QSAR-Co

QSAR-Co software is an open source standalone tool, which is designed to develop robust classification-based QSAR models. The main motive of developing this software is to tackle one of the key problems faced during QSAR model development. The problem being to develop a single robust classification model employing the chemicals with experimental responses (biological response/toxicity/any property of interest) determined under different experimental/theoretical conditions, such as, multiple biological systems like E. coli, Acinetobacter baumannii, Burkholderia cepacia etc., AND/OR having different bioactivity measure type (i.e. IC50, Ki, EC50 etc.) AND/OR against multiple biological enzyme targets (like AChE, BACE1, CDK5, MAO-B, GSK-3b). Here, QSAR-Co is a short form for “QSAR with conditions” being one of the key features of this software, though one can also develop simple classification-based QSAR models with no conditions. Thus, the important goal of this software is to smoothly develop classification-based QSAR models handling both the cases, i.e., with or without conditions, while providing several essential functionalities for model development and screening query chemicals/databases.

There are two modules available in the software, i.e., 1) Model Development module and 2) Screen/Predict module. As the name suggests the first module supports the development of robust classification models and the second module helps in predicting the response class for a set of new query chemicals or screening a chemical database employing the already developed classification models.