On a simple approach for determining applicability domain of QSAR models

Abstract Quantitative structure–activity/property/toxicity relationship (QSAR/QSPR/QSTR) modeling has been used in medicinal chemistry, material sciences, environmental fate modeling, risk assessment and computational toxicology for a long time. The Organization for Economic Co-operation and Development (OECD) has recommended that for application of validated QSAR models for prediction of new data points, there is a strict requirement of defining the applicability domain (AD) according to the Principle 3 . The AD is a theoretical region in chemical space encompassing both the model descriptors and modeled response which allows one to estimate the uncertainty in the prediction of a particular compound based on how similar it is to the training compounds employed in the model development. The AD is an important tool for reliable application of QSAR models, while characterization of interpolation space is significant in defining the AD. An attempt is made here to suggest a simple method for defining the X-outliers (in the case of the training set) and identifying the compounds that reside outside the AD (in the case of the test set) employing the basic theory of the standardization approach. Further, a standalone application named “Applicability domain using standardization approach” (available at http://dtclab.webs.com/software-tools and http://teqip.jdvu.ac.in/QSAR_Tools/ ) has been developed. The present study reports that the web application can be easily used for identification of the X-outliers for training set compounds and detection of the test compounds residing outside the AD using the descriptor pool of the training and test sets.