Building the infrastructure for open science: A conversation with Carole Goble

Author: Guy Jones

Chief Editor of Scientific Data

This is the seventh and final blog post in our series in which researchers share their experiences with open science practices, and reflect on the impact that sharing open data, code, and protocols can have.  

Carole Goble is Professor of Computer Science at the University of Manchester and a leading figure in open science, FAIR data, and research infrastructure. A co-author of the FAIR Principles and co-founder of the UK Software Sustainability Institute, she has spent nearly three decades building the tools, platforms, and collaborations behind transparent, reusable research. We spoke with her about why infrastructure, incentives, and culture must evolve together. 

Can you briefly introduce yourself and your area of expertise?  

My specialisms are semantic technology, metadata, computational workflows, and disrupting scientific communication through open science. For over a decade I've been joint Head of Node of ELIXIR-UK, the UK's national Node for ELIXIR, the European research infrastructure for life science data. We have 31 organisations in the UK, while ELIXIR spans 22 countries across Europe and coordinates the management and sustainability of life science data. 

I'm a co-author of the FAIR Principles and a founding partner of the European Virtual Institute for Research Software Excellence. I’ve spent much of my career developing open science infrastructure, including FAIRDOM-SEEK, myExperiment, and its successor, WorkflowHub. Until recently I was co-lead of Health Data Research UK, and I serve on the G7 Open Science Working Group as the UK expert. 

How did you first become interested in open science practices, and what motivated you to adopt them?  

“The Human Genome Project showed that science at scale depends on bringing data together and sharing it openly.” 

                                          Carole Goble

I began my career in health data, which isn't open at all, and moved into bioinformatics in the late 1990s. The life sciences have a long history of sharing results in databases for the common good. The Human Genome Project showed that science at scale depends on bringing data together and sharing it openly. 

That set up an obvious question: why not approach everything this way? Through the 2000s there was real momentum, alongside a deeper question about whether scholarship is just about publishing papers. More and more information was becoming available, and people were doing more analysis with computers and software, but almost none of it survived beyond publication; it disappeared the moment it went into the paper. So how could we introduce living papers that share the underlying analysis and data? 

That's when I started myExperiment, a way to share your workflows and analysis. The idea that we should share how we did something took off, not just to pass on know-how, but to show transparently how the work was done. This ties into research integrity and reproducibility: there was a bit of a reproducibility crisis at the end of the 2000s. People publish papers, but how can you stand behind what they did if they haven't shared the data or analysis behind it? 

These aren't new ideas, they go back a long way, but I became part of a community that drove the push for transparency, reuse, and recognising the value of artefacts beyond the written narrative: data, software, standard operating procedures. A lot of my work was in the biosciences because they were so open to it. 

Have you ever faced any particular challenges or barriers in undertaking open science practices?  

The barriers aren't technical. Brian Nosek's Strategy for Culture Change pyramid sums it up, though it has one flaw: the shape implies a hierarchy, with policy tiny at the top and infrastructure broad at the bottom, as if you build it in layers. I don't think that's right. It's really a series of interconnected circles because you have to work on all of them at once to move the needle. Policy isn't enough on its own, you need the infrastructure that lets things be shared, the skills that make it easy, and the incentives that give me a reason to make my work not just available but genuinely reusable. Making something available is easy; making it reusable is not. 

We see this constantly: counterfeit sharing, where you say you've shared something but haven't because there isn't the metadata for anyone to use it. That's why we push so hard on the FAIR aspects. Then there's something less like sharing than flirting: just enough metadata to interest you, but not enough to reuse anything. So, you approach the people who created it, take them to dinner. Eventually they say, ‘Yes, we can work together, and you can put me on your paper’ because the paper is still the currency. 

It's easy to say something is open, but how open is it really? People still operate on incentive models built around authorship rather than reuse and citation, and they're genuinely bad at citing, whether it's data, software, or credit for the people who did the work. That feeds the same problem. You have to make it normative. That's the top of Nosek's pyramid: it becomes standard practice. 

There's one piece missing from his pyramid: usefulness. He has possible, usable, mandatory, normative - but not useful. All the infrastructures I've built are still going. We keep trying to kill off myExperiment, it's so old and we've got WorkflowHub now, but people still use it. I still see papers published around it because it's useful. The challenge is fitting the pieces together: getting policy pushed through, the incentives right, perceptions shifted. It's hard, and that includes publishers as much as funders. The funders are actually pretty good. It's the publishers who could play a bigger role. 

Is there a particular success story or example you’re proud of that illustrates the benefits of open science?  

WorkflowHub and FAIRDOM have been great successes for sharing a particular kind of software, and because the infrastructure itself is open, people set up their own instances. Another is the TeSS training portal, a central hub for life sciences training, and an example of both shared software and shared practice. The platform underneath it was taken up by high-energy physics, CERN in particular, and by other countries: Australia, Canada, the Netherlands, Sweden, each reusing it for their own training. Now we've come together in a European Commission–funded programme to build a single shared platform: instead of each running our own instance, we run one shared service. That's the mTeSS-X project. 

This is what happens when you make things open: other people, organisations, countries, and disciplines reuse them and work with you to deliver. That matters when nobody funds training portals. Everyone says training and skills matter, but you can never get the money. This is where open proves its worth collectively. 

How has open science shaped collaboration, sustainability, and real-world impact throughout your research and educational work? 

“The benefits [of open science] are enormous. Organisations feel part of the team, which keeps the work going, and working openly lets you sustain what you do: the more platforms we develop, the more people use them and the more momentum you build.” 

                                                                      Carole Goble 

My entire career is built on it. We're serial collaborators who've built everything jointly. ELIXIR as a whole has around 300 organisations across its 22 countries, open working on a scale that's hard to imagine. The benefits are enormous. Organisations feel part of the team, which keeps the work going, and working openly lets you sustain what you do: the more platforms we develop, the more people use them and the more momentum you build, not just in how research becomes open, but in finding new income streams. To be blunt, when the chips are down in one country, another steps in to help. 

My measure of importance is impact. Did it make a difference? Could people share their workflows and collaborate around their data? If so, that's a success. Think about it the other way: you publish a paper, nobody cites it, what's the point? You've added a line to a CV. There has to be something that actually comes out of it. 

If we'd just done it in our own little bubble, who would know about it? Who would join us to develop it? We wouldn't have had the adoption, sustainability, innovation, or collaboration. We wouldn't have been part of ten Horizon Europe projects right through Brexit, when the UK wasn't even associated with it. Think how many people saw their careers end because the funding dried up. We carried on. That's a testament to our open practices. 

Are there any gaps in support or resources about open science you’d like to see addressed?  

The UK has shown remarkable foresight. It funded the Software Sustainability Institute in 2010, and the eScience Programme led by Tony Hey, started around 2000, is twenty-five years old now. That programme laid the foundations for much of today's eScience activity, for the work I’ve done, and for many of the national digital research infrastructures now being built across the physical sciences, biosciences, and beyond. The European Commission has been enormously helpful too. So, the funders have genuinely driven this; they've had the vision. 

I'd like the publishers to be more engaged. Picture four corners. The funders are one, and they often engage. The communities are another, and they’re very engaged. Then the institutions, even more conservative than publishers: they only change when forced from outside, and that force usually comes from publishers and funders, because institutions respond to money and fame. The fourth corner is the publishers. They hold the fame ticket, and the ticket to access and reuse: can I get the method and the data? If publishers had more of a voice in how we sustain the resources that archive that data, we might build more impetus to hold funders to continued support. 

So, we could get far more from publishers and institutions than we currently do. We could also achieve much more with a systemic way of managing our open research infrastructures. At the moment, they're funded piecemeal. We're still using an early-twentieth-century funding model to sustain international infrastructures, and that's a big problem. 

What advice would you give to researchers considering open science practices, and what misconceptions need to be challenged? 

If you're thinking about making things available, I'd first ask: are you reinventing something that already exists? Do you really need to build it? What's the cost of reuse? The second is to find out what support your organisation offers to prepare your data and software for reuse. It takes time and resources, so build it in from the start. There's no point pretending it's all easy and free. It isn't, but the payoffs in collaboration and citation are big, and there's plenty of evidence that papers people can access get cited more. So, build it into your processes and make it normative but expect to resource it or to push your organisation to. 

If you do nothing else, stick it in a database and put it on GitHub. Use an electronic lab notebook, anything that makes it a bit better. Don't just say, ‘No, I won't do it at all.’ We need to train up the next generation and build these skills within institutions — because institutions are where researchers live. 

Remember that FAIR doesn't mean open. It means accessible. You can be FAIR without making everything available. It's also not free: everything has to be paid for, and somebody carries the cost. Another misconception is that if something's open it can't be worth anything. Actually, it's worth a great deal. Open software powers industry. OpenUK has good data on how much UK business depends on open infrastructure. Open data is what powered AlphaFold. If that data hadn't been available, AlphaFold would be an algorithm pootling around in a corner. Open data drives this new generation of AI-driven research, and it's hugely valuable. 

One final point: open needs to be protected. Making it open doesn't mean people can steal it. Open also carries a responsibility of citizenship. Without it, we get a tragedy of the commons. In fact, we've already entered one, where people exploit open resources for their own gain. Open is vulnerable if it isn't protected. It's for the public good, so don't just expect it to still be there. Don't take it for granted. 

Learn more about open science and sharing research data, code and protocols & methods openly.

Carole Goble, Professor, School of Computer Science, University of Manchester, UK 

Carole Goble is Professor of Computer Science at the University of Manchester and a leading figure in open science, FAIR data, and research infrastructure. A co-author of the FAIR Principles and co-founder of the UK Software Sustainability Institute, she has spent nearly three decades building the tools, platforms, and collaborations behind transparent, reusable research. Professor Goble serves on the G7 Open Science Working Group as the UK expert.  

Related content:

• Open science conversations: 

1. Building trust through transparency: An open science conversation with Geir Kjetil Sandve 
2. Open science, altruism and impact: An interview with clinical geneticist Zornitza Stark
3. Bioinformatician Johannes Koester on embodying the spirit of open science 
4. How to do science that matters: Computational biologist Casey S. Greene on purpose-driven open science practices 
5. Open science before it had a name: An interview with molecular biologist Steven Henikoff 
6. From code to chemistry: A conversation with Connor W. Coley on open science practices and reproducible AI research

•  Best practices for transparency and reuse: 

1. How to share your research protocols and methods openly 

2. How to share your research code openly 

• Supporting open science practices: 

1. Why share your research data? 

2. Why sharing protocols matters 

3. Why sharing your code matters 

Don't miss the latest news and blogs, sign up to The Researcher's Source Monthly Digest!