RETAIN Fellowship
The Research Data Management Implementation in the Neurosciences (RETAIN) Fellowships
Establishing sustainable research data management (RDM) in neuroscience is challenging due to its interdisciplinary nature and the diversity of data — from cellular experiments to clinical and home-based studies. Despite growing efforts, such as those by NFDI, researchers often lack the time, resources, or tools to implement FAIR-compliant workflows. The RETAIN Fellowship addresses this gap by empowering early- and mid-career researchers — those closest to the data — with dedicated 24-month, 50% FTE positions. Fellows receive training in RDM best practices and work within their labs to assess needs, identify obstacles, and implement tailored, sustainable solutions. They create workflows, onboarding materials, and templates that are integrated into daily research environments. Guided by NeuroCure’s Coordinator for Value and Open Science fellows help turn their labs into RDM model environments. RETAIN thereby strengthens both RDM infrastructure within the NeuroCure community and individual careers—laying the foundation for broader adoption of FAIR practices across neuroscience and beyond. (10.5281/zenodo.13771491)
In 2024 fellowships were awarded to the following researchers:
Dr. rer. nat. Daniel Parthier
Project start date: January 1, 2024
Affiliation
Charité - Universitätsmedizin Berlin I Neuroscience Research Center
Research Focus
Project Summary: This preclinical data management project addresses the complexity of organizing and standardizing heterogeneous datasets in neuroscience research involving in vitro and in vivo electrophysiology and confocal/STED imaging. Key challenges include the integration of diverse proprietary and open-source electrophysiological -recording formats, fragmented metadata structures, and inconsistent file naming conventions across experiments and researchers. The project spans multiple acquisition systems generating diverse types of in vitro data with limited standardization. Deliverables include the implementation of a FAIR-compliant workflow centered around open formats like NWB (Neurodata Without Borders), creation of a lab-wide metadata schema, integration of electronic lab notebook documentation entries, and simplified metadata capture. Additional goals include the development of a low-barrier, Python-based tool to support standardization of data formats, analysis, and documentation, ensuring reproducibility and compatibility with public repositories as well as increasing awareness for data management in the lab environment through structure data onboarding process. This initiative will bridge technical gaps, improve data reusability, and support long-term open data sharing in cellular and systems neuroscience.
M.Sc. PhD cand. Jojo Vanhoecke
Project start date: January 1, 2024
Affiliation
Charité - Universitätsmedizin Berlin, Department of Neurology, Movement Disorders and Neuromodulation Unit (AG Kühn) - subgroup: Interventional and Cognitive Neuromodulation Lab
Research Focus
This clinical data management project addresses critical challenges in organizing, standardizing complex, and preserving multi-modal datasets from deep brain stimulation (DBS) in patients with movement disorders, such as Parkinson’s disease, dystonia and essential tremor. Key challenges include heterogeneous data formats (e.g., JSON, .mat, .eeg), fragmented metadata, an inconsistent data documentation across 20 years of diverse, high-value recordings, including local field potentials (LFPs) from hundreds of patients with DBS leads from different companies targeting the subcortical structures such as the subthalamic nucleus. Data acquisition has in the past 5 years expanded into chronic recordings in 150 patients through the sensing-enabled Medtronic Percept neurostimulator, and with the addition of smart devices more recently into recordings in the home environments, creating new data flow complexities. This RETAIN project aims to develop a robust, FAIR-compliant Data Management Plan grounded in the Brain Imaging Data Structure (BIDS) standard. Deliverables include semi-automated pipelines to convert Percept data into BIDS-like structure, a lab-wide data dictionary to harmonize metadata, integration of electronic lab notebooks, and standard procedures for assessing and improving data quality — such as signal fidelity and synchronization accuracy. These outputs will support reproducible research and pave the way for GDPR-compliant open data sharing of adaptive neurostimulation datasets, ultimately enhancing clinical care of Parkinson’s patients through precision DBS.