One data infrastructure for One Health surveillance

Many diseases jump from animals to humans; think of the coronavirus, salmonella infection or Q fever. That is why society benefits from timely detection and surveillance of food-related diseases. It is also important that the various parties conducting research have access to an environment where they can share data. This would be a boost for One Health surveillance, an integrated approach to monitoring the health of humans, animals, and ecosystems.

Data often unfindable

And that presents a challenge, says Dirkjan Schokker, researcher at Wageningen Bioveterinary Research (WBVR): ‘This is relevant for humans, because animal pathogens can infect people through food. WBVR, Wageningen Food Safety Research (WFSR), and the RIVM conduct genome research from their own perspectives. In an ideal world, the data from all studies would be stored according to the FAIR principles: findable, accessible, interoperable, and reusable. But according to Dirkjan, this is not yet the case: ‘Suppose data was collected in response to a suspected outbreak. In many cases, it was not stored in a way that made it easy to find when a sample was collected and where it came from. The data was stored somewhere in a folder belonging to the researcher, but not in a central location.’ A new approach was therefore needed to share data from WBVR, RIVM, and WFSR. Dirkjan: ‘The first challenge was to make sure you had your own dataset in order. This led to the second challenge: ensuring that datasets could be linked.’

Solution for better surveillance

As part of the Wageningen Common Data Solutions programme, Dirkjan and his colleagues developed a solution to improve One Health surveillance. First, we looked for an open-source platform for managing, organising, and sharing data. iRODS is very good at that. The advantage was that we already work with this platform within WUR, and so does the RIVM. We then started filling a database with data relating to a case. We now have around two hundred samples in the database.’ Researchers from the three organisations can request and view genome sequences and associated metadata and, depending on their authorisation, use them for their own analyses. But the data itself remains where it is, Dirkjan emphasises.

Leaving a familiar environment

The database will be further expanded in the coming period. The more complete it becomes, the deeper the insights that can be gained, Dirkjan predicts: ‘Suppose you are a researcher who wants to know more about Salmonella outbreaks between 2020 and 2025. Soon, you will be able to find all the sequences and metadata from the various outbreaks in one place.’ This does require researchers to abandon the familiar Windows environment and transfer their data to the new research infrastructure. Dirkjan acknowledges that this is still an issue: ‘Researchers certainly see the advantage of having a single repository for data on animal health, food safety and human health. But it requires a change in behaviour, and that is not always easy.’

Learning by doing

Currently, Dirkjan and his colleagues are working hard on linking up with the RIVM system. Technically, it is not a complicated task, but the rules for data exchange are strict. ‘We are now at the point where we need to figure out how we can share data. That is a matter of learning by doing. My approach is to try to organise what needs to be organised. Challenges will always arise, and then we will see what we need to solve them.’ A follow-up project could build on what has been achieved in this WCDS project. If it were up to Dirkjan, it wouldn't stop there: ‘COVID is already a thing of the past, but it remains important to be prepared for the next pandemic. It's good that we're working together in an EU project: UNITED4SURVEILLANCE. The foundation we've laid will certainly not be lost.’