Comparative QSPR study of food preservatives using topological indices and regression models

Abstract

Abstract Food preservatives play a crucial role in extending the shelf life of food products. Understanding their physicochemical properties can help in designing more effective and safer preservatives. In this study, we use a Quantitative Structure Property Relationship (QSPR) approach based on topological indices to develop a predictive model for certain physicochemical properties of food preservatives. We compare the performance of linear and curvilinear regression models to understand which provides the best prediction model. Among the tested models, the cubic regression model demonstrated superior predictive performance. Of all the models tested, the cubic regression model had the best predictive capabilities such as $$R^2 = 0.9998$$ for vapour density and $$R^2 = 0.9039$$ for molecular weight. To validate our findings, we employ the developed model to estimate the properties of an existing food preservative, the propionic acid. Our results offer valuable insights that can aid in the development of new and improved food preservatives.