You are here:
An evaluation of selected (Q)SARs/expert systems for the Prediction of Skin Sensitization Potential (WC10)
Citation:
Fitzpatrick, J. AND G. Patlewicz. An evaluation of selected (Q)SARs/expert systems for the Prediction of Skin Sensitization Potential (WC10). Presented at WC10, Seattle, WA, August 21 - 24, 2017.
Impact/Purpose:
Poster presentation at the 10th World Congress on Alternatives to Animal Testing.
Description:
One of the alternative approaches to assessing skin sensitization hazard is through the use of (Q)SARs/expert systems. Here we evaluate the predictive performance of two expert systems (TIMES-SS and VEGA) and two SAR rulebases (OASIS protein binding alerts and Toxtree’s reactivity domains). To evaluate them, a dataset of 473 substances tested in the local lymph node assay (LLNA) was compiled, comprising 295 sensitizers and 178 non-sensitizers. The overall predictivity of TIMES-SS and VEGA were 63% and 64% respectively. Both programs were far better at predicting sensitizers than non-sensitizers. The SAR rulebases from Toxtree and OASIS showed similar results for overall predictivity with 69% and 64% accuracy, respectively. The OASIS alerts were unique in that they were better at predicting non-sensitizers than sensitizers. This abstract may not reflect U.S. EPA policy.