Evolution of Quantitative Structure–Activity Relationships ((Q)SAR) for Mutagenicity
Andrew J. Teasdale · 2021
This chapter covers advances made to the overall accuracy and performance of the two main types of in silico quantitative structure–activity relationship ((Q)SAR) platforms that occurred as a result of the introduction of the ICH M7 guideline that exemplify the huge importance of cross-industry collaborative efforts in ensuring that (Q)SAR systems remain fit for purpose. Following the introduction of the original Ashby–Tennant SAR framework, development of proprietary and non-proprietary in silico (Q)SAR models for the prediction of mutagenicity has progressed steadily across various industries with mixed success depending on the particular use case. Primary Aromatic Amines are often present in starting materials used to synthesize pharmaceuticals and hence may be present as drug impurities. Ideally iterative improvements to the in silico (Q)SAR platforms would continue to codify this "expert knowledge" that would eventually decrease still further the requirement for human intervention in the interpretation of the in-silico predictions.