Running datathons for dementia
by Guest Author on 23 Sep 2019
What’s a datathon – and why should you care? Professor John Gallacher, the Director of Dementias Platform UK, reveals the inspiration behind their popular data-tastic events and the benefits they offer.
We started the datathon events to help equip scientists with the skills that they need to use our secure online environment, which contains health data for dementia research. In a nutshell, for three days up to 50 data scientists work together in small teams to begin tackling a research question using our data resource.
Dementias Platform UK (DPUK) launched five years ago with a mission to enhance research into dementia, one of the greatest health challenges facing this generation. Cohort studies, where researchers study people over a long period, are crucial for improving our understanding of dementia. But smaller studies can’t answer the toughest questions on their own. That’s why we created a secure online data store which contains 35 cohorts and upwards of three million study participants’ health research data, known as the DPUK Data Portal.
As we say goodbye to old-fashioned paper and data becomes increasingly digital, I think it’s important that people are familiar with using online data systems. But we discovered that while we’d built something that we thought was a good resource, people often didn’t know how to use it – hence our datathons!
First and foremost, our datathon’s are about teamwork. Attendees split into teams of four to six people and together decide on a question that they will work on over the three days. While the teams have free reign to choose what they focus on, each datathon has a theme to provide a starting point for exploring the data – such as cognitive decline, or healthy ageing and dementia prevention.
All scientists are welcome, whether they have experience in dementia research or not. This creates a diverse mix of skills and backgrounds in the room. For early-career researchers, it’s a fantastic opportunity to build a network of multi-disciplinary contacts. I think that this social element is extremely important for peer to peer learning. Not only that, but previous attendees found it to be a stimulating environment and ultimately, an enjoyable experience.
As part of the data management process, new additions are cleaned to make them as easy to use as possible. For example, two cohort studies may categorise gender differently – male and female or women and men – and that makes them harder to compare. Where possible, we make sure that the data are speaking the same language.
But even with this process, the data we have available are still highly complex. We begin every event with workshops that introduce our online data platform and equip scientists with the skills they need to carry out their research. Then it’s time to get stuck into exploring and analysing the data.
A highlight of the datathons is that they bring people together who are experts in the types of data we have, who know what they’re talking about. Cohort volunteers have provided a wealth of data – including blood samples, brain scans and genetic data – to be stored within our data portal. I’m passionate about supporting analysts to handle these different types of data, as dementia research must be approached from multiple angles to answer the most difficult questions.
The end of the event doesn’t mean the end of the team’s work. We provide continued access to the data for a further eight weeks so that they can continue to develop their projects with the aim of publishing the results. Being a relatively new programme, we’re a way off from seeing that but some teams have progressed their ideas into full proposals and plan to continue research using the data.
Datathons are just one part of the DPUK programme, and our aim is to help scientists generally access and analyse large complex datasets. Through our Data Portal we provide a solution that ensures data security and allows global access.
Once online systems become the go-to option for data management and access, the need within the scientific community will be different, and the datathon will need to evolve. We’re not quite there yet, but I’m optimistic that this is the future. Come along to a datathon near you and be part of the change.
Find out more and apply for a free place at the next datathon.
In their first five years, Dementias Platform UK (DPUK) have changed the dementia research landscape by bringing technology, expertise and volunteers together to accelerate the development of new treatments. The Data Portal is a free remote-access resource in which researchers work with data from multiple cohort studies.
No comments have been posted