Here, we report on a new ontological resource aimed at standardizing and organizing the chemical toxicological databases in the OECD QSAR Toolbox which is free predictive toxicology software. The OECD Toolbox supports the import and export of data from the International Uniform ChemicaL Information Database (IUCLID), a system for managing data on intrinsic and hazard properties of chemical substances and registration within REACH. IUCLID implements the OECD Harmonised Templates which are standard data formats for reporting results. Even if the OECD harmonized templates provide a substantial basis for such data mapping, however, they are not well formalized, leaving much space to free text entering. The use of ontology will introduce the knowledge representation of toxicological data with the possibility of automatic reasoning, will help remove uncertainties with synonyms. Each attribute of the toxicological datasets will be associated with the ontology entry, the restriction rules and hierarchy between classes will be defined.
Olga Tcheremenskaia* (A), Romualdo Benigni (A), Cecilia Bossa (A), Monika Batke (B), Thomas Baier (B), Doris Hirmann (C) and Terry Schultz (D)
(A) Istituto Superiore di Sanità, Environment and Health Department, Viale Regina Elena 299,Rome 00161, Italy;
(B) Fraunhofer Institute for Toxicology &Experimental Medicine, Nikolai-Fuchs-Str. 1, 30625 Hannover, Germany;
(C) Computational Assessment Unit, European Chemicals Agency, Annankatu 18, P.O. Box 400, FI-00121 Helsinki, Finland;
(D) Environment, Health and Safety Division, Organisation for Economic Co-operation and Development 2, rue André Pascal – 75775 Paris
We use the DL species of the Web Ontology Language (OWL DL) supported by the Protégé OWL editor and the BioPortal import plugin.
The work started with a critical evaluation of all existing related projects. For terms collection for the carcinogenicity endpoint ontology we have re-used some ontology classes from following resources: the OECD Template 72: Carcinogenicity, OECD picklists, toxicological ontology elaborated within the OpenTox project (B. Hardy, 2010 ; O. Tcheremenskaia , R. Benigni, 2012) including the ITEM Organs system and Effects ontology, INHAND initiative (International Harmonization of Nomenclature and Diagnostic Criteria for Lesions in Rats and Mice) (J. Vahle, A. Bradley et al. 2009), some existing ontologies freely available at the Bioportal ontology depository such as SNOMED, NCI Thesaurus, Clinical Terms Version 3, Mouse pathology and Mouse adult gross anatomy. The OECD harmonised template available as xml schema have been used as the basis for the development of the ontologies.
This project started with ontology development for carcinogenicity. The main classes of the ontology are corresponding to the OECD harmonised template XML schema: Administrative data, Applicant’s summary and conclusion, Data source, Materials and Methods, Results and Discussion. For the moment, more than 5000 terms have been collected from related recourses in order to define free text fields present in the harmonised template and to collect as much synonyms as possible
For example, Species class includes terms imported from the OECD template, NCI Thesaurus, Mouse anatomy ontology and Clinical Terms ontology.
Each entry of the carcinogenicity experimental databases including in the Toolbox has been associated with the ontology using the OWL hierarchy relationships and restriction rules.
The complete ontologies will be released by the end of 2012 and will include ontologies for Carcinogenicity, Repeated Dose Toxicity and Reproductive/Developmental Toxicity. The ontology will be used within the OECD QSAR Toolbox for better integration and standardization of experimental data that will facilitate the data gap filling for the predictive toxicology approach.
This work is funded by ECHA/OECD Program: “Multiple framework contract with re-opening of competition for the provision of scientific support services – ONTOLOGY” (ECHA/2011/25)
B. Hardy et al. (2010), Collaborative Development of Predictive Toxicology Applications, Journal of Cheminformatics, 2:7; doi:10.1186/1758-2946-2-7.
O. Tcheremenskaia , R. Benigni et al. (2012) OpenTox Predictive Toxicology Framework: toxicological ontology and semantic media wiki-based OpenToxipedia. Journal of Biomedical Semantics, 2012. 3(Suppl 1): p. S7.
J. Vahle, A. Bradley et al. (2009) The International Nomenclature Project: An Update. Toxicol Pathol, August 2009; vol. 37, 5: pp. 694-69