Could you please share the entire presentation?
Hi Gautam, thanks for sharing these articles.
I can speak about our experience in Brazil, Anvisa does accept predictions from OECD QSAR Toolbox but not as a standalone tool - if looking to meet ICH M7 you need to combine an expert-rule based model (e.g. the ISS model for in vitro mutagenicity) with a statistical model (e.g. CAESAR for which results can be seen through VEGA), and every prediction needs to be followed by an expert review. But this is true for any software that is used for toxicity predictions, we always need to perform an expert review on the software output, and no tool is sufficiently predictive to be used as a standalone tool.
Hi Fernanda,
A big thank you for your expert insights. They are extremely helpful in our decision-making and in determining the way forward.