Establishing a data management plan

Requirements of funding bodies


The European requirements: H2020 programme

Adopting a data management plan, prior to giving open access to scientific data when the conditions are met, was first confined to a few fields supported by the H2020 programme before being extended to all projects from 2017 on.
Now, any beneficiary of H2020 funding must present a data management plan (see the model) within six months of the beginning of the project. It should be updated as the project progresses, depending on decisions (patent application, new partners, etc.). When there is no valid reason to the contrary, such as intellectual property or classified information, data must be made available, especially when it “validates results presented in scientific publications”.

These requirements also apply to researchers funded by the ERC (European Research Council) which is a significant pillar of the H2020 program (17% of the total budget). The ERC has developed its own DMP framework, available here .

The costs of open data, including those generated by the application of a management plan, are eligible for H2020 funding. Two distribution licenses (CC0 and CC-BY) are recommended. They are based on a very broad conception of data reuse, covering commercial uses. This choice is influenced by a vision that open science is important for innovation in the private sector.

Data preservation

There are no specific directives about data preservation when a project is finished. For the Magenta Project supported by H2020 and involving 10 institutions including the CEA, Zenodo was chosen for the long-term preservation of data from the project.

Yet Zenodo does not present itself as an archiving platform able to provide data readability over time. For more information on data retention, see the following articles.

Projects funded by the ANR

In line with the French open science plan, the open science policy of the French National Research Agency (ANR) has requested a Data Management Plan (DMP) for projects funded since 2019.

This policy reflects the logic applied to European projects: the document must be provided within 6 months after scientific work begins. Data management plans for projects lasting over 30 months must be regularly updated and presented in 3 versions.

The creation of the DMP OPIDoR portal, a collaboration tool, facilitates preparing a data management plan. For more information see the article here.

The model put forward by the ANR serves as a guide, but is not mandatory. It is possible to use a different model, either from the institution or a partner.

At this stage, data management plans are not an essential selection or support criterion for different projects. It is, however, a deliverable and may be a condition of the payment of the final due for payment at the end of the project (limited to 10% of the total budget grant).

Funding data management costs

Data storage and management costs are eligible for funding up to 5 years after the end of a project. Costs are eligible for funding, but they do not justify an increase in budget.

Projects funded by the Swiss National Science Foundation

Since October 2017, project leaders must provide a data management plan. This is a condition of funding being paid and can be updated throughout the project. The final version must be made available in the P3 database of the SNF.
The SNF drafted a checklist to help researchers to identify which warehouses meet the FAIR principles. In their declaration of principle, the Swiss foundation requires researchers to archive and share data used in their research in public data repositories in accessible and reusable formats.

Projects funded by the US National Institutes of Health (NIH)

The National Institutes of Health (NIH) finances biomedical and health research in the United States.

A “data sharing plan” is required for projects funded from $500,000. Costs for sharing and archiving can be indicated in the funding application and are eligible for funding. Not publishing data must be duly justified.

Requirements are stricter than in France: data must be shared no later than when work is accepted for publication by a publisher and funded bodies must retain the data for three years after the end of the contract. Data from big studies can be distributed gradually as they are available or published. More information here.

Useful resources