Aligning with NIH Data Sharing Mandates

Understanding NIH Data Sharing Mandates

The National Institutes of Health (NIH) has robust policies on data sharing to promote open science, transparency, and reproducibility. These are primarily outlined in the NIH Data Management and Sharing (DMS) Policy, effective since January 25, 2023. It applies to all NIH-funded research generating scientific data, replacing the earlier 2016 Data Sharing Policy. Below, I'll break it down for you, focusing on key requirements. This is based on official NIH guidance (as of my last update in 2023; always verify the latest at NIH Grants Policy or the DMS Policy page).

1. Scope and Applicability

• Who it affects: Applies to all extramural research (grants, cooperative agreements) producing scientific data, regardless of funding amount (> $0 threshold since 2023). It includes intramural and some contract research.


• What counts as "scientific data": Digital recordings of observations (e.g., images, text, audio, or video) from instruments, sensors, or models, not including lab notebooks, preliminary analyses, or physical specimens. Metadata and documentation are included.


• Exceptions: Limited to cases like proprietary data, legal restrictions (e.g., HIPAA), or if sharing would cause harm. Even then, justification is required.

2. Key Requirements

• Data Management Plan (DMP):


• Required for all new and competing renewal applications submitted on or after January 25, 2023.


• Submit a DMP (up to 2 pages) describing how data will be managed, preserved, and shared. It must address:


• Types of data to be generated.


• Standards for data formats and metadata (e.g., FAIR principles: Findable, Accessible, Interoperable, Reusable).


• Storage, backup, and security measures.


• Timeline for sharing (data should be made available no later than the end of the award or upon acceptance for publication; maximum 1 year post-publication in most cases).


• Mechanisms for access (e.g., public repositories like NIH's Generalist Repository or domain-specific ones like GenBank for genomics).


• Reviewed during peer review but not scored; non-compliance can affect future funding.


• Data Sharing:


• Data must be shared to allow validation of results or further research, unless restricted.


• Use established repositories (e.g., dbGaP for controlled access, Zenodo, or Figshare for open access). NIH provides a list of repositories.


• Include a Data Availability Statement in publications linking to shared data.


• Budget for sharing costs (e.g., repository fees) in grant applications; allowable up to the end of the project period.


• Timeline:


• Plan submission: With the grant application.


• Data sharing: As soon as possible, but no later than project end + 1 year (or publication + 1 year for unpublished data).


• Compliance reporting: Annual progress reports and final reports must confirm adherence.

3. Aligning Your Practices

• Audit your current workflow: Review your projects for data types and check if they generate "scientific data." Identify gaps in documentation, storage (e.g., use GitHub or OSF for versioning), and metadata.


• Tools and Resources:


• Use DMPTool (dmptool.org) for templates tailored to NIH.


• Adopt FAIR data principles; tools like DataDryad or Dryad help with deposition.


• For sensitive data, consider controlled-access options via dbGaP, including access agreements.


• Training and Compliance: NIH offers webinars (e.g., via NOT-OD-22-198). Track updates via the NIH Extramural Nexus blog.


• Consequences of Non-Compliance: Can lead to withholding of funds, restrictions on future awards, or required remediation.

If your query was cut off (e.g., "NIH" and more funders like NSF or EU Horizon), provide details for others. For personalized advice, consult your institution's grants office or NIH program officer, as requirements can vary by institute (e.g., NIAID has additional rules). Let me know how else I can help!