Summary
Leading QSAR models provide supporting documentation in addition toIn three case studies, we explore the information provided by the ePISuite, t.e.S.t., and VeGA platforms for evaluating the bioconcentration factor (BCF) for three example compounds. We address the central practical challenges of how to determine whether a toxicity prediction for a compound is reliable, as well as how to discuss the predictions and supporting information so that the evidence and reasoning can be understood, reviewed, and potentially accepted by others. This investigation is the first inter-laboratory and cross-institutional review exercise on the reliability of QSAR results in the evaluation of an animal model. As far as we know, it is the first exercise of this nature.In sections 2-6 below, we start by explaining the rationale for this exercise: why the practical case-by-case use of QSAR models is an important focus for investigative review, especially in light of the eU regulatory demands from ReACH. We then describe the methods of this investigation, including the significance of the three example compounds as case studies, the features of the QSAR platforms used, and, finally, the issue of reliability for experimental BCF results and predictions.taking each compound as a separate case study, we describe the BCF predictions offered by each QSAR platform and the range of information provided to support those predictions. We discuss and illustrate the use of this information by toxicologists to reach a decision on the reliability of the prediction. We then discuss the decisions on reliability made by the 28 participants and the comments they made about their decisions. Having