RFA-RM-26-017: Pilot Projects Enhancing Utility and Usage of Common Fund Data Sets (R03 Clinical Trial Not Allowed)

RFA-RM-26-017: Pilot Projects Enhancing Utility and Usage of Common Fund Data Sets (R03 Clinical Trial Not Allowed)

If you are building research ideas around existing biomedical data ecosystems, this call is unusually practical: it rewards teams that can quickly produce evidence by combining NIH Common Fund data at scale and then turning outputs into reusable scientific value. It is a one-year pilot structure, so reviewers expect specificity, execution discipline, and a clear argument that the project improves data usability rather than chasing a broad programmatic vision.

Key details

What this opportunity is designed to fund

This NOFO is a data-ecosystem grant and review posture different from many translational clinical grants. It explicitly prioritizes utility, integration, and reproducibility of existing public resources. The language in the program text is clear that the intended use is to run small, concrete projects that can accelerate knowledge from multiple Common Fund data sets and generate outputs the broader community can reuse. In review terms, this matters because feasibility and usefulness of data integration are central, not just whether your science is imaginative.

The expected winning projects tend to share three features:

• They start with strong, auditable evidence that the relevant data are already available, accessible, and fit the scientific question.
• They produce a defined deliverable within 12 months (a workflow, a harmonization pipeline, a pilot validation, or a reusable analysis approach).
• They include a concrete plan for sharing outputs and methods.

The Common Fund list in the NOFO is extensive and includes resources such as 4D Nucleome, GTEx, HuBMAP, HMP, and MoTrPAC. The proposal must integrate at least two CF resources, with MoTrPAC-only submissions constrained by the stated exception language. This makes proposal framing non-negotiable: you need to define a hypothesis that is impossible to pursue with a single dataset alone.

Who should apply and why this is not just a generic R03

The opportunity is not for every early-stage project. It is for teams comfortable working with pre-existing high-dimensional data, multiple repositories, and explicit validation goals. If your group normally depends on lab-generated data and does not have strong data engineering capacity, this can still work if your concept is small and focused, but you need to prove that your team can execute integration and interpretation on schedule.

This matters if you are choosing between opportunity types:

• If your project needs prolonged wet-lab characterization and major new sample collection, it may be more expensive and too broad for a pilot R03.
• If your project is primarily data-intensive and can be scoped to a one-year test, this is likely a stronger match.
• If your team has already built a prototype pipeline but needs evidence of broader utility and externalization, this call is often a better fit than submitting a full standalone hypothesis-heavy grant.

The NOFO is broad by subject matter but narrow by mechanism: any topic can qualify, but only if the Common Fund data strategy is central and convincing.

Eligibility and compliance details you should plan around

The official eligibility section is deliberately broad by organization class but strict on structure and status:

• U.S. entities can apply; foreign entities cannot apply as direct recipients.
• HEIs, nonprofits, for-profits, governments, and certain tribal and public organizations are generally eligible.
• The mechanism is New submission only (R03, not a resubmission pathway in this framing).
• Clinical trials are explicitly not allowed.
• Foreign subawards/subcontracts are not allowed.

The restriction on international subawards is important. Teams with intended overseas collaborators should structure them as unfunded collaboration or in-kind support only where appropriate and ensure no disallowed financial agreements are in the scope.

The same applies to PI eligibility nuances:

• PD/PI must hold an eRA Commons account.
• PI can be any qualified individual with skills, knowledge, and resources for the planned work.
• There are explicit exclusions for certain current DCC-associated personnel from specific related programs (including key personnel associated with specific Common Fund ecosystem leadership roles, according to the NOFO).

From a practical perspective, the biggest compliance trap is administrative, not scientific:

• SAM registration (including UEI)
• eRA Commons account alignment (SO and PD/PI roles if applicable)
• Grants.gov account and submission-ready privileges

The NOFO notes registration delays can exceed six weeks, so these are best started before the project itself.

Timeline and money model

The call opened 2026-05-23 and had a deadline of 2026-06-23 in the official table. The expiration date appears as 2026-06-24, indicating the deadline window is the primary submission gate.

Financially, each proposal can request up to $200,000 in direct costs for the one-year term. NIH states the FY 2026 Common Fund pool for this NOFO at about $4,000,000, with 10–13 awards anticipated depending on merit and budget.

For applicants, there are two planning implications:

• A strong application can be high-impact with a conservative budget because this is a pilot class and the scope is short.
• Experimental components are allowed, but budget share is constrained when experimentation is proposed (the NOFO states experimental work should be limited to around 20% of requested budget in related sections).

You should therefore design the project around data-centric outputs first, then reserve bench or validation work for essential feasibility checks rather than broad discovery blocks.

How to structure a competitive application

Because review criteria are heavily method-driven, a weak narrative but technically sound structure will fail later than a strong narrative with weak methods. Your narrative should be built around the NOFO’s own review logic.

1) Start with data-grounded significance

Your opening section should answer:

• Which scientific gap cannot be addressed without combining two or more Common Fund resources?
• Why are those two specific resources complementary?
• What new hypothesis emerges from the integration that cannot be tested with one dataset?

Keep this explicit. Reviewers repeatedly score significance against clarity and novelty, but they also check that your hypothesis is actually enabled by your required data architecture.

2) Build one research design that fits 12 months

Every month should have a planned deliverable. A common winning pattern looks like:

• Month 1: finalize access checks and harmonization strategy, including preprocessing and metadata assumptions.
• Month 2: run baseline integration and feasibility checks; pre-register analysis pathways and QC thresholds.
• Month 3–8: execute the main analyses, refine models, and produce first outputs.
• Month 9–10: test robustness, sensitivity, and negative controls where possible.
• Month 11–12: prepare data products, repository handoff, and final interpretation in a format useful to Common Fund stakeholders.

This is just a template. If your project requires less computation but more synthesis, invert the timeline, but preserve that each phase produces measurable progress.

3) Align each work package with review criteria

The review text groups scoring into Significance, Rigor/Approach, and Investigator/Environment, with additional considerations. Translate this directly:

• Significance: link your project to real-world utility improvement for a named Common Fund resource.
• Innovation: show novel integration or analytic leverage, not just another re-analysis for a single dataset.
• Approach: describe reproducible steps, assumptions, missing-data handling, and validation logic.
• Rigor: if human-subjects work exists, describe handling of covariates and protection of privacy.
• Resources: clarify compute, storage, and access-level readiness.

4) Plan data sharing from day one

The NOFO requires explicit planning for data management and sharing. Reviewers look for specificity in

• how outputs will be deposited,
• where tooling will be hosted,
• expected formats and metadata standards,
• who can reproduce the workflow.

If your project creates a pipeline or curation product, include a clear open-access delivery path and documentation package. If you do not, this is a common area where projects look technically competent but strategically weak.

5) Keep experimental spending bounded

If you include experimental validation, cap spending to the recommended portion and document exactly why these experiments are essential to evaluate hypotheses generated from secondary data. Do not dilute 12-month computational milestones with long-running experiments that delay deliverables.

Common mistakes and how to avoid them

1. Treating the NOFO as a generic hypothesis-only grant. Avoid broad objectives. If integration, utility, and dissemination are not the spine of the project, the fit is poor.

2. Using only one data set. The minimum requirement is at least two eligible Common Fund data sets, with explicit justification. State this in the abstract and Project Summary in plain language.

3. Undervaluing compliance readiness. SAM and eRA Commons issues are preventable causes of late or noncompliant submission. Start registration checks immediately.

4. Ignoring reviewer criteria on feasibility. A beautiful computational idea without a timeline and milestone plan is often scored down in Approach and Rigor.

5. Submitting late or relying on corrected submissions. The NOFO clarifies late submissions and changed applications are treated strictly; aim to file early enough for corrections.

6. Submitting foreign collaborations as funded subawards. If non-U.S. partners are essential, re-check whether their role can remain unfunded, advisory, or non-subcontracted.

Who should not apply

Skip this call if:

• you need multi-year translational outcomes
• the project depends mainly on data collection rather than reuse
• your team cannot justify access to multiple eligible Common Fund datasets before the award start
• your submission timeline cannot complete regulatory/accounting prereqs
• your plan includes clinical trials or prohibited funding arrangements

This is not an exclusion of quality science; it is a scope alignment issue.

Preparation plan from inquiry to submission

A practical sequence that fits this cycle:

Week 1–2:

• Read eligibility and restrictions line by line.
• Confirm PI role and eRA Commons access.
• Pull candidate CF datasets and check access gates.

Week 3:

• Freeze scientific question.
• Choose two to three Common Fund resources with complementary biological axes.
• Build a one-page feasibility matrix (variables available, sample compatibility, ethical controls, compute needs).

Week 4–5:

• Draft significance and innovation sections.
• Draft approach with methods, controls, and fallback plans.
• Prepare preliminary reproducibility and data management plan.

Week 6:

• Build application materials with strict page limits and template instructions.
• Ask internal compliance team to review budget split and registration status.

Week 7:

• Internal pre-submission review and final QA.
• Submit early enough to allow corrections before 2026-06-23.

Use this plan to reduce rework. In NIH pilot settings, application quality is mostly a function of clarity and feasibility, not volume.

Practical fit analysis for different applicants

Universities and research labs

This call works well for teams with experienced data analysts and data governance familiarity. It is strongest when the project has a clear path to reusable outputs and not just one-off analysis.

Small biotechnology or analytics firms

Applicable if the company has an internal team that can deliver integration and methods work in a short time. For-profit applicants are eligible but should avoid spending budget on generic commercialization sections not connected to utility goals.

Local governments and nonprofit institutions

These organizations are eligible and can be strong if they partner with strong technical leads and can show project management readiness.

DCC-connected teams

Review conflict lines carefully; some CFDE leadership roles may be ineligible as applicants. Verify PI history to avoid an avoidable policy mismatch.

FAQ based on the NOFO

Are clinical studies allowed? No, clinical trials are not allowed. Other human-subjects analyses are possible if compliant, but the trial pathway is not permitted.

Do I need exactly two datasets? At minimum two Common Fund datasets are required for eligibility. You can add others if they materially improve the design.

Can I include experimental validation? Yes, but the NOFO emphasizes pilot analysis and limits the experimental component to a bounded share of budget.

What is the project term? One year max.

Can I submit more than one application? Yes, if each is scientifically distinct. NIH will not accept duplicate/high-overlap applications pending one another.

Is there support during submission? Yes: NIH and Grants.gov technical channels are listed in the official contacts.

Official links

• Official NOFO text (primary): https://files.simpler.grants.gov/opportunities/c1d89214-0c56-4030-9166-31c9bd209052/attachments/edb440ad-ecd5-494d-bbd6-52240b642c98/RFA-RM-26-017-Full-Announcement.html
• NIH Common Fund programs page: https://commonfund.nih.gov
• NIH How to Apply - Application Guide: https://grants.nih.gov/grants/how-to-apply-application-guide.htm
• NIH eRA Commons: https://public.era.nih.gov
• Grants.gov: https://www.grants.gov

Final takeaway

RFA-RM-26-017 is not a broad “more money to study anything” grant. It is a deliberately narrow pilot mechanism for teams that can turn existing NIH Common Fund resources into new scientific leverage quickly and responsibly. If your best value proposition is data integration that expands reuse, this call is strategically good fit; if your proposition is primarily generation of new data and long-cycle validation, another mechanism is usually a better match.

For 2026–2027 planning, this is one of the few opportunities where the funding window is short but the strategic upside is high: if your project can prove utility in one year with reproducible outputs and transparent sharing, you get both review signal and portfolio-level visibility without committing to a longer, heavier grant.