Machine learning–based validation of Human Phenotype Ontology (HPO) terms for systemic autoinflammatory diseases using the Eurofever Registry: the ODINO project
6 affiliations
- 1Rheumatology and Autoinflammatory Diseases, IRCCS Istituto Giannina Gaslini, Genoa, Italy
- 2Clinical Bioinformatics Unit, IRCCS Istituto Giannina Gaslini, Genoa, Italy
- 3Department of Pharmacy and Biotechnology, University of Bologna, Bologna, Italy
- 4Department of Genetics, University of Groningen, University Medical Center Groningen, Groningen, Netherlands
- 5Department of Molecular Genetics and Cytogenomics, CHU Montpellier, CeRéMAIA, IRMB, Univ Montpellier, INSERM, Montpellier, France
- 6Gaslini Trial Center, IRCCS Istituto Giannina Gaslini, Genova, Italy
Abstract
Background: The Human Phenotype Ontology (HPO) terms for systemic autoinflammatory diseases (SAIDs) were revised in 2022, but their diagnostic accuracy has not yet been validated in real-life cohorts.
Objectives: To evaluate the diagnostic accuracy of HPO terms in a real-life cohort of patients enrolled in a large international registry for SAIDs through a machine learning (ML) approach.
Methods: Clinical and genetic data were extracted from the Eurofever Registry (EF) for 2,576 patients. Due to limited sample sizes for rarer conditions, analyses focused on FMF, MKD, TRAPS, CAPS and PFAPA. The 223 clinical EF variables were codified into HPO terms, including terms with missing HPO correspondence. Supervised ML models (Elastic Net, k-Nearest Neighbors, Random Forest, and XGBoost) were trained to classify the five diseases using both multiclass and one-vs-rest (OVR) strategies.
Results: A full HPO correspondence was found for 185 EF terms. Clinically relevant variables frequently missing from HPO, including “ancestry” and “fever length”, were incorporated into the final models. On the held-out test set, the XGBoost OVR framework achieved the following recall values: 87.2% for FMF, 90.4% for PFAPA, 81.1% for CAPS, 55.7% for TRAPS, and 55.6% for MKD. Feature importance analyses highlighted “fever length” as key discriminative variable. Error analysis revealed an enrichment of patients carrying the R92Q TNFRSF1A variant among false-negative predictions for TRAPS. Unsupervised analysis showed no clear separation between FMF patients carrying one versus two pathogenic MEFV mutations.
Conclusions: HPO-based phenotypic modeling supports accurate classification of the 5 most frequent recurrent fevers included in this study. The HPO database should be enriched by including the missing terms, especially “fever length”, “ancestry”, and the symptoms frequencies according to Eurofever. An ontology-driven diagnostic tool applicable to both clinical research and computational settings for recurrent fevers is now available.
Related
- LitGene: an interpretable transformer model integrating text and ontology for gene representation learning
- A Mechanistic Framework for Somatic Mosaicism in Autoinflammatory Diseases Beyond Mendelian Inheritance
- Risk Stratification and Patient Trajectories in CAPS: The Important Role of the Autoinflammatory Disease Damage Index
- MAOMAO: an ontology-guided FAIR resource for harmonized peptide toxicity data
- Deep learning coupled with scalable domain-specific structural validation expands RNA virus discovery from metatranscriptomes.
BibTeX
@article{matuccicerinic2026machine,
title = {{Machine learning–based validation of Human Phenotype Ontology (HPO) terms for systemic autoinflammatory diseases using the Eurofever Registry: the ODINO project}},
author = {Caterina Matucci-Cerinic and Giacomo Cavalca and Matteo Vergani and Roberta Caorsi and Marielle van Gjin and Guilaine Boursier and Eugenia Mosci and Giovanni Fiorito and Paolo Uva and Marco Gattorno},
journal = {Frontiers in Immunology},
year = {2026},
month = oct,
doi = {10.3389/fimmu.2026.1936099},
url = {https://doi.org/10.3389/fimmu.2026.1936099},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Machine learning–based validation of Human Phenotype Ontology (HPO) terms for systemic autoinflammatory diseases using the Eurofever Registry: the ODINO project
AU - Caterina Matucci-Cerinic
AU - Giacomo Cavalca
AU - Matteo Vergani
AU - Roberta Caorsi
AU - Marielle van Gjin
AU - Guilaine Boursier
AU - Eugenia Mosci
AU - Giovanni Fiorito
AU - Paolo Uva
AU - Marco Gattorno
PY - 2026
DA - 2026/10/12
JO - Frontiers in Immunology
DO - 10.3389/fimmu.2026.1936099
UR - https://doi.org/10.3389/fimmu.2026.1936099
AB - Background: The Human Phenotype Ontology (HPO) terms for systemic autoinflammatory diseases (SAIDs) were revised in 2022, but their diagnostic accuracy has not yet been validated in real-life cohorts. Objectives: To evaluate the diagnostic accuracy of HPO terms in a real-life cohort of patients enrolled in a large international registry for SAIDs through a machine learning (ML) approach. Methods: Clinical and genetic data were extracted from the Eurofever Registry (EF) for 2,576 patients. Due to limited sample sizes for rarer conditions, analyses focused on FMF, MKD, TRAPS, CAPS and PFAPA. The 223 clinical EF variables were codified into HPO terms, including terms with missing HPO correspondence. Supervised ML models (Elastic Net, k-Nearest Neighbors, Random Forest, and XGBoost) were trained to classify the five diseases using both multiclass and one-vs-rest (OVR) strategies. Results: A full HPO correspondence was found for 185 EF terms. Clinically relevant variables frequently missing from HPO, including “ancestry” and “fever length”, were incorporated into the final models. On the held-out test set, the XGBoost OVR framework achieved the following recall values: 87.2% for FMF, 90.4% for PFAPA, 81.1% for CAPS, 55.7% for TRAPS, and 55.6% for MKD. Feature importance analyses highlighted “fever length” as key discriminative variable. Error analysis revealed an enrichment of patients carrying the R92Q TNFRSF1A variant among false-negative predictions for TRAPS. Unsupervised analysis showed no clear separation between FMF patients carrying one versus two pathogenic MEFV mutations. Conclusions: HPO-based phenotypic modeling supports accurate classification of the 5 most frequent recurrent fevers included in this study. The HPO database should be enriched by including the missing terms, especially “fever length”, “ancestry”, and the symptoms frequencies according to Eurofever. An ontology-driven diagnostic tool applicable to both clinical research and computational settings for recurrent fevers is now available.
SN - 1664-3224
ER -