An Introduction to Our Work

We are the Biomedical Informatics Group, part of the institute for Adaptive and Neural Computation in the School of Informatics at the University of Edinburgh.

“We develop novel statistical and machine learning approaches to help us better understand human disease”

Most of our current research focusses on working with mixed collections of data derived from research studies of human disease. Over the last 5-10 years it has become increasingly common for these data to be comprised of several different data types. In many cases integrating these “multi-modal” data allows us to learn jointly from the data, typically this increases signal, minimises noise, and allows us to effectively use the power of much larger and complex representations of disease.

Our research uses data derived directly from groups of patients suffering from specific conditions who have consented to have some of their data shared for research purposes. These data are made available to researchers in a carefully controlled manner to preserve patient privacy and to ensure that the data is used only for the purposes originally intended. Some example of disease based initiatives that we work with are:

• The Parkinson’s Progression Markers Initiative (PPMI)
• The Simons Foundation for Autism Research Initiative (SFARI)
• The Cancer Genome Atlas (TCGA)

We also work with initiatives that are population rather than specifically disease based. Once again access to these data is strictly regulated and conducted in accordance with the strict usage conditions of the data provider, preserving the privacy and acknowledging the consent of the people involved. These initiatives contain hugely valuable and varied information about participants extending into the thousands or 100s of thousands in number. Some examples of population based initiatives that we work with are:

• Generation Scotland
• UK Biobank
• Genomics England
• All Of Us

You can find out more about our work by visiting our research page.