Better data - better neuroscience
Heterogeneity, complexity and growing volumes of neuroscience data, paired with the demand to re-use valuable experimental data, make it increasingly important to provide infrastructural support for the data management tasks of neuroscientists.
Emerging solutions need to be integrated and further developed to achieve a coherent and efficient data management framework for neuroscientists, addressing the specific needs of the researchers.
The NFDI Neuroscience consortium aims at building a community to develop the conceptual and practical basis for managing research data for the neurosciences.
The consortium is supported by the Neurowissenschaftliche Gesellschaft (NWG), the Bernstein Network for Computational Neuroscience, the Deutsche Gesellschaft für klinische Neurophysiologie (DGKN), and EBRAINS.
Building a community-centered process where scientists benefit from and contribute to improving the research data management infrastructure.
NFDI Neuroscience aims to help researchers addressing their needs with existing solutions and to trigger and coordinate the development of new solutions where necessary. Community involvement on different levels is essential for this process – from the contributions of individual researchers to the representation of large collaborative research projects and institutions.
Offering a unifying framework for standardizing data handling that seamlessly connects the lab data procedures to remote data services.
NFDI-Neuro will advance and disseminate a federated interoperable ecosystem for data and for reproducible research. To guide the goals of NFDI, Neuro, we have conducted a community survey and published its results (DOI:10.12751/g-node.w5h68v) serving as basis for the workplan. Various communication channels have been used for dissemination of the survey, including those of professional societies and social media. The survey was aligned with the survey conducted by our partner NFDI4BIOIMAGE and now serves as basis for a follow-up world-wide survey developed by the INCF.
Setting up a work program with a unifying approach and common objectives that addresses the RDM needs in different neuroscience subdomains.
Advancing previous RDM work:
NFDI-Neuro advances an existing ecosystem of standards and services and entails their dissemination to a broader community: data and metadata models, provenance tracking through version control and the executable storage of processes with DataLad and container image technology applicable for various data types and not limited to a particular spatial scale and hence with potential for the entire NFDI. Preceding collaborations between Fraunhofer SCAI, Charité, FZJ with synergistic aims in the EOSC project Virtual Brain Cloud and in the Human Brain Project led to the delivery of promised infrastructure results and to significantly funded follow-up projects such as eBRAIN-Health in collaboration with the European Platform for Neurodegenerative Diseases (epnd.eu). The challenges of integrating different semantic frameworks are being addressed by our partners in internationally leading roles. Efforts in the standardization of acquisition protocols and procedures are a crucial part of RDM, which does not imply a quest for a single neuroimaging standard. Making data FAIR includes the establishment of workflows for annotation, quality control, and versioning. Data quality assurance has a broad set of diverse aspects, of which the most important ones are being addressed by the project, including completeness of data sets, compliance to naming and data set structure standards, signal quality, presence of reference measurements, and accessory data for artefact correction procedures. Automation ensures reproducibility and re-usability. Workflows that require manual interaction can be made reproducible by tracking the user decisions with DataLad as recently demonstrated in our publication in Neuroimage.
Training and Dissemination:
Working groups, transfer teams, workshops, interactive decision trees, demonstrators, and extensive educational material are just a few examples. NFDI-Neuro has organized in its preparation phase 10 webinars, 5 community workshops, a special issue on NFDI, a published community survey, and has conducted weekly or biweekly community meetings over the past two years with an average attendance of >40 persons per event. Training materials are discoverable even at an international scale via the INCF Training Space and the EBRAINS portal.
Nine Use Cases have been selected from existing large-scale collaborative research projects in Germany – thus representing state-of-the-art requirements of the German Neuroscience community. Use Cases are cross-cutting the various RDM domains covered in the work program.
One of the NFDI-Neuro aims is “Addressing ethical and legal challenges linked to the development and use of digital twins for research, including: a) Representation, bias, consent, inclusivity of research and promotion of diversity. b) Legal capacity shared and supported decision-making, communication, trustworthiness, and trust.” Details can be found in the work program.
Data protection and sharing:
Data protection measures comprise technical and organizational measures including the conclusion of various types of processing, sharing and service agreements for processing or storage of data in research infrastructures.
Added value for the community
Our consortium is significantly contributing to several cross-cutting goals of NFDI. We focus on advancing operational solutions and bringing them to many labs in Germany.