A medical ontology is a computational representation that contains, in an explicit way, the available knowledge on a particular disease, the secondary diseases and existent relationships between them (comorbidity), and allows extracting information. This is a very powerful tool since the concepts are clearly identified and the information is independent of language or synonyms that can be used to designate the same concept.
MS is a longstanding chronic disease and, to understand it, it is necessary to integrate a lot of information coming from different sources: scientific articles, databases or medical information in clinical notes.
In the article published in Plos One, researchers used text-mining tools and analyzed in parallel anonymized data from the clinical notes of 600 patients with MS from Hospital Clínic and scientific articles related to this disease published in PubMed, the search portal of the National Library of Medicine of the United States. The concepts related to the illness and other data needed to develop the ontology were extracted.
To validate this specific ontology, researchers crossed the results of the two approaches (patient records and articles referenced in Pubmed EM) to find comorbidities. They found out that the coexisting diseases were the same. It also allowed them to discover relationships that were not described in the texts.
This medical ontology, the first one for Multiple Sclerosis, is already available for the scientific community in the repository of ontologies “BioPortal” of the National Center for Biomedical Ontology”, and will help to advance the clinical and translational research disease.
Article reference:
Malhotra A, Gündel M, Rajput AM, Mevissen HT, Saiz A, Pastor X, Lozano-Rubi R, Martinez-Lapiscina EH, Zubizarreta I, Mueller B, Kotelnikova E, Toldo L, Hofmann-Apitius M, Villoslada P.
PLoS One. 2015 Feb 9;10(2):e0116718. doi: 10.1371/journal.pone.0116718. eCollection 2015. PMID: 25665127