(Q)SAR Assessment Framework: Guidance for the regulatory assessment of (Quantitative) Structure Activity Relationship models and predictions, Second Edition

Investment Rating - The report does not provide a specific investment rating for the industry or companies involved Core Insights - The (Q)SAR Assessment Framework aims to create a systematic and harmonized approach for the regulatory assessment of (Q)SAR models, predictions, and results based on multiple predictions, applicable regardless of the modeling technique, predicted endpoint, or regulatory purpose [19][20] - The framework emphasizes that the assessment of (Q)SARs should extend beyond model validity, as valid models can yield unacceptable predictions under certain conditions, necessitating dedicated assessments for individual predictions and results from multiple predictions [20][23] Summary by Sections 1. Assessment of (Q)SAR Models (Model Checklist) - The Model Checklist evaluates models based on OECD principles, including defined endpoints, unambiguous algorithms, defined applicability domains, and measures of goodness-of-fit, robustness, and predictivity [33][50] - Each principle is further detailed with assessment elements to ensure compliance and transparency in model evaluation [33][41][50] 2. Assessment of (Q)SAR Predictions (Prediction Checklist) - The Prediction Checklist establishes four principles for assessing predictions: correct inputs, substance within the applicability domain, reliability of predictions, and fitness for regulatory purposes [60] - Each principle is broken down into specific assessment elements to facilitate thorough evaluation [60][70] 3. Assessment of a (Q)SAR Result Derived from Multiple Predictions (Result Checklist) - The Result Checklist includes additional assessment elements to evaluate the integration of predictions for determining final results, focusing on the reliability and applicability of the predictions [24][70] - The assessment process is designed to streamline evaluations while ensuring that all relevant factors influencing prediction reliability are considered [24][70] 4. Final Considerations - The report includes updates to the (Q)SAR model reporting format (QMRF) and (Q)SAR prediction reporting format (QPRF), reflecting the newly established OECD (Q)SAR Prediction Principles [25] - The document encourages the application of the QAF principles in case studies and further research to enhance the understanding and implementation of (Q)SAR methodologies [26][27]