A new artificial intelligence model that could significantly enhance disease classification and treatment strategies, especially for rare and orphan diseases, has been developed by researchers at the Hebrew University of Jerusalem.
According to the research published in the Journal of Biomedical Informatics, the model achieved an impressive 89.4 percent ROC Area (Receiver Operating Characteristic Curve) resulting in the identification of 515 disease candidates predicted to possess previously unannotated subtypes.
The ROC curve is used to assess a test’s overall diagnostic performance and compare two or more diagnostic tests used to test the absence or presence of a disease.
Distinguishing diseases into distinct subtypes is crucial for study and effective treatment strategies. Nevertheless, many disease annotations are incomplete, requiring laborious expert medical input. This challenge is especially pronounced for rare and orphan diseases, where resources are scarce.
Utilizing the extensive database of approximately 23,000 diseases documented in the Open Targets Platform, the researchers derived new capabilities to predict diseases with subtypes using direct evidence.
The Open Targets Platform (OT) integrates biomedical, genetic, and biochemical datasets, including 23,000 diseases, to empower disease ontologies, classifications, and potential gene targets.
The study was led by PhD student Dan Ofer and Prof. Michal Linial from the Department of Biological Chemistry at The Life Science Institute at Hebrew University.
“This project demonstrates the incredible potential of machine learning in expanding our understanding of complex diseases,” said Ofer. “By leveraging advanced models, we can uncover patterns and subtypes that were previously hidden, ultimately contributing to more precise and personalized treatments.”
The methodology enables a robust and scalable approach for improving knowledge-based annotations and provides a comprehensive assessment of disease ontology tiers, the researchers said.
“We are excited about the potential of our machine learning approach to revolutionize disease classification,” said Prof. Linial. “Our findings can significantly contribute to personalized medicine, offering new avenues for therapeutic development.”
The research paper titled “
- Automated annotation of disease subtypes
” can be accessed by clicking here.
The research was partially supported by ISF grant 2753/20 (M.L), The Milgrom Family Support Fund of the Hebrew University of Jerusalem 3015004508.
