A QSPR analysis and curvilinear regression models for various degree-based topological indices: Quinolone antibiotics

Abstract

Topological indices play an essential role in defining a chemical compound numerically and are widely used in QSPR/QSAR analysis. Using this analysis, physicochemical properties of the compounds and the topological indices are studied. Quinolones are synthetic antibiotics employed for treating the diseases caused by bacteria. Across the years, Quinolones have shifted its position from minor drug to a very significant drug to treat the infections caused by bacteria and in the urinary tract. A study is carried out on various Quinolone antibiotic drugs by computing topological indices through QSPR analysis. Curvilinear regression models such as linear, quadratic and cubic regression models are determined for all topological indices. These regression models are depicted graphically by extending for fourth degree and fifth degree models for significant topological indices with its corresponding physical property showing the variation between each model. Various studies have been carried out using linear regression models while this work is extended for curvilinear regression models using a novel concept of finding minimal RMSE. RMSE is a significant measure to find potential predictive index that fits QSAR/QSPR analysis. The goal of RMSE lies in predicting a certain property of a chemical compound based on the molecular structure.