gms | German Medical Science

Introduction: Developing heart patches for repairing the heart muscle in patients with heart failure as well as spontaneously contracting engineered human myocardium (EHM) for drug discovery are core expertise of the Institute of Pharmacology and Toxicology in Göttingen [1], [2]. This translational biomedical research presents challenges beyond the scope of standard tools for research data management (RDM) due to complex, heterogeneous data from highly specialized techniques. Here, we describe our conceptualization for managing data from cardiac tissue engineering, focusing on the feasibility of already established RDM tools from the infrastructure framework provided by our group in the context of various biomedical consortia [3]. ????

Methods: To understand the RDM needs within the described biomedical context, we initially conducted interviews with experts from relevant working groups. Their insights were then translated into user stories using personas to capture user needs effectively. In parallel, we analyzed the data structure of a clearly defined, yet heterogeneous and complex dataset studying EHMs for up to twelve months. Finally, we assessed the feasibility of applying locally existing RDM infrastructures and software tools. We implemented the EHM dataset into these systems and evaluated their ability to meet the diverse stakeholder requirements identified earlier.

Results: Our requirements analysis revealed that the scope of what RDM should provide differed greatly across user groups. For instance, users with programming skills prioritized automatic extraction of standardized data with machine-readable metadata for AI-assisted analysis. In contrast, non-programmers preferred user-friendly interfaces for data exploration and visualization across various data types. The exemplary EHM dataset included longitudinal microscopic and contraction data alongside end-point measurements, particularly from OMICs techniques, adding challenges concerning large file volumes and the need for extensive preprocessing workflow documentation. It also confirmed that while in most working groups there are established, if not necessarily standardized digital workflows focusing on each lab’s specific experimental expertise, there is still a lack of infrastructure for FAIR sharing of data with (internal) collaborators. Encouragingly, the EHM data structure could be successfully mapped to the ISA-tab format within SEEK [4], fulfilling user needs for comprehensive crosslinking and granular access control. Microscopic data was uploaded to OMERO [5] for enhanced visualization, while large data files were stored in a research data archive with cross-references and metadata maintained in SEEK. However, this approach revealed limitations regarding search functionalities and seamless data integration across experiments.

Discussion: Sustainably reusing existing RDM infrastructure offers cost advantages in software development and maintenance. However, it is crucial to balance this with providing sufficient additional values for users to foster their engagement in FAIR data sharing. In our case, SEEK serves as a valuable foundation for organized data storage. Yet, additional tools should address data integration, high-performance computing (HPC) connectivity, advanced search functionalities, and visualization approaches as expressed by the designated users. We have initiated explorative prototype development for further enhanced data integration, while also contributing valuable feedback to the collaborative open-source SEEK project based on our experience.

Acknowledgments: Funded by DFG through CRC1002 (INF), Z projects of CRC1190, CRC1565 and Germany’s Excellence Strategy - EXC2067/1- 390729940.

The authors declare that they have no competing interests.

The authors declare that an ethics committee vote is not required.