Lhasa Limited shared knowledge shared progress

Establishing best practice for the application of a novel statistical-model to evaluate potential mutagenic impurities under ICH M7

pdf fileBarber CG; Hanser T; Kruhlak NL; Stavitskaya L; Vessey J;

The (ICH) M7 guideline can allow in silico assessments for mutagenicity to be used which should consist of a statistical-based method (QSAR) and an expert rule-based method (SAR). In this work, we present best practice for the application of a new statistical-based QSAR software, Sarah Nexus, when applied to the assessment of genotoxic potential of pharmaceutical impurities. To establish best practice, optimal model application settings were assessed using two independently-derived external validation sets containing 513 (19% positive) and 809 (42% positive) compounds. The same optimal settings of 10% sensitivity (defines the relative impact of positive signals) and 10% equivocal (the threshold under which the signal is not strong enough for the model to make a call) were identified for both cases, resulting in balanced accuracy of 74 and 77%, sensitivity of 64 and 68%, and specificity of 86 and 85% respectively for the two validation sets. While coverage is reduced to 81-86% under these conditions, this is considered an acceptable trade-off in favour of overall improved accuracy and, in particular, improved sensitivity.

Presented by Chris Barber at SOT, Phoenix, USA; 27th March 2014.


© 2019 Lhasa Limited | Registered office: Granary Wharf House, 2 Canal Wharf, Leeds, LS11 5PS, UK Tel: +44 (0)113 394 6020
VAT number 396 8737 77 | Lhasa Limited is registered as a charity (290866)| Company Registration Number 01765239 (England and Wales).

Thanks to QuestionPro's generosity, we now have survey software that powers our data intelligence.