Abstract:
Quantitative structure–activity relationship (QSAR) is the statistical correlation of physicochemical properties of the structure with biological activity. QSAR study involves two main steps: first is the generation of descriptors, and second is building and validating the models. We have developed a python script that effectively uses descriptor and activity data in building and validating best QSAR model. Although available software provides descriptor calculation, lack of either the components needed for cross-validation or proper workflow for QSAR analysis is the bottleneck. In this paper we report, the validation results of this software using data available for MMP 13 inhibitors and anti-malarial compounds. We have found that the values obtained using this script correlate well with that calculated from Microsoft excel and reported QSAR models.