Application of Zagreb Index Models in Predicting the Physicochemical Properties of Unsaturated Fatty Acids

This paper develops models of the Zagreb index suitable for unsaturated fatty acids, which are crucial in performing metabolic functions in all living organisms. An algorithm-based methodology was brought into practice to optimize computation and data processing. Degree-based topological indices derived from the M-polynomial were computed using SPSS. Using linear regression analysis, the study proved that these indices are relevant to some physical properties of unsaturated fatty acids. The QSPR (Quantitative Structure-Property Relationship) models were developed to measure the efficiency of the models, making a correlation with four physical properties (LogP, Enthalpy, Molar Refractivity and Polarizability) and the indices. Right from the analyzed data, several multivariate linear regression models were built to find possible significant effects. It was concluded that the computed feature values adequately predict these physical properties' values, with the features' estimates showing great concentration on the observed values. Thus, the estimates regrettably relied on these values.