Abstract
The purpose of this chapter is to follow the evolution of what has occurred over time in the ontologies published in response to the COVID-19 pandemic. Correctness and completeness of ontologies on the schema and instance level are important quality criteria in their selection for an application. To help both the librarians and the users, there is a need of a framework for the comparison of different semantic data sources in the COVID-19 pandemic. Meanwhile, online services and/or applications based on ontologies or SKOS-based COVID-19 thesauri are still rare. As an emerging technology in libraries, an all-integrating ontology for coronavirus disease knowledge and data refers to the continuing development of an existing technology. In spite of using ontologies in the Semantic Web, meanings of concepts and relationships are still largely unrealized in terms of obtaining accurate and timely information about COVID-19. But the nature of causal relationships on this disease is made accessible through ontologies as the material in which its main concepts are supported.
TopIntroduction
In clinical studies, an ontology is a human- and computer-interpretable set of terms and relations that represent entities in a specific biomedical domain and how they relate to each other. A semantic web ontology expresses a data model using an ontology web language (OWL). It provides a set of classes, properties, and restrictions that can be used to represent the datasets available to a system and secure their interoperability with other systems and under different contexts. It can also be specialized to create new classes and properties to model provenance for different applications and domains. Recently, the Coronavirus Infectious Disease Ontology (CIDO) has used text mining to identify chemical drugs and antibodies effective against at least one human coronavirus infection in vivo or in vitro. After mapping the identified anti-coronavirus drugs to ontologies (DrugBank) three drugs with most connections, dasatinib, chlorpromazine, and anosomycin can be identified in Figure 1.
Figure 1. Anti-coronavirus drugs map according to the DrugBank oncology. Three drugs with most connections are dasatinib (an inhibitor of murine leukemia), chlorpromazine (anti-coronavirus properties not fully understood), and anosomycin (antibiotic targeting ribosomes and blocking protein synthesis (Barbacid & Vazquez, 1974)). Source: (Liu et al., 2020).
In ontology engineering, a domain model is a formal representation of a knowledge domain with concepts, roles, datatypes, individuals, and rules, typically grounded in a description logic (Greenfield et al., 2020). An epidemiological data model of the COVID-19 must include its primary data domains:
Such COVID-19 data domains remain with standardization challenges that must be solved before data can be effectively integrated across domains for epidemiology studies. So in the coronavirus data domain, studying the datasets from the COVID-19 can help get an insight into the ontologies and vocabularies used in the domain. Some of the ontologies that a data publisher needs to have in mind are: the WHO COVID-19 Rapid Version CRF semantic data model (COVIDCRFRAPID) (Bonino, 2020), the COVID-19 Surveillance Ontology (COVID19) (de Lusignac et al., 2020a)(de Lusignac et al., 2020b), COviD-19 Ontology for Cases and Patient information (CODO) (Dutta & DeBellis, 2020), COVID-19 Ontology (COVID-19) (Domingo-Fernández et al., 2020), or the COVID-19 Infectious Disease Ontology (IDO-COVID-19) (Babcock et al., 2020).
A COVID-19 ontology engineering area can consider glossaries on the pandemic (https://www.btb.termiumplus.gc.ca/publications/covid19-eng.html), and open resource repositories (Hu et al., 2020), to consistently utilize clinical presentations, speciðc ðndings, responses to treatment, biomarkers, and molecular characterizations as taxonomic features—either to build into a disease classiðcation all of the indicated disease attributes (Babcock et al., 2020) or separate them as annotations with metadata, evidence, and provenance (He et al., 2020a). In the first case, COVID-19 ontologies can be used to define a strategy for constructing a taxonomy of coronavirus diseases incorporating both high-throughput genetic and molecular data as well as clinical data; in the second, COVID-19 ontologies are regarded for as community-based tools for knowledge representation and reasoning (familiar terms in artificial intelligence at work in the late 1990s biomedical resource Gene Ontology (GO)).
Key Terms in this Chapter
COVID-19 Disease: Viral respiratory sickness caused by the 2019 SARS COV-2 Coronavirus.
Coronavirus: A high-risk type of common virus that infect human, typically leading to an upper respiratory disease.
Knowledge Graph: A graphical representation tool focusing on nodes which represent the environments relevant for teams and communities to accelerate discovery.
Semantic Data Model: A method of structuring data that includes the use of ontologies to encode rich semantics in the form of machine-readable Resource Description Framework (RDF) metadata.
Scenario: A set of queries that an ontology must be able to answer, that often have the form of story problems.
Knowledge Organization System (KOS): A set of elements which can be used for describing (indexing) objects, or browsing collections, often structured and controlled; some of them are thesauri, taxonomies, classification schemes and subject heading systems.
Biomedical Ontology: A domain ontology that represents a body of explicit declarative knowledge about human biomedicine.
Biomedical Thesaurus: The reference terminology underpinning the integration of molecular and clinical related information within a unified biomedical informatics framework.
Ontology Engineering: Refers to the following sequence of activities, ontology development process, ontology life cycle, and methodologies, tools and languages required for building ontologies.