integration and Machine Readability between OMERO and ARC

Bioimaging datasets are acquired in a variety of file formats that often demand specialized software to be managed according to RDM (Research Data Management) standards. OMERO [1] (Open Microscopy Environment Remote Objects), a (meta)data management platform, specialized in microscopic imaging, is designed for this purpose and is a widely utilized tool in microscopic imaging [2]. More generic approaches like ARC [3] (Annotated Research Context), based on FDO [4] (FAIR Digital Object) principles, provide data annotation for heterogeneous data types respecting their individual research context. By design, an ARC can handle diverse data by combining a hierarchical directory structure with metadata annotation defined by the ISA (Investigation, Study, Assay) abstract model [5] but does not yet provide capabilities for working with complex microscopic images. Here, our work presents various approaches to integrate these technologies to make the bioimaging data management interoperable. To make use of the respective strengths of OMERO and ARC, raw data, related metadata and analysis results are converted to be consumable by each. However, both OMERO and ARC have different folder structures and therefore, the OMERO-ARC interoperability requires mapping of respective folder structures. To this end, we developed a recommended mapping schema, workflows and a software plugin to transfer (meta)data from OMERO to ARC and ARC to OMERO. The omero-arc plugin [6] being developed is based on the community tools omero-cli-transfer [7] and ARC Commander [8] and the goal is to automate the workflow of (meta)data exchange between OMERO and ARC. This way, the multimodal (meta)data in an ARC can be transferred and visualized in the graphical representation in OMERO to which the bioimaging users are accustomed, while changes in OMERO will be archived and versioned in the PLANT dataHUB [9]. For smoother interoperability of metadata and machine readability, an additional layer of metadata exchange and interoperability between OMERO and ARC can be a valuable step towards the goal of Fair Digital Objects. To this end, we are exploring plugins like omero-rdf [10] that creates a stream of RDF (Resource Description Framework) triples for metadata in OMERO and outputs it in various formats including json-ld & RO-Crate [11]. On the other hand, metadata from ARC can already be exported as a json-ld file with semantically rich metadata using various tools it provides to export ARC as RO-Crate. An ongoing effort of standardizing the contextual information in the bioimaging domain including metadata from OMERO will make the metadata machine readable, actionable and interoperable.