Advancing FAIR practices in biomedical research: challenges, methods and innovations for sustainable data management

Bernabé, C.H. (2025)

Doctoral Thesis

This thesis presents a comprehensive strategy to enhance FAIR (Findable, Accessible, Interoperable, and Reusable) data practices in biomedical research, focusing on rare diseases. It addresses the challenges of making sensitive, fragmented data FAIR, highlighting the crucial role of data stewards and collaborative "Bring Your Own Data" (BYOD) workshops in developing best practices. To structure these efforts, this work introduces GO-Plan, a goal-oriented method for creating clear FAIRification objectives. The thesis also examines how foundational ontologies can improve data interoperability. A key finding demonstrates that restructuring knowledge graphs (KGs) with ontology based conceptual models enhances the consistency of machine learning predictions without sacrificing performance.

• FAIR principles
• FAIRification
• Ontologies
• Goal-based modelling
• Rare Diseases

• All authors: Bernabé, C.H.
• Supervisor: Mons, B.
• Co-supervisor: Roos, M.
• Committee: Spruit, M.R.; Cornet, R.; Reisen, M.E.H. van; Guizzardi, R.S.S.; Souza, V.E.S.
• Qualification: Doctor (dr.)
• Awarding Institution: Faculty of Medicine, Leiden University Medical Center (LUMC), Leiden University
• Date: 2025-11-26