Providing the Missing Link: The Exposure Ontology ExO

1. Providing the Missing Link: The Exposure Ontology ExO RASS December 11, 2013 Elaine Cohen HubalChemical Safety for Sustainability Research Program Disclaimer. Although this work was reviewed by EPA and approved for presentation, it may not necessarily reflect official Agency policy.

2. What Is Exposure Science? • The bridge between the sources of chemical, physical and biological agents and human health • Provides crucial information to estimate real-life risks to health and to identify the most effective ways to prevent andreduce these risks. www.isesweb.org

3. TRANSPORT, TRANSFORMATION, and FATE PROCESS MODELS Exposure for Risk Evaluation: Approaches SOURCE / STRESSOR FORMATION ADVERSE OUTCOME • Questionnaire based metrics (epidemiology) • Surrogate exposure metrics (ambient measures) • Exposure measurement (direct or point-of-contact) • Biomonitoring (NHANES) • Modeled estimates (indirect or scenario evaluation) EXPOSURE MODELS TRANSPORT/ TRANSFORMATION DOSE PBPK MODELS ENVIRONMENTAL CHARACTERIZATION EXPOSURE ACTIVITY PATTERN • Individual • Community • Population 2

4. Biological Receptor Stressor Outcome Perturbation Perturbation Environmental Source Ambient Exposure Population Disease Incidence/Prevalence Environmental Source Personal Exposure Disease State (Changes to Health Status) Individual Internal Exposure (Tissue Dose) Dynamic Tissue Changes (Tissue Injury) Tissue Dynamic Cell Changes (Alteration in Cell Division, Cell Death) Dose to Cell Cell Dose of Stressor Molecules Dynamic Changes in Intracellular Processes Biological Molecules Systems Biology: Exposure at All Levels of Biological Organization Cohen Hubal, JESEE, 2008

5. Susceptibility (Genetic Variants / Epigentic Modifications) Exposure for Translation Biological Insight (Toxicity Pathways) Environmental Factors (Exposure) Improved Measures of Individual Etiological Processes and Individual Exposures Key Perturbations Key Targets Personal Risk Profile Biomarkers Indicators Metrics Information Extrapolation for Risk Assessment Education Personal Risk Management Public Health Policy Prevention Cohen Hubal, et al. JTEH, 2010

6. Knowledge Systems – Enabling Hypothesis Development • Computational Techniques – Two Branches A combination of discovery and engineering (mechanistic)-based modeling approaches required for hypothesis development and testing • Knowledge-discovery • Data-collection, mining, and analysis • Required to extract information from extant data on critical exposure determinants, link exposure information with toxicity data, and identify limitations and gaps in exposure data. • Mechanistic (dynamic) simulation • Mathematical modeling at various levels of detail • Required to model the human-environment system and to test our understanding of this system.

7. Exposure-Hazard Knowledge System • Translation of HTP hazard information requires holistic risk assessment knowledge system • Include ontologies, databases, linkages • Facilitate computerized collection, organization, and retrieval of exposure, hazard, and susceptibility information • Define relationships, allow automated reasoning, facilitate meta analyses • Standardized exposure ontologies required to • Develop biologically-relevant exposure metrics • Design and interpret in vitro toxicity tests • Incorporate information on susceptibility and background exposures to assess individual and population-level risks

8. Schematic of ontologies, databases and ontology/database linkages needed for the efficient development of a Foods-for-Health Knowledge System MC Lange, et al. (2007) A multi-ontology framework to guide agriculture and food towards diet and health. J Sci Food and Ag 87(8)1427-34.

9. Exposure Data Sources http://actor.epa.gov Peter Egeghy, NERL

10. Exposure Data Landscape Network of exposure taxonomy used in ACToR; Egeghy et al, 2011

11. Exposure Data Landscape Number of unique chemicals by data type in ACToR; Egeghy et al, 2011

12. Exposure Data Collection and Access: ACToR, Aggregated Computational Toxicology Resource http://actor.epa.gov/ ACToR API Chemical ID & Structure QC, Inventory Tracking DSSTox Chemical Chemical ID, Structure ToxRefDB ACToR Core ToxCastDB ExpoCastDB Tabular Data, Links to Web Resources In Vivo Study Data - OPP ToxCast Data – NCCT, ORD, Collaborators Exposure Data – NERL, NCCT Internet Searches

13. Exposure Data Collection and Access :ExpoCastDB Goals • Consolidate observational human exposure data, improve access and provide links to health related data • House measurements from human exposure studies • Encourage standardized reporting of observational exposure information • Provide separate interface with inner workings of ACToR • Facilitate linkages with toxicity data, environmental fate data, chemical manufacture information • Provide basic user functions • Visualization (e.g., scatterplots, probability plots, goodness-of-fit) • Obtain summary statistics and estimate distributional parameters • Download customized datasets http://actor.epa.gov/

14. Exposure Data Collection and Access: ExpoCastDB Generic_chemical table in ACToR • Four initial studies from National Exposure Research Laboratory • Full raw data sets available for download • Browse data capability • Descriptive statistics capabilities 1 Laboratory method 1 N Technique / sampling method Measure (Chemical) N N N N 1 Location_CV (cont vocab.) N N Medium (serum, air, soil, etc.) 1 1 N Sample N N N N Location N N N N N Exposure Taxonomy (Assay_ category_ CV table in ACToR) N N N 1 N Subject N N N N N N N Study N N Sources N Study_CV cont. vocab 1 N N

15. ExO: An Ontology for Exposure Science

16. Exposure Data Collection and Access: Design and Evaluation of the Exposure Ontology, ExO Background: • Significant progress has been made in collecting and improving access to genomic, toxicology, and health data • These information resources lack exposure data required to • translate molecular insights • elucidate environmental contributions to diseases • assess human health risks at the individual and population levels Aim: • Facilitate centralization and integration of exposure data to inform understanding of environmental health • Bridge gap between exposure science and other environmental health disciplines Vehicle: • Carolyn Mattingly, Mount Desert Island Biological Laboratory • LRI seed funding, followed by NIEHS RO1

17. Ontologies • An ontology is a formal representation of knowledge within a domain and typically consists of classes, the properties of those classes, and the relationships between these • Many fields are developing ontologies to • Organizing and analyzing large amounts of complex information from multiple scientific disciplines • Provide unprecedented perspective • Enable more informed hypothesis development (Gruber, Int. J. Human-Computer Studies, 1995) (http://www.obofoundry.org/)

18. Design and Evaluation of the Exposure Ontology: ExO • Develop an exposure ontology consistent with those being used in toxicology and other health sciences • Facilitate centralization and integration of exposure data to inform understanding of environmental health • Bridge gap between exposure science and other environmental health disciplines • Initially focus development on human exposure to chemicals • Ultimately, provide domains that can be extended to encompass exposure data for the full range of receptors and stressors

19. Exposure Ontology Working Group

20. Phase II Phase III Phase IV Phase I Phases of Exposure Ontology Development Disseminate exposure ontology for public feedback Initial pilot curation to identify major concepts Full working group evaluation of draft ontology Model relationships among data concepts Expand test data set to evaluate extensibility of conceptual model and cross -reference existing ontologies Iterate data model refinement and curation

21. Definitions of Central Concepts • Exposure Stressor - An agent, stimulus, activity, or event that causes stress or tension on an organism and interacts with an exposure receptor during an exposure event. • Exposure Receptor - An entity (e.g., a human, human population, or a human organ) that interacts with an exposure stressor during an exposure event. • Exposure Event - An interaction between an exposure stressor and an exposure receptor. • Exposure Outcome - Entity that results from the interaction between an exposure receptor and an exposure stressor during an exposure event. Mattingly et al, submitted

22. Biolog. Agent Relational View of Selected ExO Domains Public Policy Chem. Agent • Source • Location • Process • Transport Path Inter- vention Exposure Outcome Exposure Stressor Biomech. Agent Disease Biolog. Response Phys. Agent Symptom Exposure Event Psychosoc. Agent Molecular Response Exposure Receptor Human Pop. Anthro- sphere • Location • Temporal Pattern • Intensity • Route • Assay • Medium • Method • Location Individual • Location • Genetic Background • Lifestage • Health Status • Socioeconomic Status • Occupation Mattingly et al, submitted

23. High-level schematic of Exposure Ontology (ExO) integration within a broader biological context. Encode Annotated with Biological Process Molecular Function Cellular Component (e.g., Gene Ontology Chemical (e.g., MeSH) Genes Gene Products Interacts with (e.g., CTD) Is a Interact via Exposure Stressor (e.g., ExO) Pathways, Networks Reactions (e.g., KEGG, Reactome) Interacts with Occur within Interacts with (e.g., CTD) Exposure Receptor (e.g., ExO) Biological System (e.g., Functional model of anatomy) Is a Assessed by Via an Exposure Event (e.g., ExO, ExpoCastDB) Exposure Outcome Phenotype (e.g., OMIM, MeSH) Results in an Mattingly et al, submitted

24. Next Steps • Open source approach • With input from the scientific community, further specify branches • Leverage existing ontologies (e.g., CHEBI and MeSH for “Chemical agent” Stressors; DO, OMIM and MeSH for “Disease” Outcomes). • Cross-referencing will underscore where ExO fits into a broader knowledge space and where it may add value to existing ontologies.

25. Chemicals Exposure Data (curated and public sources) chemical-gene interactions chemical-disease relationships Genes Diseases gene-disease relationships functional annotations Beyond EPA: Pilot Curation of Exposure Data into CTD Carolyn Mattingly pathway data

26. Acknowledgements ExpoCastDB -- Richard Judson, Peter Egeghy, Sumit Gangwal • ExO– • Carolyn Mattingly, • Tom McKone, • Judy Blake, • Mike Callahan