Read-Across and Quantitative Structure–Activity Relationships (QSAR) for Making Predictions and Data Gap-Filling

Read-across is originally a non-statistical grouping approach for data gap filling. The grouping of chemicals may be done based on similarities in structural features, physicochemical properties, absorption/metabolism/distribution properties, etc. Based on the similarities to the source compounds with a known target property, predictions of the property can be made for a query chemical. Grouping and read-across have mostly been used in predictive toxicology. Quantitative structure–activity relationship (QSAR) is a statistical approach to correlate a target property with structural features with appropriate internal and external validation for the acceptability of the derived quantitative correlations as per the Organisation for Economic Co-operation and Development (OECD) criteria. QSAR models have been found to be useful for the predictions in medicinal chemistry, regulatory toxicology, materials sciences, food and agricultural sciences, nanosciences, etc. Both read-across and QSAR are used for regulatory prediction purposes in the absence of experimental data.