Methods for reliability and uncertainty assessment and for applicability evaluations of classification- and regression-based QSARs.

Abstract

This article provides an overview of methods for reliability assessment of quantitative structure-activity relationship (QSAR) models in the context of regulatory acceptance of human health and environmental QSARs. Useful diagnostic tools and data analytical approaches are highlighted and exemplified. Particular emphasis is given to the question of how to define the applicability borders of a QSAR and how to estimate parameter and prediction uncertainty. The article ends with a discussion regarding QSAR acceptability criteria. This discussion contains a list of recommended acceptability criteria, and we give reference values for important QSAR performance statistics. Finally, we emphasize that rigorous and independent validation of QSARs is an essential step toward their regulatory acceptance and implementation.

Summary of a workshop on regulatory acceptance of (Q)SARs for human health and environmental endpoints.

“…Three review papers were circulated among participants before the workshop and provided a common background and helped in the discussions (Cronin et al 2003a(Cronin et al , 2003bEriksson et al 2003).…”

Noncovalent interactions between hydroxylated polycyclic aromatic hydrocarbon and DNA: Molecular docking and QSAR study

“…The applicability domain of the developed QSAR model was assessed by the Williams plot, i.e., the plot of standardized residuals ( ) versus leverage (Hat diagonal) values (h i ) (Eriksson et al, 2003). h i value of a chemical in the original variable space and the warning leverage value (h * ) are defined as:…”

Theoretical Prediction of Antiproliferative Activity against Murine Leukemia Tumor Cell Line (L1210). 3D‐Morse Descriptor and its Application in Computational Chemistry

“…For these reasons, in this study we plotted the leverage values calculated for every compound (X-axis) versus standard residuals (Y-axis). Then, the domain of applicability of the model was defined as a squared area within the AE 2 range for standard residuals and the leverage threshold h* [53,54]. This kind of plot allows a graphical detection of both the outliers and the influential chemicals in a model.…”