Quantitative structure-activity relationship (QSAR) modeling has been a time-tested method for exploring mechanisms of biological and toxicological actions and their quantitative predictions as applied to diverse fields. Chemical read-across is a similarity-based approach for data gap filling mainly applied in regulatory toxicology. Recently, QSAR and read-across techniques have been merged to a new emerging field of Read-across Structure-Activity Relationship (RASAR) that uses the chemical similarity concepts of read-across (unsupervised step) and finally develop a supervised learning model (like QSAR) (Luechtefeld et al., 2018). The RASAR method has so far been used only in case of graded predictions or classification-based modeling. In this work, we attempt, for the first time, to apply RASAR for quantitative predictions using a case study of androgen receptor binding affinity data. We have previously developed a partial least squares (PLS) QSAR model employing only 2D descriptors using the data set collected from the FDA EDKB database employing the <b>DTC-QSAR</b> tool, available from https://dtclab.webs.com/software-tools. We also obtained chemical read-across predictions using the features selected in the above model employing a tool developed by us and available from https://sites.google.com/jadavpuruniversity.in/dtc-lab-software/home. While making the read-across predictions, we have computed a number of error-based and similarity-based measures. We have then clubbed these additional measures along with selected descriptors from the previously developed QSAR model, redeveloped new PLS models from the training set, and predicted the endpoint using the test set data. Interestingly, many of these models outperform the internal and external validation quality of the original QSAR model. The quality of read-across predictions for the test set is also enhanced in the RASAR models. A RASAR model has the advantage over read-across predictions in providing easy interpretation and indicating quantitative contributions of important chemical features. The strategy described here should be applicable to other biological/toxicological/property data modeling.

First report of quantitative RASAR modeling applied to the androgen receptor binding affinity: Towards an approach of easy interpretability and efficient transferability of quantitative prediction models

Fall 2022 abstracts, posters, and presentations will launch in mid-September.

ACS Fall 2022

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.