The CoMPaSS-NMD Atlas Platform: A Secure Multimodal Infrastructure for Harmonized Data Collection, Exploration, and Future Artificial Intelligence-Driven Stratification in Hereditary Neuromuscular Disorders.
Authors
Affiliations (11)
Affiliations (11)
- FINCONS GROUP AG, Lugano, Switzerland.
- Department of Data Science and Engineering, Silesian University of Technology, Gliwice, Poland.
- Department of Biomedical, Metabolic and Neural Sciences, University of Modena and Reggio Emilia, Modena, Italy.
- Friedrich-Baur-Institute at the Department of Neurology, LMU University Hospital, LMU Medizin, Ludwig-Maximilians-Universität München, Munich, Germany.
- John Walton Muscular Dystrophy Research Centre, Translational and Clinical Research Institute, Newcastle University, and Newcastle Upon Tyne Hospitals NHS Foundation Trust, Newcastle upon Tyne, UK.
- TECNALIA, Basque Research and Technology Alliance (BRTA), Donostia-San Sebastián, Gipuzkoa, Spain.
- Folkhälsan Research Center, Helsinki, Finland.
- Deep Blue S.r.l., Rome, Italy.
- Department of Translational Medicine, IGBMC (Institut de Génétique et de Biologie Moléculaire et Cellulaire), Inserm U1258, CNRS UMR7104, University of Strasbourg, Illkirch, France.
- CeGaT GmbH, Tübingen, Germany.
- Molecular Medicine and Neurogenetics Unit, IRCCS Fondazione Stella Maris, Pisa, Italy.
Abstract
Hereditary neuromuscular disorders exhibit marked phenotypic and molecular heterogeneity, which weakens genotype-phenotype correlation and complicates timely diagnosis, prognosis, and patient stratification for clinical trials. The Horizon Europe project CoMPaSS-NMD aims to develop artificial-intelligence-based stratification tools by integrating clinical, genetic, histopathological, and imaging data from clinical reference centers across Europe and the United Kingdom. We developed CoMPaSS-NMD Atlas, a secure, modular web platform for pseudonymized data collection, multimodal data management, and controlled exploration. Clinical, technical, and data-management experts defined requirements; contributing centers co-designed electronic clinical and histopathology forms. Clinical phenotypes used the Human Phenotype Ontology (HPO), while imaging, histopathology, and sequencing data followed international standards. The platform supports role- and attribute-based access, pseudonymized patient registration, metadata management, structured clinical data entry, MRI and histopathological uploads, and a dedicated genetic-sequencing workflow. Cohort-level exploration is driven by structured metadata and HPO terms, enabling hierarchical phenotype navigation, ontology-aware data navigation, and cross-domain queries spanning clinical, genetic, imaging, and histopathological features. Users can download raw files in their original formats, along with structured metadata reports in CSV or PDF, according to the permissions defined for data protection. The Atlas provides a foundation for harmonized, standardized prospective data collection and for building large datasets for artificial intelligence- and machine learning-based patient stratification. With secure governance, traceability, and multimodal integration, it is designed to support diagnostic and stratification research in hereditary neuromuscular disorders; clinical validation is ongoing.