Dr Sarah Bauermeister is getting ready to lead (at least) two more DPUK datathons before the year is up. She’s been known to get up at 4.45am because she’s so enthusiastic about her work, and is totally passionate about developing younger scientists in the dementia research field. We asked her why datathons are such a great opportunity for them.
"In my experience working with junior researchers, I notice that many simply haven’t had the opportunities to work with big and complex datasets. This might because they’re working in small research departments, or they only work with a single health study cohort collected at their institution. This is totally understandable but will become an increasingly outdated way of working. When I originally developed the concept of the DPUK datathons, I had junior scientists – each with stacks of potential and ideas – in mind.
In what is our flagship datathon at Swansea University, analysts and data scientists will be exploring key dementia research themes with a treasure trove of free tools. We guide them with research questions and provide the data to work with. Fundamentally, it’s about collaboration and support – exactly what junior researchers need.
The support we’ll offer
At our flagship datathon in Swansea we offer attendees access to 4-5 large cohort datasets, a full range of statistical tools (eg R, Python, Stata, SPSS), a virtual desktop and a selection of suggested research themes to take forward. But critically – they’re not left on their own. They are given chance to discuss in groups which theme they would like to work on. Then, using their own skills and some new ones, they’ll work on chosen datasets to answer a research question within their theme over the three days.
Using multiple methodologies across multiple types of data, bringing together the brightest minds in one room over three days means only one thing, exciting outcomes are going to happen. - Dr Sarah Bauermeister
Senior researchers are on hand to support throughout the three days. At the end of the datathon, the attendees are guided towards carrying on with their research by the scientific leads. Like our previous datathon, we expect these outputs to lead to full DPUK project proposals and publications.
Our research focus in Swansea
In Swansea we want to look at large datasets to find patterns in the data which will help us understand ‘healthy brain ageing’. We’re going to be looking at those people who have been followed for a long time in population studies and do not go on to develop dementia. What did they do that was successful? What did they eat? What diseases did they not have? What medications did they take or not take? In particular we’re interested in finding out about which other diseases and medical conditions ‘clump’ together. We're interested in knowing about what places an individual at a higher risk of dementia. These are the ‘comorbidities’ of dementia and this is becoming a key research area. We want to know about genetic links too. With innovative analysis techniques, DPUK’s population health study cohorts can tell us all this, and that’s why we’re supporting junior scientists to work with them.
Fundamentally, our aim with the DPUK datathon initiative is that research into the most pressing research questions will not be restricted by institution limits or, funder. We want to support and we want to collaborate. Our next datathon is an open door to all data scientists who are curious and want to tackle dementia head on."
You can find out more and apply for the Swansea datathon here.
What to read next
Swansea, the hub of the UK’s world leading informatics expertise, is where you’ll hear the hum of the hundreds of UKSeRP servers that are the home to the DPUK Data Portal. The virtual space of the Data Portal is, in reality, run from Swansea University’s state-of-the-art Data Science Building overlooking the Bristol Channel. Here, Mark Newbury is one of the people you’re very likely to be regularly in touch with. Mark’s on hand to help bring researchers to data, because, as all researchers know, it’s not quite as simple as a click of a button, just yet.
A branch of artificial intelligence (AI) which is based on training computers to learn patterns has the ability to transform our understanding of dementia. David Llewellyn co-leads the first DPUK datathon next month. He explains why machine learning could be the beginning of the end for this devastating condition.
25 March 2019
Ruby recently started as a postdoctoral researcher in DPUK. Alongside pursuing her research in factors affecting mental health and wellbeing, she contributes her valuable experience of working with cohort data to the developers working on the Data Portal upgrades.
Although the more commonly known ‘hackathon’ has been used to address societal-level problems from homelessness to corruption, we think that it’s only recently that the data science community is turning its attention to dementia. The DPUK datathon series might well be one of the first for the field in fact. Chris, DPUK’s Data Project Manager, and Sarah, DPUK’s Senior Data Manager, take us through the three key ingredients for a successful datathon.