Phenotype Ontologies


Group Logo

A medical phenotype terminology is a catalog of specific signs, symptoms, imaging findings, and other abnormalities seen in medical practice. Individual items (e.g. Fever or Hepatosplenomegaly) are referred to as terms. An ontology is a terminology that additionally provides computational definitions that specify the meaning of the terms as well as their interrelationships, forming a foundation for computational searching and analysis. Whereas a disease ontology (also known as nosology) describes disease entities such as Marfan syndrome, phenotype ontologies describe the individual manifestations of diseases, which in the case of Marfan syndrome might include Ectopia lentis, Aortic dilatation, and Scoliosis.

This project of GA4GH aims to bring together existing international efforts to develop and promote standard language and tools for recording patient clinical phenotypes and exploiting phenotype data for diagnostics and translational research.

Currently numerous terminologies and ontologies are in use, reflecting the disparate needs of different communities involved in translational research and patient care in many fields of medicine. The Rare Disease Research Consortium has recently developed a set of recommendations surrounding disease and phenotype ontologies for rare disease, and suggests the use of the Human Phenotype Ontology for the annotation of phenotypic abnormalities of rare diseases, recognizing that other resources may be suitable in certain situations. IRDiRC has supported the International Consortium for Human Phenotype Terminologies (ICHPT), which was created with input from several groups including primarily Orphanet, the Human Phenotype Ontology team, and the Online Mendelian Inheritance in Man (OMIM) team. ICHPT has provided a mapping between the core set of terms in HPO, PhenoDB, Orphanet, Elements of Morphology, POSSUM, SNOMED, MeSH, and MeDRA. Additionally, the HPO is currently being incorporated into the UMLS (United Medical Language System). Additional IRDiRC recommendations included continuing to ensure interoperability with ORDO, the Orphanet Rare Disease Ontology, and OMIM, Online Inheritance in Man, which are two complementary rare disease nosologies.

GA4GH intends to bring together global stakeholders in the area of phenotype terminologies and ontologies from other areas of medicine including oncology, and in the future common (complex) diseases and pharmacogenomics. Amongst the terminologies of interest in the domain of cancer are the International Classification of Diseases for Oncology (ICD-O), SNOMED CT, the National Cancer Institute’s Thesaurus, and the resources of the National Cancer Informatics Program. The HPO additionally provides a limited coverage of cancer with focus on rare diseases. Remaining challenges include not only integration and development of phenotype ontology resources but also of frameworks and standards for exchange of cancer phenotype data, agreement on the range of data to be captured (e.g., pathology, biomarkers, imaging, other clinical features), and the development of bioinformatics algorithms to optimally include phenotypic aspects into the analysis of genomic data that is now beginning to become standard in the care of cancer patients.