Benchmarking Assessment of Salmonella Mutagenicity QSAR Models for Potential Use Under ICH M7.pdf
This poster reviews an FDA study, in which they evaluated the performance of two freely-available, open source (Q)SAR programs, Toxtree and OECD Toolbox and three newly-released, commercial (Q)SAR programs, Leadscope Expert Alert System, Sarah Nexus, and Symmetry as potential candidates for qualifying pharmaceutical impurities. To effectively assess their performance, an in-house Salmonella mutagenicity database of 3979 compounds and a highly-curated version of the Hansen dataset of 3700 compounds were used for benchmarking. These sets are comprised of drug-like and industrial chemical examples containing a variety of functional groups as well as less-common atoms. These assessments will improve and facilitate the ability of FDA/CDER to interpret the quality and reliability of in silico data submitted under ICH M7 for the qualification of pharmaceutical.
Presented by representatives of the FDA at SOT, Phoenix, USA; 23rd - 27th March 2014.