Share Data, Models and Software

Neuroscientific data - be they experimental or simulated data, models or software - are multimodal, heterogeneous, and organised in many different ways. Although most of these data use similar terms and concepts, it can be challenging to standardise these terms without losing the study-specific aspects of each data type. The process of organising, categorising and integrating data is called data curation. Data curations allows for the data to be found, compared and/or reused by other researchers. The EBRAINS Curation Service supports you in this process. Together, we will advance Open Science by making data "open".

"Open data" are shared without any restrictions on the reuse of data to promote transparency, accelerate scientific research, and encourage new collaborations. In some cases, there may be legitimate reasons to shield data from the public, for example, to protect sensitive data (e.g., safeguard the privacy of subjects), and to reduce the risk of being scooped by others. The “A” in the FAIR Data Guiding Principles (Wilkinson et al., 2016) stands for “Accessible under well-defined conditions”. The goal is to strive to make data “as open as possible, and as closed as necessary”, which ensures the safeguarding of data and prevents their misuse without compromising the possibility for reuse.

EBRAINS has clear guidelines on how data are shared. We ensure that all data are accompanied by licenses to allow for such restrictions; we have several options in place for releasing data under embargo until a scientific article is accepted for publication (read more); and we safeguard data via our access control solutions for sensitive human data (read more).

The FAIR guiding principles also serve to guide researchers in preparing data for data sharing and managing data and metadata appropriately. The EBRAINS Curation Service follows these FAIR principles in the curation of data and supports researchers and scientific publishers in their endeavours to improve the Findability, Accessibility, Interoperability, and Reusability (FAIRness) of their data.

The FAIRness of data depends on the amount of information that is shared, what format it is shared in, and the number of connections that can be made with other existing data. EBRAINS offers two levels of FAIR data curation that meet the needs of any researcher (see figure). Regardless of which track the researcher chooses, all data and metadata are integrated into the EBRAINS Knowledge Graph (KG)

The KG is a powerful semantic network developed to bring together information from different fields and illustrate their relationships. The KG is a graph database in which data, models, and software are integrated, and it is a powerful tool for connecting seemingly unrelated data (e.g. connecting datasets to software, analysis and visualisation tools).

All research products that are shared via EBRAINS are curated, a process in which important information, better known as metadata, is extracted and integrated into the KG. To give structure to metadata, we rely on metadata schemas that describe how metadata relate to each other. EBRAINS has adopted the flexible, scalable and extendable openMINDS metadata framework. openMINDS is grounded in standardised terminologies and ontologies, increasing the interoperability of datasets, models and software in the EBRAINS KG and with resources outside of the EBRAINS research infrastructure. openMINDS is continuously developing to accommodate curation-specific needs for particular research areas. One example of such an extension is SANDS. The SANDS extension was developed to enable the standardisation of anatomical locations of neuroscience data, both semantically using names of brain regions, and spatially via coordinates, allowing datasets to be linked to the atlases hosted by EBRAINS

The Knowledge Graph contains heterogeneous data, annotated with extensive domain-specific metadata, can be extracted via two ways that are tailored to the ability and preference of the user. The intuitive search user interface (UI) offers a low-threshold option for any researcher. The search UI has search filters that allow the user to refine their search based on data type, research modalities, methods, species, data accessibility, and keywords. For the advanced user, data and metadata can be accessed via an Application Programming Interface (API) and is compatible with many programming languages, enabling a versatile downstream implementation of search results.

More and more journals require a data availability statement on how data can be found and reused. EBRAINS caters to the needs of journal authors wishing to share data related to a manuscript submission to a peer-reviewed journal.

We offer several solutions for sharing your data alongside a manuscript, with various levels of accessibility that can help with the peer-review process whilst keeping the data secure.
For example, we can provide a temporary access to the dataset for reviewers, without making the data publicly available until the paper is published. Data can also be released under embargo, which allows others to find the dataset, but restricts them from using the data, until the paper is accepted.

All datasets curated and stored through EBRAINS will receive a DOI and are released to the public, making your dataset a published, citable resource.

Data papers are citable scientific papers that describe scientifically valuable datasets in more detail to promote the sharing and reuse of these resources. Many journals, such as Nature Scientific Data and Frontiers, now offer the possibility to publish data papers.

Data papers require datasets to be structured and annotated with enough metadata to make them understandable to others. They also require a unique URL, but preferably a DOI, that points directly to the published data files, and require a license that outline the terms and conditions for reuse.

All datasets that have gone through the EBRAINS curation process meet these requirements, making EBRAINS a suitable repository for dataset related to data paper. EBRAINS is also listed as a recommended repository for Springer Nature journals.

EBRAINS is committed to conducting responsible research and innovation. Part of that commitment is ensuring that any data shared through our platforms meets national, local and European Union ethical, social and legal requirements. All data providers looking to share data on EBRAINS platforms must complete the EBRAINS ethical compliance survey. The evaluation performed by the EBRAINS Ethics Compliance Management team is not an ethics review; EBRAINS does not provide ethical approval or seek to replace your local approval processes, we merely check that we can legally and ethically share your data on our platforms.

For identifiable data obtained from living human subjects, compliance with the EU General Data Protection Regulation (GDPR) must also be assured. De-identified, anonymised and strongly pseudonymised human data is additionally safeguarded via the Human Data Gateway (HDG) service. For the access to GDPR-sensitive data in the EBRAINS Knowledge Graph, users need to comply with a set of conditions, including the acceptance of an additional Data Use Agreement. We are also developing a service for sensitive data (SSD) that makes the sharing of pseudonymous and raw data possible. This service will be launched in 2023.

For questions or assistance, contact: hbp.compliance@dmu.ac.uk

To maximise the benefits of data sharing, data should be stored in a sustainable long-term repository. Following the FAIR data guidelines, EBRAINS offers such a repository. EBRAINS uses the OpenStack Object Storage (swift) system, provided by the Swiss National Supercomputing Centre (CSCS) as a part of the FENIX infrastructure. Researchers who are sharing their data through the EBRAINS curation service will be able to store their data free of charge for 10 years via the long-term data storage service.
If your data is already stored at a different sustainable data repository, you can still make use of the EBRAINS curation service. Metadata accompanying such data can be integrated into the Knowledge Graph and linked to the data via a persistent identifier (e.g. a DOI) once the external repository has been validated by the EBRAINS curation service.