An ontology for representing hematologic malignancies: the cancer cell ontology

Lucas M Serra,Alexander D Diehl,William D Duncan

doi:10.1186/s12859-019-2722-8

Lucas M Serra, Alexander D Diehl + Show 1 more

Open Access

https://doi.org/10.1186/s12859-019-2722-8

Copy DOI

Abstract

BackgroundWithin the cancer domain, ontologies play an important role in the integration and annotation of data in order to support numerous biomedical tools and applications. This work seeks to leverage existing standards in immunophenotyping cell types found in hematologic malignancies to provide an ontological representation of them to aid in data annotation and analysis for patient data.ResultsWe have developed the Cancer Cell Ontology according to OBO Foundry principles as an extension of the Cell Ontology. We define classes in Cancer Cell Ontology by using a genus-differentia approach using logical axioms capturing the expression of cellular surface markers in order to represent types of hematologic malignancies. By adopting conventions used in the Cell Ontology, we have created human and computer-readable definitions for 300 classes of blood cancers, based on the EGIL classification system for leukemias, and relying upon additional classification approaches for multiple myelomas and other hematologic malignancies.ConclusionWe have demonstrated a proof of concept for leveraging the built-in logical axioms of the ontology in order to classify patient surface marker data into appropriate diagnostic categories. We plan to integrate our ontology into existing tools for flow cytometry data analysis to facilitate the automated diagnosis of hematologic malignancies.

Highlights

Within the cancer domain, ontologies play an important role in the integration and annotation of data in order to support numerous biomedical tools and applications
[5] As interest and support has shifted over the years, the scope of the Cell Ontology (CL) has shifted to focus primarily on vertebrate cell types with special attention to hematopoietic cell types. [6, 7] The CL links to other ontologies within the Open Biological and Biomedical Ontology (OBO) foundry via relations from the Relations
[26] Our ontology imports the entirety of the CL, which indirectly imports modules from the Protein Ontology (PRO), the Chemical Entities of Biological Interest ontology (CHEBI), the Phenotypic Quality Ontology (PATO), the Cell Line Ontology (CLO), the Relation Ontology (RO), the National Center for Biotechnology Information taxonomy (NCIT), the Uber Anatomy Ontology (UBERON) and the Gene Ontology (GO)

Summary

Introduction

Ontologies play an important role in the integration and annotation of data in order to support numerous biomedical tools and applications. [2] Briefly, ontologies at their core are semantic terminologies that exist as two types: reference ontologies, which embody established knowledge via rich, precise meanings for terms in a domain, and application ontologies, which are designed for a specific purpose and weave together sets of related classes from reference ontologies in order to represent the entities of complex domains. Technically a terminology with ontology-like features, the NCI Thesaurus covers 110,000 terms in 36,000 concepts in the cancer research domain and arose from a need to integrate varied data systems through a unified coding system. Technically a terminology with ontology-like features, the NCI Thesaurus covers 110,000 terms in 36,000 concepts in the cancer research domain and arose from a need to integrate varied data systems through a unified coding system. [21, 22]

Methods

Results

Discussion

Conclusion