FLEX4HEALTH project - representative image

Science cluster

LS RI - Life Sciences

Summary

In recent years, large multi-center cohort studies and exposome research consortia have led the way in creating successful analysis networks using metadata catalogues and federated analysis.

FLEX4HEALTH aims to scale federated health data analysis across Europe by integrating leading research

Infrastructures (RIs) - BBMRI, ELIXIR, and EIRENE - with established Open Science platforms, such as MOLGENIS for FAIR-compliant data cataloguing and DataSHIELD for secure, privacy-preserving analytics. By bridging these RIs with interoperable tools and standards, and increasing their technical readiness levels, FLEX4HEALTH will strengthen Europe’s capacity for sustainable, collaborative, cross-border health research while ensuring data sovereignty, inclusivity, and compliance with FAIR and Open Science principles.

Research domains:
Life Sciences
Partner(s):
University Medical Center Groningen – UMCG, Barcelona Institute for Global Health - ISGLOBAL
Project team member(s):
Morris Swertz, Tim Cadman, Juan Ramon González, Martine Vrijheid

Challenge

Rapid advancements in biotechnology and information technology offer enormous potential for personalised healthcare by leveraging diverse phenotypic, environmental, and molecular data.
Until recent times, EU researchers could use large cohort studies for this research, such as LifeLines, ALSPAC, or NFBC cohort studies. However, to reach sufficient statistical power, diversity and longitudinality, it is essential to combine and jointly analyse data from many cohort studies across regions and countries. The integration and analysis of extensive, sensitive health datasets remain challenging due to technical complexity, privacy concerns, and fragmented resources.

Solution

Building on initiatives, such as DataSHIELD and MOLGENIS, FLEX4HEALTH aims to improve the accessibility and technical readiness of these tools, and consolidate the development work from large EU consortia using multi-centre cohort studies, in particular EUCAN-connect, ATHLETE, LongITools and LifeCycle. 

The project will firstly create a scalable governance where all stakeholder groups can direct the infrastructure while promoting diversity and inclusivity. Secondly, it will facilitate seamless integration of diverse health data sources (from BBMRI, ELIXIR, ECRIN, EIRENE), and provide comprehensive training materials, thus promoting wide adoption within the broader scientific community, including research institutions, healthcare providers, and citizen scientists.

Enhancements will also focus on incremental user interface improvements, performance optimisation, and integrating new functionalities. 

Finally, FLEX4HEALTH will implement an onboarding campaign to introduce ESFRI community members into the new developments, addressing essential gaps to enable analysing complex OMICS and imaging data.

Scientific Impact

By enhancing interoperability, governance and scalability, FLEX4HEALTH will significantly expand the reach and usability of federated analysis across biomedical domains. The project will reduce key analytical gaps - especially for complex omics and imaging data, which are priorities for BBMRI, ELIXIR, and EIRENE - while leveraging synergies with initiatives such as GDI, EOSC4Cancer, ERDERA, and EUCAIM. This will allow federated AI model training in medical imaging, demonstrating the practical applicability of federated analysis in cutting-edge research areas. 

It will reduce operational and support load, by simplifying operations, integrating into existing support networks and optimising training/documentation using ESFRI solutions, to enable scaling up. In particular, it will reduce barriers for smaller institutions lacking extensive IT resources by integrating and disseminating via the Life Science ESFRI infrastructures, making use of their tools and training options. Moreover, further streamlining the operations and data maintenance in MOLGENIS catalogue and DataSHIELD to enable upscaling with the ESFRIs will mean increasing technical readiness from TRL-5 to TRL-7.

Keywords
federated learning, federated AI model training, federated analysis, omics data, imaging data, data catalogue, MOLGENIS, DataSHIELD, FAIR data, Open Science
Project start date:
Project duration:
24 months

Principal investigator

Morris Swertz - FLEX4HEALTH Principal investigator's picture
Morris Swertz
UMCG - University Medical Center Groningen
BIO

Prof.dr. Morris Swertz is professor at UMCG where he holds the chair "Big FAIR data in biomedicine". In 2008 he founded the Genomics Coordination Center (now 40+ staff) supporting the Dept. of Genetics and international consortia. His research focuses on methods and tools for DNA, RNA and phenotype data analysis. He developed the MOLGENIS framework, including VIP for genome analysis and Armadillo for federated data. He holds IT leadership roles in BBMRI-ERIC, ERN registries, Health-RI, and EU consortia like Solve-RD, EJP-RD, and the Genome Data Infrastructure.

QUOTE
"FLEX4HEALTH scales federated health analysis by bridging RIs with platforms like MOLGENIS and DataSHIELD. By advancing FAIR cataloguing and privacy-preserving analytics, we reduce barriers for smaller institutions and enable secure, cross-border Open Science innovation for all researchers."
Resources