Unlocking complex diseases through federated data governance, FAIR multi-modal data, and AI

Venkata Satagopam*

University of Luxembourg

venkata.satagopam [at] uni.lu

Abstract

Neurodegenerative diseases (e.g., Parkinson’s and Alzheimer’s) and immune-mediated diseases (e.g., Inflammatory Bowel Disease, Multiple Sclerosis, and Rheumatoid Arthritis) are highly complex in their etiology, presenting major challenges for early diagnosis, patient stratification, and the discovery of robust biomarkers. Addressing these challenges requires access to high-quality, longitudinal health data and well-characterised clinical cohorts — yet such data are typically distributed across hospitals and research centres in different countries, under heterogeneous legal, ethical, and technical constraints.

This talk will present federated approaches to data governance, management, and analysis, and how translational medicine informatics enables the integration of multi-modal data — clinical records, multi-omics, imaging, and sensor/mobile data — while keeping sensitive data at its source. A central theme will be the substantial effort required for data curation, harmonisation, and FAIRification (making data Findable, Accessible, Interoperable, and Reusable) to enable cross-cohort and cross-border analysis.

I will showcase how statistical and Machine Learning (ML) methods applied to such multi-layered data can stratify patients by disease severity and progression, identify biomarkers, and underpin clinical decision support models. Complementing these data-driven approaches, disease maps will be discussed as a means to uncover the underlying mechanistic models and co-morbidities of these complex diseases.

Together, these efforts illustrate how federated, FAIR, and AI-ready ecosystems — aligned with European research infrastructures and regulatory frameworks — can accelerate translational research and improve patient outcomes in complex disease areas.

Keywords: Federated data analysis, FAIR data