As a result of a partnership with Astrazeneca, Computer Science MSc students working under the supervision of Dr Gavin Brown, Professor Andy Brass and Professor Uli Satler, have been given the opportunity to take on some interesting and challenging projects in the field of BioHealth Informatics.

The projects are sponsored by Dr James Weatherall, Astrazeneca's Global Director for Biomedical Informatics, and will see students working in close collaboration with staff at the Alderley Edge research site.

One such project, being undertaken by MSc student Elisabeta Marinoiu, will apply feature selection to the problem of adverse event selection. Feature selection is a statistical technique used in machine learning to automatically identify which features (biomarkers) within a given set of data are relevant, irrelevant or redundant in the context of other features. These techniques can be employed to help practitioners identify critical features and understand the interdependencies between them.

Elisabeta's project is informed by the challenge of personalised medicine. During the pharmaceutical development of a drug it is vital that all safety signals are investigated and considered. Within a single clinical trial, patients are treated at different times and at different doses of a drug and adverse events are reported over the course of the trial.

Elisabeta's project will analyse time-dependent data from a clinical trial in order to select the key features of a patient (age, sex, BMI etc.) that predict whether a patient experiences a particular adverse effect, and when. The project will focus on a framework of mutual information and will deliver a prototype implementation of the solution.

Commenting on the projects Dr Brown said:

"The partnership with Astrazeneca is providing a unique opportunity for students here in Manchester whilst facilitating knowledge transfer from academic research to industry practice. We're proud to be working with AZ on a topic as pressing and important as cancer prognosis."