Community needs for FAIR pathogen data

This study by the Pathogen Data Network identifies systemic barriers and specific training priorities among infectious disease stakeholders, revealing that limited funding, data aggregation challenges, and a need for bioinformatics education are the primary impediments to achieving FAIR pathogen data, thereby providing an evidence-based roadmap for community-responsive support and infrastructure development.

The Gist

Imagine a massive, global library dedicated to the world's most dangerous germs—viruses, bacteria, and parasites. This library holds the blueprints to every disease we've ever faced. If we could read these blueprints perfectly and share them instantly, we could stop outbreaks before they start and cure diseases faster.

However, right now, this library is a bit of a mess. The books are scattered in different rooms, some are written in code no one understands, and many are locked behind paywalls or missing pages. This is the problem the Pathogen Data Network (PDN) is trying to solve. They want to make this data FAIR: Findable, Accessible, Interoperable (works together), and Reusable.

To figure out how to fix the mess, the PDN team asked 136 experts (scientists, doctors, teachers, and data wizards) a simple question: "What's stopping you from using this data effectively?"

Here is what they found, explained through some everyday analogies:

1. The Big Three Roadblocks

The experts didn't say the problem was that the library was too small or the books were too hard to read. Instead, they pointed to three huge, structural issues:

• The Empty Wallet (74%): The biggest problem is money. Imagine trying to build a highway to connect all the library rooms, but you keep running out of gas. Scientists need funding to buy computers, pay for software subscriptions, and hire people to organize the data. Without the budget, the project stalls.
• The Tower of Babel (68%): The second issue is data silos. Imagine every country and hospital has its own version of the library, but they all use different filing systems. One group files by color, another by size, and another by the smell of the book. Trying to combine them is a nightmare. The data exists, but it's stuck in isolated boxes that don't talk to each other.
• The Missing Mechanics (52%): The third issue is a shortage of skilled people. You can have the best library in the world, but if you don't have librarians who know how to catalog the books or mechanics who can fix the shelves, the system fails. There just aren't enough people trained to handle this complex data.

2. What They Need to Learn

The survey asked, "If you could take a class to fix these problems, what would you study?"

• The Top Request: Bioinformatics for Infectious Disease. Think of this as learning the "universal translator" for germs. Scientists want to learn how to use computers to decode the genetic language of viruses so they can spot patterns and predict outbreaks.
• The "How-To" Guide: Many people wanted training on the Pathogens Portal. This is the PDN's new "central hub"—a single website where all the data and tools live. People want a user manual on how to navigate it without getting lost.
• The Difference Between Teachers and Researchers:
  • Researchers wanted "advanced tools." They asked for classes on Machine Learning and AI—basically, teaching computers to do the heavy lifting of finding patterns in the data.
  • Teachers/Educators wanted "real-world stories." They asked for case studies. They wanted to see how data saved a life in a specific outbreak so they could teach that story to their students.

3. How They Want to Learn

When asked how they wanted to learn these skills, the answer was clear: Keep it flexible.

• Virtual Short Courses (68%) and Webinars (66%) were the most popular.
• Why? Scientists are busy. They don't have time to fly to a conference for a week. They want to learn in 30-minute chunks while they are at their desks, just like watching a quick tutorial video on how to fix a leaky faucet.

4. The Most Important Tool

When asked, "What is the single most helpful thing the PDN can give you?" the answer was the Pathogens Portal.

• The Analogy: Imagine if all the scattered library rooms were finally connected by a giant, magical elevator that takes you directly to the right book. That is what this portal is. 72% of people said this central hub is the most essential thing they have.

The Bottom Line

The paper concludes that the problem isn't that the technology is too hard or the data is too scary. The problem is systemic.

It's like trying to build a skyscraper when you have the blueprints but no money for bricks, no crane to lift them, and no team to assemble them. The technology (the FAIR principles) is ready, but the infrastructure and the people to support it are lagging behind.

The Takeaway: To stop future pandemics, we don't just need better computers; we need better funding, better ways to connect different databases, and a massive training program to turn more scientists into data experts. The PDN is now using this survey to build that support system, ensuring that when the next germ threatens us, the library is open, organized, and ready for everyone to use.

Technical Summary

1. Problem Statement

Datasets related to infectious diseases are critical for public health decision-making and genomic medicine. However, their full potential is unrealized due to persistent barriers preventing effective sharing, integration, and reuse. While the FAIR principles (Findable, Accessible, Interoperable, Reusable) are widely recognized as the solution, the specific infrastructural and educational needs of the global pathogen data community have not been systematically characterized.

Current barriers extend beyond technical constraints to include:

• Insufficient institutional support and funding.
• Lack of expertise in submission processes and data curation.
• Data sensitivity and ethical concerns.
• Limited computational and bioinformatics skills.
• Siloed data systems that hinder aggregation.

There is a critical need to move from general assumptions to an evidence-based understanding of community priorities to design effective interventions for data stewardship.

2. Methodology

The study employed a cross-sectional stakeholder survey designed by the Pathogen Data Network (PDN), an international consortium focused on building a FAIR ecosystem for infectious disease data.

• Survey Design: A 10-item questionnaire developed de novo based on PDN strategic goals. It utilized multiple-choice, Likert-type, and open-ended questions covering demographics, institutional context, perceived barriers, and training needs.
• Participant Recruitment:
  • Target Population: Researchers, educators, bioinformaticians, healthcare professionals, and policymakers working with human health-related pathogen data.
  • Channels: Email campaigns (reaching ~10,000 individuals, including 7,000 faculty and 4,000 NIH/NIAID awardees), social media, PDN mailing lists, and outreach to 70+ global organizations (e.g., CDC, WHO, ECDC).
  • Timeline: Open from February 20 to April 25, 2025.
• Data Collection & Analysis:
  • Sample Size: 198 initiated, 136 valid completed responses (filtered for those answering role-specific questions).
  • Analysis: Descriptive statistical analysis (frequencies/proportions) using R (v4.5.1). Qualitative review of open-ended responses. No inferential hypothesis testing was performed; the goal was to characterize relative frequencies and majority trends.
  • Ethics: IRB exempt (Cold Spring Harbor Laboratory); informed consent obtained; no personal identifiers collected.

3. Key Contributions

• Empirical Baseline: Provides the first systematic, data-driven diagnostic of the barriers and needs within the global pathogen data community.
• Differentiated Needs Analysis: Explicitly contrasts the priorities of researchers (technical/analytical focus) versus educators (foundational/pedagogical focus), offering a nuanced view for training design.
• Resource Validation: Validates the Pathogens Portal (an integrated PDN tool) as the most critical resource for the community.
• Reproducible Framework: The survey instrument, analysis code, and anonymized raw data are publicly available, serving as a reusable template for assessing needs in other scientific communities.

4. Key Results

A. Respondent Demographics

• Roles: 58% Researchers/Analysts/Healthcare professionals; 26% Educators; 7% Developers.
• Geography: 57% United States; significant representation from Switzerland, Norway, UK, and India.
• Institutions: 66% Higher Education; 13% Government; 11% Research Institutes.
• Career Stage: 57% Mid-career professionals.

B. Primary Barriers to FAIR Data

Respondents identified structural and systemic constraints as the primary impediments, rather than purely technological ones:

1. Limited Funding: 74% (insufficient budgets for tools, experiments, and subscriptions).
2. Data Aggregation Challenges: 68% (difficulty combining data across siloed systems).
3. Shortage of Skilled Personnel: 52%.

• Notable Finding: Technological limitations and data misuse concerns were rated as "not significant" by the majority (60–61%), suggesting trust in frameworks exists but capacity is lacking.

C. Training Priorities

• Top Priority: Bioinformatics for infectious disease research (68%).
• Secondary Priorities: Guidance on the Pathogens Portal (51%) and data visualization/communication (50%).
• Emerging Interest: Machine Learning (ML) and AI applications (39%).
• Role-Based Divergence:
  • Researchers: Prioritized technical subjects (ML, data management planning, FAIR principles).
  • Educators: Prioritized foundational content (real-world case studies, conceptual overviews).

D. Preferred Delivery Formats

• Virtual Short Courses: 68%
• Webinars: 66%
• Insight: There is a strong preference for flexible, distributed online learning over in-person courses (53%), likely reflecting time and funding constraints.

E. Resource Importance

• Pathogens Portal: Ranked as the most essential PDN resource by 72% of respondents.
• Other valued resources included metadata standards (43%) and validated analytical workflows (37%).

5. Significance and Implications

• Systemic vs. Technological: The findings challenge the assumption that FAIR data barriers are primarily technical. Instead, they highlight that funding, workforce development, and inter-institutional coordination are the critical bottlenecks.
• Strategic Alignment: The results provide a roadmap for organizations (like PDN) to align infrastructure and outreach with empirically identified needs. Specifically, investing in bioinformatics training and centralized portals yields the highest community return.
• Future Directions: The paper argues for sustained investment in training and institutional continuity. It suggests that future interventions must address the "human element" (skills and staffing) alongside tool development.
• Community Engagement: By identifying the Pathogens Portal as the central hub, the study validates the strategy of centralized data integration. It also highlights the need for tailored training tracks for researchers (technical depth) and educators (pedagogical breadth).

In conclusion, the paper establishes that while the community recognizes the value of FAIR data, realizing this potential requires addressing systemic resource gaps and providing targeted, flexible, and role-specific training, with the Pathogens Portal serving as the cornerstone for interoperability.