- Create a framework for federated services to advance health data science and support the deployment of advanced analytics by researchers
- Integrate with and leverage opportunities across the infrastructure programmes to add value and maximise user benefit
- Collaborate with partner organisations to establish a community that can drive standards and best practice while encouraging innovative approaches to federation and its application
- Improve transparency around the technical engineering and governance of federated analytics to make it easier for developers from other disciplines to get involved
Trusted Research Environments (TREs) are secure locations in which data are safely available for researchers to analyse, however it is difficult for researchers to perform analyses across multiple TREs. Variability amongst TRE implementations and the analysis tools that use them prevent this federated approach to data access and analysis. Federated analytics at scale requires a solution that’s straightforward for existing, independent systems to adopt, that can cope with the variety of system implementations, and works within the Five Safes framework that enables data services to provide safe research access to data.
During the TRE-FX and TELEPORT pilot projects supported by DARE UK, proof of concept solutions to support federated analytics were developed. This work demonstrated how to streamline the use of analysis tools across multiple TREs while ensuring that the access is safe.
The Federated Analytics Programme will build on this work to ensure that all HDR UK Research Driver Programmes, and the wider scientific community, have access to advanced federated analytics services.
The Federated Analytics Programme is working hand in hand with ELIXIR UK and is adopting many ELIXIR technologies, most notably RO-Crate (a research data packaging initiative), and building on the work of the Five Safes RO Crate (part of the TRE-FX work) and infrastructure for computational workflows, such as WorkflowHub.
The intention is to re-use existing standards, especially those from the Global Alliance for Genomics and Health (GA4GH), into the solutions we develop. All our developments will be in the open and with permissive open source licences, for example HUTCH. Technology decisions will be taken to support the needs of the HDR UK Research Driver Programmes and our other partner programmes including BY-COVID and EOSC-ENTRUST.
HDR UK funded institutions:
- Swansea University
- University of Dundee
- University of Manchester
- University of Nottingham
- Arjuna Technologies
- BC Platforms
- DARE UK
- Receive updates from the Federated Analytics Programme by joining the mailing list.