Cross Institute Datathon: Asthma
Initiated by the Southampton Institutes, the latest Datathon programme was hosted at the Data Science Institute, Imperial College, creating the opportunity for diverse teams from Web Science, Medical Research, Data Science and Bioinformatics to explore the integration of social machines with medical data around the theme of asthma.
This interdisciplinary effort crosses Southampton campuses, includes the Southampton Biomedical Research Unit and works alongside colleagues at Imperial.
Prof Peter Smith, Director of the Institute for Life Sciences summarised the event: “Datathons are a unique opportunity to bring together divergent minds to consider complex datasets and the challenges of assimilating different formats. It was a great day, enhanced by the fantastic facilities and spectacular screens at the KPMG Data Observatory Facility, Imperial.”
The goal of the two-day meeting was to develop a pathway for integrating data. Such material has many sources; traditional medical research datasets, open data, personal data, mobile/sensor data, environmental data, crowd annotation and community discussions. Within an ethical framework our goal is to fuse and develop these datasets to deliver novel insights on severe asthma as it is experienced by patients.
This first cross-disciplinary, cross-institute meeting was a scoping exercise where participants shared information about the various datasets that might be drawn upon for a future data hackathon. As in all cross-disciplinary work, researchers have to first learn each other’s languages before considering which datasets can realistically be used for the practical experiments.
The impact of this approach was nicely summarised by Prof Dame Wendy Hall, Director of the Web Science Institute: “It is always the data integration process that throws up the difficult social and technical issues that need to be overcome before the real analytical work can begin. But it is also the linking of the different datasets across common entities where the jewels of new knowledge might be found.”
Imperial College were gracious hosts to this first Datathon, making available their outstanding resources for the visualisation of data and the expertise of their staff, ECRs and students. Prof Yike Guo, Data Science Institute, Imperial College London said “Through this data hackathon, we have successfully established a concrete social machine model in healthcare. We are keen to further develop this important research and to build a close collaboration with Southampton in data science research”.
Prof Ratko Djukanovic, Director of the NIHR Southampton Respiratory Biomedical Research Unit and close colleague of Prof Guo in U-BIOPRED (Unbiased BIOmarkers in PREDiction of respiratory disease outcomes), an international project to better understand asthma, commented “This was a wonderful event, to my knowledge the first such gathering; I look forward to working in the near future with web scientists to provide a bigger, social dimension to our biomedical data on asthma and our understanding of this disease.”