Desiderata for an ontology of diseases for the annotation of biological datasets.

Olivier Bodenreider,Anita Burgun

doi:10.1038/npre.2009.3531.1

Olivier Bodenreider, Anita Burgun

Open Access

https://doi.org/10.1038/npre.2009.3531.1

Copy DOI

Abstract

AbstractThere is a plethora of disease ontologies available, all potentially useful for the annotation of biological datasets. We define seven desirable features for such ontologies and examine whether or not these features are supported by eleven disease ontologies. The four ontologies most closely aligned with our desiderata are Disease Ontology, SNOMED CT, NCI thesaurus and UMLS.

Highlights

Ontologies have been developed for the annotation of biological datasets from multiple perspectives including functional annotation of gene products (Gene Ontology), molecular sequences (Sequence ontology) and phenotypes (Mammalian Phenotype Ontology, Phenotypic Quality Ontology)
The objective of this study is to propose a list of desirable features for an ontology of diseases suitable for the annotation of biological datasets, and to analyze a list of candidate terminologies through the framework provided by these features
Support for the desirable features ranges from 32% (OMIM, Logical Observation Identifiers Names and Codes (LOINC)) to 68% (NCIt, Unified Medical Language System (UMLS))

Summary

Introduction

Ontologies have been developed for the annotation of biological datasets from multiple perspectives including functional annotation of gene products (Gene Ontology), molecular sequences (Sequence ontology) and phenotypes (Mammalian Phenotype Ontology, Phenotypic Quality Ontology). The NCI Thesaurus was developed for the annotation of cancer research and includes many diseases, but its focus on cancer can be a limitation for use in other domains. Terminologies have been long been developed for the purpose of annotating clinical records, including the International Classification of Diseases (ICD) and SNOMED CT. These terminologies have not been widely adopted by biomedical researchers for annotating disease entities in biological datasets. Neither terminology is free of intellectual property restrictions and a license or fee may be required for their use, which represents a limiting factor

Objectives

Methods

Discussion

Conclusion