Catalysing a National Biomedical Data Science Solution
A toolkit for galvanising communities, bridging institutional silos, and building shared research data science infrastructure and skills.
The problem
For Scottish (and UK!) biomedical data science to remain competitive in the modern research landscape, there is an urgent need to break down all barriers currently preventing equitable and efficient access to regional data skills, training, and infrastructure.
The challenges are interconnected. Researchers often struggle to access appropriate compute and storage, but they may also lack access to training, skilled support, and the expertise of peers at other institutions. Universities invest in storage and compute infrastructure, yet may lack the resources to employ the experts needed to maintain such infrastrucutre. Within biomedical resarch communities, data scientists and software engineers often lack sustainable career pathways, resulting in loss of talent and expertise to other sectors. Consequently, the full scientific value of publicly funded biomedical data remains unrealised because the systems to connect people, skills, and infrastructure across institutions simply do not exist.
At a time when UK universities are under enormous financial pressure, this fragmentation creates scientific as well as economic losses. Institutions that share resources, coordinate training, and pool expertise will be more cost-effective, agile, and competitive than those that continue to duplicate in isolation.
This represents a coordination problem — spanning infrastructure, training, people, and institutional culture. Using the Scottish Biomedical Data Science context as proof-of-concept, this toolkit presents a practical approach to enabling the necessary coordination.
What this initiative set out to do
This initiative sought to galvanise the Scottish biomedical data science community: to identify barriers and enablers across infrastructure, training, and people, and to turn widespread frustration into positive, coordinated action. It had three core objectives:
- Collect shareable information about existing biomedical data science training, skills, and infrastructure provisions accessible to researchers across Scottish institutions
- Identify the cultural, financial, and technical barriers that prevent equitable access to and use of Scottish biomedical data science infrastructure and skills
- Co-design effective solutions for a future-proof, distributed biomedical data science enterprise — with multi-stakeholder commitment to action
Crucially, the initiative aimed to bring together key stakeholders who often do not find themselves in the same room:
- Individual researchers spanning different career stages and disciplines — from bench scientists and bioinformaticians to early-career fellows and established group leaders
- National and international infrastructure providers — HPC centres, cloud platforms, and data service organisations
- Communities of practice — networks working on FAIR data, research software engineering, and data stewardship
- Funders and government agencies
- Higher education, research institute and industry representatives
- Research software engineers and data scientists
- Legal team representatives — contracts, IP, and data governance specialists
- Core facility managers — the people who keep research infrastructure running day-to-day
What began as my practical frustration — difficulty accessing appropriate compute, data infrastructure, training, and bespoke expertise while establishing my independent research group — became a national conversation about how Scotland should organise its biomedical data science capabilities: the hardware and platforms as well as the people, skills, and collaborative culture needed to make them work.
Over approximately 18 months, this initiative moved through a structured sequence:
- Problem identification — recognising a systemic gap
- Evidence gathering — surveys and stakeholder consultations
- Community building — a virtual roundtable to evaluate key evidence and gather further information
- Workshop design — translating evidence into a structured, decision-oriented event
- Workshop delivery — bringing together the full spectrum of stakeholders listed above
- Collective ownership — senior stakeholders agreeing to carry the initiative forward
Key outcomes
The workshop achieved two objectives specified at the start:
A prototype for a co-designed framework structured around five pillars: infrastructure, governance, authentication and access, data ownership and intellectual property, and careers and training. This framework was agreed in principle by all participants.
A roadmap for action with defined phases over 12 months: working group assembly, landscape review, gap analysis, use case development, and a formal strategic pitch to government and funders.
Why this matters beyond Scotland
While this initiative originated in the Scottish biomedical data science community, the underlying challenge is broader. Any country, region, or research domain that needs to coordinate data infrastructure across autonomous institutions will face similar dynamics:
- Fragmented infrastructure, training, and expertise with no coordination layer
- Institutions duplicating costs and effort when sharing would deliver savings and agility
- Training provision that is invisible across institutional boundaries
- Financial pressures making coordination essential rather than optional
- The need to influence institutional leadership without formal authority
This toolkit captures what I learned so that others can adapt it to their own contexts — whether that is another nation’s biomedical data ecosystem, a multi-university research consortium, or any domain where resources, training, and people need to be coordinated across institutional boundaries under financial constraint.
“As a Future Leaders Fellow and somebody in a relatively privileged position, I saw it as my responsibility not only to do interdisciplinary science but to ensure that the conditions for doing such science exist for all researchers across Scotland. It turned out there was a lot of substrate and enthusiasm for a shared Scottish Biomedical Data Science effort, and all I had to do was act as the catalyst behind a collective and coordinated effort.”
— Dr Ralitsa Madsen, UKRI Future Leaders Fellow, University of Glasgow
How to use this site
| Section | What you will find |
|---|---|
| The Journey | A structured timeline showing how this initiative evolved from idea to action |
| The Framework | The co-designed five-pillar framework for national biomedical data science infrastructure |
| The Toolkit | A practical, phase-by-phase guide for running similar initiatives |
| Lessons Learned | Honest reflections on what worked, what did not, and what I wish I had known |
| Resources | Templates, frameworks, and checklists you can adapt |
Who is this for? This toolkit is designed for researchers who see a systemic problem and want to do something about it; for research managers and institutional leaders who need to coordinate across organisations; for funders designing infrastructure calls; and for policymakers seeking evidence-based models for national research data strategy.
Workshop delivery was funded by a UKRI Flexible Creative Fund award to Dr Ralitsa Madsen, Future Leaders Fellow (Round 8) at the University of Glasgow and the CRUK Scotland Institute.