Mathematical Representation of Vasorelaxant Effect of Benzopyranes Using Box-Jenkins Model

Drug design is generally achieved through trial and error methods, which is a time and resource consuming process and novel methods are needed to improve it. Mathematical modelling is one of the possible solutions to speed up this process. In this study, we have presented Box-Jenkins model, to predict the vasorelaxant activity (pEC50) of a set of benzopyrane compounds. We used the HyperChem software to extract the molecular features of these compounds (51 molecules); then these features were used to generate the Box-Jenkins model. The dataset was divided into three groups: 37 for training, 9 for prediction and 5 for validation. The absolute relative deviation of the predicted values was 4.8% and 4.1% for validation and predicton sets, respectively. The correlation coefficient between the predicted points and the experimental values was 0.9657, revealing that the proposed model is capable of representing pEC50 of benzopyranes and could be used to speed up drug discovery processes.