Data Science

Theme leads

Research Focus

The Data Science Theme is dedicated to transforming the treatment of mental health disorders through the integration of routine clinical data, research datasets, and advanced digital technologies. By combining diverse data sources and applying cutting-edge analytical methods, the programme aims to enable personalised care pathways that improve patient outcomes and reduce healthcare costs. The work builds on the foundation established by the previous Biomedical Research Centre (BRC) Informatics and Digital Health Theme, with a renewed focus on precision mental health and real-world clinical application.

Regional Context

The research is led from Oxford, with strong national and international collaborations. Key academic partners include Harvard University, the University of Toronto, and the Centre for Addiction and Mental Health, alongside industry collaborators such as Janssen and Akrivia Health. These partnerships provide access to global expertise, advanced infrastructure, and diverse patient populations. The Theme is also aligned with national data science initiatives, including Health Data Research UK’s DATAMIND, ensuring integration with broader efforts to enhance mental health research through data-driven innovation.

Research Aims

The Theme is structured around three interconnected work packages (WPs), each designed to address a critical aspect of data-enabled mental health care:

WP3: Risk Stratification in High-Risk Populations
This work package will refine and expand existing predictive models for adverse outcomes in schizophrenia, bipolar disorder, and forensic settings. Tools such as OxMIS, OxMIV, FoVOx, and OxRec will be enhanced to improve their clinical utility, linking risk assessments more directly to treatment planning and safety strategies.

WP1: Translational Precision Mental Health
This work package focuses on the integration of multi-modal data including clinical records, imaging, outcome measures, and digital device data to create a large, research-ready cohort. This infrastructure will support targeted recruitment into diagnostic and therapeutic studies, enabling more precise and effective interventions.

WP2: Digital Phenotyping
Building on existing platforms such as True Colours, this work package will capture remote data from electronic devices to better understand mental health trajectories. It will apply artificial intelligence (AI), natural language processing (NLP), and informatics to analyse behavioural and physiological signals, using data from established cohorts such as the Bipolar Disorder Research Network and OXTEXT.

Impact

The Theme is positioned to deliver significant translational impact by embedding data science into routine mental health care. By enabling personalised treatment strategies, the work aims to reduce the trial-and-error nature of current approaches, improve clinical outcomes, and optimise resource use within the NHS. The development of scalable digital tools and predictive models will support earlier intervention, better risk management, and more efficient service delivery. In parallel, the Theme will continue to build research capacity through training, fellowships, and collaboration with institutions such as the Alan Turing Institute, ensuring a sustainable pipeline of expertise in this rapidly evolving field