You are here:
Development & Use of Adverse Outcome Pathways for 21st Century Decision Support
Citation:
Edwards, S. Development & Use of Adverse Outcome Pathways for 21st Century Decision Support. International Symposium on Chemical Risk Prediction & Management and National Symposium of Ecotoxicology, Wuxi, CHINA, April 26 - 28, 2016.
Impact/Purpose:
Adverse Outcome Pathways can expand and enhance the use of ToxCast and other in vitro toxicity information by providing a mechanistic link to adverse outcomes of regulatory concern, but the expert-driven development of AOPs is labor-intensive and time-consuming. This work is intended to complement that process by creating a broad array of computationally-predicted AOPs (cpAOPs) by data mining of publicly available data. This increases the coverage of AOPs and provides minimal information in many cases where nothing is available otherwise. It also provides a starting point for expert-driven development of AOPs and thereby potentially accelerating that process.
Description:
The Adverse Outcome Pathway has emerged as an internationally harmonized mechanism for organizing biological information in a chemical agnostic manner. This construct is valuable for interpreting the results from high-throughput toxicity (HTT) assessment by providing a mechanistic connection to adverse outcomes of regulatory interest. To facilitate the development and use of AOPs, an international knowledgebase (AOPKB; https://aopkb.org) was developed to house this information with the public release of the first module (AOP-Wiki; http://aopwiki.org) in September 2014. AOP development can be viewed as a stepwise process of defining the AOP, evaluating the AOP, and quantitatively describing the AOP. The time required for the later steps is considerably greater than earlier steps, so mechanisms for crowd-sourcing AOP development and approaches for prioritizing which AOPs advance through later steps is important. The AOP development process can be further accelerated via computational approaches to create hypothetical AOPs for subsequent evaluation by experts. Early examples of the resulting computationally-predicted AOPs (cpAOPs) have been evaluated by comparing against previously-defined modes of action such as liver toxicity via CCl4 exposure and provide annotation information for >200 EPA ToxCast assays. These cpAOPs can be rapidly assessed by any scientist working in mechanistic toxicology to assemble putative AOPs. Detailed AOP development can be focused on problems where high precision is needed whereas putative AOPs may be sufficient for decisions when less quantitation and/or weight of evidence is required. By combining AOPs with chemical-specific information such as high throughput toxicity testing results and ADME predictions, we can recapitulate the MOA for a given chemical from reusable components allowing more extensive use of AOPs in risk assessment. [This is an abstract or a proposed presentation and does not necessarily reflect EPA policy. Mention of trade names or commercial products does not constitute endorsement or recommendation for use.]