Prediction of Melting and Boiling Points of Fatty Acids and Their Derivatives Using Quantitative Structure-Activity Relationship (QSAR) Methodology

Abstract: Lipids contain fatty acids as fundamental components and their boiling and melting temperatures matter significantly for industrial uses. The researchers employed the Quantitative Structure-Property Relationship (QSPR) method for predicting melting and boiling points in fatty acids along with their derived substances. Measuring properties of chemical structures with molecular descriptors enabled the development of QSPR models that received validation by experimental results. During the training phase the predictive accuracy reached high levels as the coefficients of determination (R²) values turned out to be 0.948 for melting points and 0.938 for boiling points. The cross-validation validation produced R² values at 0.925 for estimating melting points and also 0.925 for estimating boiling points which shows robust predictive capability. Due to their reliable nature QSPR models exhibit strong performance in predicting thermal characteristics of fatty acids for uses in biodiesel production as well as food processing and cosmetics industries.