Oxford-GSK collaboration in Biostatistics and Artificial Intelligence in Medicine

The GSK-Oxford Collaboration in Biostatistics and Artificial Intelligence in Medicine will draw upon biostatistical and artificial intelligence (AI) expertise across Oxford, whilst prioritising the development of innovative new tools and methods that translate to medical and biopharmaceutical research.

The programme is led by Professor Chris Holmes Professor of Biostatistics in Genomics at the Big Data Institute together with Dr Nicky Best, Vice President and Head of Statistics and Data Science Innovation.

By supporting small teams of researchers and data scientists working closely with senior Oxford and GSK scientists, the collaboration will harness Oxford’s strengths in statistics mathematics, engineering science and AI to address major challenges and bottlenecks in developing new medicines.

Oxford and GSK have identified the following thematic areas of initial interest:

• Overcoming model misspecification
• Uncertainty in large language models
• Intelligent data gathering
• Clinical trials, models and simulation
• Causal reasoning
• Multiomic machine learning across biobanks.

Principal investigators leading the research teams are:

• Professor David Clifton (Department of Engineering Science)
• Professor Robin Evans (Department of Statistics)
• Associate Professor Yarin Gal (Department of Computer Science)
• Associate Professor Agni Orfanoudaki (Saïd Business School)
• Professor Tom Rainforth (Department of Statistics).

New methodologies arising from the partnership will enable researchers to utilise previously inaccessible patterns and associations from vast amounts of complex, multi-layered health data. The approach will inform better design, decision-making and interpretation in clinical research, speeding up discoveries that improve outcomes for patients.