DataHub Workflow Demonstration

On this page you will find workflow demonstrations showing key components of our harmonized research data management and the tools used. These videos are intended to give a first impression and cannot replace working through our workflow tutorial or a personal meeting with the DataHub Team. The videos were recorded in November 2024.

Storage

Video: Preparing data for storage in the DataHub. Template folder structures and the BIDS Standard are applied. Data is submitted to version control by Git + Git-LFS and exchanged with the private TAM-GitLab. The TAM-GitLab is the central platform to store and share data and code, for collaborative work and for automatization. It provides access to the TAM storage resources in the MaRC3 HPC environment.

Compute

Video: Using the DataHub's central compute resources. Affiliated researchers can access the MaRC3 compute cluster directly via commandline to submit compute jobs to the cluster management system. TAM members have privileged access to the TAM-owned resources or they use the pooled idle capacity of the whole cluster. The JupyterHub offers a beginner-friendly and browser-based interface to the HPC cluster for interactive compute and data visualization. Both environments are prepared for seamless data exchange with git.

Publication

Video: Publishing research data and code with DOI on the DataHub Repository. TAM's DataHub Repository is a curated research data repository for neuroscience and psychology. Researchers can annotate their finalized data with additional metadata to increase findability and reusability. Accessibility can configured down to file-level. The metadata schema extends the interoperable Dublin Core standard with domain-specific metadata. Researchers worldwide can find published research by browsing / searching the repository and via the metadata transferred to the DataCite registry.