Development of QSAR-based (MLR/ANN) predictive models for effective design of pyridazine corrosion inhibitors

Twenty pyridazine derivatives with previously reported experimental data were utilized to develop predictive models for the anticorrosion abilities of pyridazine-based compounds. The models were developed by using quantitative structure-activity relationship (QSAR) as a tool to relate essential molecular descriptors of the pyridazines with their experimental inhibition efficiencies. Chemical descriptors associated with frontier molecular orbitals (FMOs) were obtained using density functional theory (DFT) calculations, while others were obtained from additional calculations effected on Dragon 7 software. Five descriptors together with concentrations of the pyridazine inhibitors were used to develop the multiple linear regression (MLR) and artificial neural network (ANN) models. The optimal ANN model yielded the best results with 111.5910, 10.5637 and 10.2362 for MSE, RMSE and MAPE respectively. The results revealed that ANN gave better results than MLR model. The proposed models suggested that the adsorption of pyridazine derivatives is dependent on the five descriptors.Five pyridazine compounds were theoretically designed.

Automated summary

In this study, predictive models for the anticorrosion abilities of pyridazine-based compounds were developed using QSAR, and the essential molecular descriptors of the pyridazines were related to their experimental inhibition efficiencies. The results showed that the ANN model outperformed the MLR model in predicting the anticorrosion abilities of the pyridazine compounds.