You are here:
Framework for computationally-predicted AOPs
Citation:
Bell, S. AND S. Edwards. Framework for computationally-predicted AOPs. Presented at Research to Regulations, RTP, NC, September 03 - 05, 2014.
Impact/Purpose:
To be presented at meeting Adverse Outcome Pathways: Research to Regulations - September, 2014.
Description:
Framework for computationally-predicted AOPs Given that there are a vast number of existing and new chemicals in the commercial pipeline, emphasis is placed on developing high throughput screening (HTS) methods for hazard prediction. Adverse Outcome Pathways (AOPs) represent an ideal framework for connecting molecular initiating events to key events measured via HTS data with adverse outcomes of regulatory importance. However, traditional AOP development is labor intensive and time consuming. We present a graph-based workflow, cpAOP-net, that enables data integration across multiple data types to identify computationally-predicted AOPs (cpAOPs). Using carbon tetrachloride (CCL4) as an example, we illustrate how using cpAOP-net a large toxicogenomic dataset can be leveraged and phenotypic traits incorporated to generate a cpAOP which closely mirrors the hypothesized mode of action based on EPA’s IRIS assessment of CCL4. This framework for cpAOP identification has the potential to generate and rank hypothetical AOPs for expert evaluation and provide data-based scaffolds to improve the efficiency of AOP development. We discuss how cpAOP-net provides a mechanism for identifying HTS targets and the role of domain experts and confirmatory studies in model improvement. The views expressed in this abstract are those of the authors and do not necessarily reflect the views or policies of the U.S. Environmental Protection Agency.