NIH Data Management & Sharing
Vendors advertise “NIH Data Management & Sharing support” like it is a certificate they bought. It is not. This page explains what the policy actually requires, why no electronic notebook is “NIH certified,” and exactly how ResearchOS supports your compliance.
There is no such thing as a certified ELN
The NIH Data Management and Sharing (DMS) Policy, in effect since January 25 2023, binds researchers and institutions that receive NIH funding. It does not bind software. There is no NIH program that audits, licenses, or certifies an electronic lab notebook, and there is no badge a tool can earn. When LabArchives or anyone else lists “NIH Data Management Mandates & Sharing support” as a feature, that is a marketing claim that decodes to one honest sentence. Our features help your researchers do the things NIH expects of them.
Compliance is something a researcher does, not something a tool is. You comply by writing a Data Management and Sharing Plan with your grant application and then following it. The notebook is where the work that the plan describes actually happens. Any ELN with the right features can legitimately say it supports the policy, and ResearchOS has those features.
What the policy actually asks for
NIH expects funded investigators to do three things.
- Plan and budget for managing and sharing scientific data, before the work starts.
- Write a DMS Plan (a two-page document) and submit it with the funding application.
- Implement the plan by managing the data rigorously during the project, then share it no later than publication or the end of the award, whichever comes first.
How ResearchOS supports it
ResearchOS supports all three expectations. The same plan / create / implement structure other ELN vendors use to map their features to the policy applies here, with the bonus that your data lives in open formats on your own disk the entire time.
1. Plan and budget
Use separate projects to model how each grant's data will be organized, and reusable methods and templates so the whole lab captures data the same way from day one. Planning the structure up front is most of what a DMS Plan's data-organization section asks you to describe.
2. Create the DMS Plan
The plan itself is a short document built against NIH's required format. Most institutions point researchers at the free DMPTool, which has the official NIH template built in. Draft it there, then keep the finished plan in ResearchOS as a note or attachment alongside the project it governs, so the plan and the data it describes live together.
3. Implement the plan
This is where the day-to-day work lives, and where ResearchOS is strongest. The table below maps what the policy and NIH's Desirable Characteristics of Data Repositories guidance expect against the ResearchOS feature that delivers it.
| What the policy expects | How ResearchOS supports it |
|---|---|
| Provenance and integrity. A trustworthy record of who did what, when. | Append-only PI audit log (_pi_audit.json), per-task Lab Notes and Results stamped with owner and time, and soft-delete Trash with a recovery window instead of instant destruction. |
| Organization and description. Data grouped and described well enough to find and reuse. | Projects, tasks, and a reusable Methods library (PCR, LC gradients, plate layouts, cell-culture schedules), plus tags and filter-driven search across your projects. |
| Access control during the project. Decide who can see and edit what. | Per-record sharing with read or edit levels, a whole-lab option, PI view-all, and a password-gated edit session fronting every PI write. |
| Open, non-proprietary formats. Data that outlives the tool that made it. | Everything is written to a folder on your machine as plain JSON, markdown, and your original image and PDF files. No proprietary container, no vendor database. See Security. |
| A shareable package. A clean bundle you can deposit when it is time to share. | One-click export of any experiment, or many at once from Search, as a print-ready PDF, a self-contained HTML bundle, or a raw re-importable ZIP. Images and attachments travel with it. |
| Oversight. A named person responsible for data management. | The PI role, with soft-write approvals, flags, and the audit trail that records them. |
Where your data actually gets shared
The DMS Plan names a repository, and that repository is where sharing happens, not the notebook. NIH does not mandate a specific one. It publishes Desirable Characteristics of Data Repositories (persistent identifiers, long-term stewardship, curation, clear reuse terms) and expects you to pick one that fits your data, either a domain-specific repository where one exists, or a generalist repository such as Zenodo, Figshare, Dryad, or Vivli otherwise.
ResearchOS gets you to that handoff cleanly with a guided deposit dialog. You pick the repository (Zenodo, Figshare, or another of your choice), and ResearchOS prefills the deposit metadata in the DataCite shape repositories expect (title, creator with ORCID, abstract, license, keywords, and the funder and award pulled from the grant you linked), bundles the curated data, and opens that repository's own upload page so you drop the bundle in and the repository mints the DOI. Because your ResearchOS data is already in open formats, there is nothing to unpack or convert first. See Depositing to a repository for the step-by-step.
Honest limits worth knowing about
ResearchOS supports your DMS compliance, and it is honest about where a feature is a manual step rather than a built-in one. None of these block compliance, but you should know them before you write your plan.
The official sources
Do not anchor a compliance decision to a vendor blog (including this one). The primary sources are short and readable.
- NIH DMS Policy overview and the policy hub at sharing.nih.gov.
- NOT-OD-21-013 is the Final NIH Policy itself.
- NOT-OD-21-014 lists the required elements of a DMS Plan. This is the real checklist.
- NOT-OD-21-016 covers desirable characteristics of data repositories.
Wondering how ResearchOS stacks up against the tool most labs are leaving? See ResearchOS vs LabArchives.