Date of Award

5-1-2015

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Information Science

First Advisor

Xiaowei Xu

Second Advisor

Imran Shah

Abstract

Over 10,000 chemicals are currently in commercial use, and hundreds of more chemicals are introduced each year. And humans are exposed to over 6,000 environmental chemicals. But only a small fraction of these chemicals have been adequately assessed for potential hazard, leading to the need to prioritize the chemicals for targeted toxicity testing. It is time-consuming and resource intensive to determine a single chemical’s potential for toxicity using animal testing. This approach is impractical for evaluating tens of thousands of chemicals. Hence, there is an urgent need to find alternatives to animal testing and a number of international efforts are underway to assess the utility of in vitro assays in addressing this challenge. EPA’s Toxicity Forecaster (ToxCast) Database and the Toxicity Reference Database (ToxRefDB) include a large number of high-throughput screening (HTS) in vitro data and in vivo guideline studies, provide a novel resource for profiling and building predictive models of chemical toxicity. This research integrated available data from diverse experimental methods, including in vitro high-throughput screening molecular assays (ToxCast), in vivo guideline animal studies (ToxRefDB), and chemical structure information (QSAR) to develop, evaluate, and use machine learning for prediction of toxicity. Tools produced by this research will enable rapid prioritization of thousands of new environmental chemicals for further testing and risk assessment. Disclaimer: The views expressed in this paper are those of the authors and do not necessarily reflect the views or policies of the U.S. Environmental Protection Agency and U.S. Food and Drug Administration. Mention of trade names or commercial products does not constitute endorsement or recommendation for use.

Download

Share

COinS