Why structure matters in rare disease research
Research in rare neuromuscular diseases is not only limited by data availability. It is often constrained by how fragmented that data is.
Cellular models, omics analyses and functional assays can each provide valuable insights. But when these datasets are generated and analysed separately, it becomes difficult to build a coherent understanding of disease mechanisms, especially in rare conditions where patient populations are small and biological variability is high.
DREAMS addresses this challenge by exploring how biological research and computational approaches can be integrated into a more structured discovery pipeline.
From biological models to complex datasets
A key element of DREAMS is the development of comparable biological models across several rare neuromuscular diseases that share underlying mechanisms, particularly related to autophagy and protein organisation.
Using induced pluripotent stem cell (iPSC)-derived skeletal muscle cells, the project generates reproducible disease models. These are characterised through transcriptomic and proteomic analyses, phenotypic screening approaches, and functional assays targeting disease-relevant processes.
Together, these methods produce multi-layered datasets that capture different dimensions of disease biology.
However, the challenge lies in how these datasets are connected and interpreted.
The role of AI in supporting data integration
In DREAMS, artificial intelligence is used to support, not replace, biological research.
Computational approaches are applied to analyse heterogeneous datasets in combination, identify patterns across different disease models, and support the identification of potential shared targets.
These analyses rely both on experimental data generated within the project and on curated external datasets.
Importantly, computational outputs are treated as hypotheses. They require further validation through experimental models, ensuring that results remain grounded in biological evidence.
Building a structured discovery pipeline
To address data complexity, DREAMS is organised around a structured pipeline that connects different stages of research.
This pipeline includes:
- Generation of cellular models and datasets: iPSC-derived muscle cells enable reproducible modelling across multiple diseases.
- Data integration and computational analysis: Biological datasets are analysed using computational methods to identify patterns and shared features.
- Target identification and prioritisation: Machine learning supports the identification and prioritisation of candidate targets.
- Experimental validation: Targets and candidate compounds are evaluated using established cellular assays.
- Extension to additional indications: The project explores whether identified mechanisms are relevant across a broader range of diseases.
- Contribution to clinical research design: Methodological work supports the development of clinical trial approaches adapted to small populations.
This structure reflects the integration of experimental and computational workflows across the project.
Why this approach is relevant for rare diseases
Rare diseases require research strategies that maximise the value of limited and heterogeneous data.
A structured pipeline helps to ensure consistency across experimental systems, enable cross-disease comparison, and connect data generation with downstream analysis and validation.
It also facilitates collaboration between disciplines, including cell biology, computational science and clinical research.
What DREAMS does and does not aim to do
DREAMS focuses on improving how research is conducted.
It does not deliver validated therapies or clinical applications within its duration. Instead, it contributes to identifying and prioritising potential targets, structuring data integration and analysis, and supporting more systematic research approaches.
Similarly, artificial intelligence is used as a complementary tool, embedded within an iterative process that combines data generation, analysis and validation.
About DREAMS
DREAMS explores how integrating biological models, multi-omics data and computational analysis can support a more structured approach to rare neuromuscular disease research.
By connecting these components within a coherent pipeline, the project contributes to improving how complex datasets are analysed and compared across diseases, helping to guide future research in a more systematic way.
