You are here:
Building Scientific Confidence in the Development and Evaluation of Read-Across - GenRA: Evaluating local validity for read-across prediction using chemical and biological information (SOT/QSAR conference)
Citation:
Patlewicz, G., J. Liu, R. Judson, R. Thomas, AND I. Shah. Building Scientific Confidence in the Development and Evaluation of Read-Across - GenRA: Evaluating local validity for read-across prediction using chemical and biological information (SOT/QSAR conference). Presented at 17th International Conference on QSAR in Environmental and Health Sciences, Miami, FL, June 13 - 17, 2016. https://doi.org/10.23645/epacomptox.5067529
Impact/Purpose:
Presentation at OSAR conference on Developing a systematic and objective read-across approach
Description:
Read-across remains a popular data gap filling technique within category and analogue approaches for regulatory purposes. Acceptance of read-across is an ongoing challenge with several efforts underway for identifying and addressing uncertainties. Here we demonstrate an algorithmic approach to facilitate read-across using ToxCast in vitro bioactivity data in conjunction with chemical descriptor information to predict in vivo outcomes in guideline testing studies from ToxRefDB. Over 3400 different chemical structure descriptors were generated for a set of 976 chemicals and supplemented with the outcomes from 821 in vitro assays. The read-across prediction for a given chemical was based on the similarity weighted endpoint outcomes of its nearest neighbors calculated using in vitro bioactivity and chemical structure descriptors, called GenRA. GenRA is based on a computational approach for: (i) defining local validity domains using chemical and bioactivity descriptors, (ii) systematically deriving endpoint read-across predictions within these domains using similarity weighted activity of nearest neighbours, (iii) objectively evaluating predicted performance using tested chemicals, and (iv) assigning read-across predictions to untested chemicals along with estimates of uncertainty. We found in vitro bioactivity descriptors were often found to be more predictive of in vivo toxicity outcomes than chemical structure descriptors. We believe GenRA is an important first step in systematizing read-across prediction of toxicity and serves as a useful tool as part of a screening level hazard assessment for new untested chemicals. This abstract does not reflect EPA policy.
URLs/Downloads:
DOI: Building Scientific Confidence in the Development and Evaluation of Read-Across - GenRA: Evaluating local validity for read-across prediction using chemical and biological information (SOT/QSAR conference)
QSAR 2016_READ-ACROSS_PATLEWICZ_090616 KMC_FINAL.PDF (PDF, NA pp, 2197.928 KB, about PDF)
SHAH_GENRA-QSAR-2016 ABSTRACT.PDF (PDF, NA pp, 100.764 KB, about PDF)