Citation:

Angrish, M., G. Woodall, S. Watford, AND P. Whaley. Role of Semantics, Ontologies, and Adverse Outcome Pathways as a Point of Integration in Chemical Assessments-Poster. NAS Evidence Integration, Washington, DC, June 03 - 04, 2019.

Impact/Purpose:

The work in this poster is an example of how literature based information identified by systematic review and evidence mapping methods can be integrated using an adverse outcome pathway framework. Specifically, information on specific topics (i.e., thyroid outcomes) can be used to bound the literature and information extracted from the information for a data-driven approach to Adverse Outcome Pathway (AOP) development and evidence integration.

Description:

The quality and utility of literature based chemical assessments has been improved by leveraging the power of systematic review (SR) and systematic mapping (SM, also referred to as evidence mapping) approaches to aggregating and evaluating evidence of health risks posed by exposure to environmental chemicals. Taking maximal advantage of SRs and SMs is currently impeded by linguistic inconsistencies resulting from different communities using different vocabularies to describe common study characteristics, requiring the systematic reviewer to anticipate all the concepts, relationships, and words related to a science question when developing a search string sensitive enough to locate all potentially relevant studies. The state-of-the-art approach, to use dictionaries and thesauruses are useful for ensuring all semantically related terms are included in a search, but they do not offer the context necessary to capture relationships between concepts, e.g. according to biological organization such as gene expression. We are therefore exploring the use of ontologies and semantic mapping as a part of evidence integration in literature based chemical assessments. An ontology is a controlled vocabulary of precisely-defined terms and the specified relationships between them, interpretable by both humans and machines. Here we give an example of how thyroid health outcome data extracted from human and animal literature studies can be matched to ontology concepts that serve as a point of integration in a semantic framework bounded by a structured Adverse Outcome Pathway (AOP) framework. When implemented, this ontological approach may solve the problem of a researcher needing perfect a priori knowledge of all relevant terms and relationships in order to query a database for comprehensive information about mechanisms of thyroid toxicity: this information is already provided in the database ontology.

Record Details: