Generating a focused view of disease ontology cancer terms for pan-cancer data integration and analysis.

T.-J Wu,S Ward,R Finney,S Srivastava,Y Pan,H Kincaid,E Mitraka,D J Crichton,W A Kibbe,Y Hu,Q.-R Chen,R Mazumder,L M Schriml,C Yan,K M Smith,M Colbert,D Meerzaman

doi:10.1093/database/bav032

T.-J Wu, S Ward + Show 15 more

Open Access

https://doi.org/10.1093/database/bav032

Copy DOI

Abstract

Bio-ontologies provide terminologies for the scientific community to describe biomedical entities in a standardized manner. There are multiple initiatives that are developing biomedical terminologies for the purpose of providing better annotation, data integration and mining capabilities. Terminology resources devised for multiple purposes inherently diverge in content and structure. A major issue of biomedical data integration is the development of overlapping terms, ambiguous classifications and inconsistencies represented across databases and publications. The disease ontology (DO) was developed over the past decade to address data integration, standardization and annotation issues for human disease data. We have established a DO cancer project to be a focused view of cancer terms within the DO. The DO cancer project mapped 386 cancer terms from the Catalogue of Somatic Mutations in Cancer (COSMIC), The Cancer Genome Atlas (TCGA), International Cancer Genome Consortium, Therapeutically Applicable Research to Generate Effective Treatments, Integrative Oncogenomics and the Early Detection Research Network into a cohesive set of 187 DO terms represented by 63 top-level DO cancer terms. For example, the COSMIC term ‘kidney, NS, carcinoma, clear_cell_renal_cell_carcinoma’ and TCGA term ‘Kidney renal clear cell carcinoma’ were both grouped to the term ‘Disease Ontology Identification (DOID):4467 / renal clear cell carcinoma’ which was mapped to the TopNodes_DOcancerslim term ‘DOID:263 / kidney cancer’. Mapping of diverse cancer terms to DO and the use of top level terms (DO slims) will enable pan-cancer analysis across datasets generated from any of the cancer term sources where pan-cancer means including or relating to all or multiple types of cancer. The terms can be browsed from the DO web site (http://www.disease-ontology.org) and downloaded from the DO’s Apache Subversion or GitHub repositories.Database URL: http://www.disease-ontology.org

Highlights

Cancer is among the leading causes of deaths worldwide and accounts for 8.2 million deaths annually [1]
The manual process of identifying the type of cancer term involved investigation of each cancer term to identify the current classification of each type of cancer from authoritative resources including primary publications, the National Cancer Institute EDRN (NCI) Dictionary of Cancer terms, NCI cancer topics and World Health Organization (WHO) classification
In Step 3 we identified if the cancer term existed in disease ontology (DO) or if the term could be mapped to a synonym of a DO term or if the term was a novel term and should be added to DO

Summary

Introduction

Cancer is among the leading causes of deaths worldwide and accounts for 8.2 million deaths annually [1]. The term cancer describes a group of diseases in which cells develop abnormal abilities to grow rapidly without limitation, invade adjoining tissues and metastasize to other parts of the body through the lymph and blood systems [2, 3]. Cancers arise from one abnormal cell that goes through multiple transformations from normal to malignant states. Non-cancerous cells grow and divide into additional cells and undergo a limited number of division cycles in a controlled manner [4, 5]. As a result of these mutations, cells proliferate without restraint, leading to tumor formation and variety of cancers [4]

Methods

Results

Conclusion