Novel two variable-QSPR analysis for authentication and typification of vegetable oils

Area of Science:

• Analytical Chemistry
• Food Chemistry
• Computational Chemistry

Background:

• Vegetable oils are crucial food components, with properties influenced by fatty acid composition.
• Characterizing oils requires understanding parameters like saponification and iodine indices.
• Predicting these properties aids in quality control and nutritional assessment.

Purpose of the Study:

• To develop a Quantitative Structure-Property Relationship (QSPR) model for predicting saponification and iodine indices in vegetable oils.
• To establish a method for classifying vegetable oils based on their fatty acid profiles.
• To explore the relationship between fatty acid composition and key oil properties.

Main Methods:

• Formulation of QSPR models using 144 vegetable oils with 1-8 fatty acid components.
• Calculation of 25,118 mixture descriptors based on fatty acid components and their weight percentages.
• Application of the Replacement Method for variable subset selection to identify optimal predictive descriptors.

Main Results:

• A novel two-variable QSPR model was successfully developed and validated.
• The model accurately predicted saponification and iodine indices for various vegetable oils.
• The approach demonstrated effectiveness in classifying oils of known composition but unknown experimental data.

Conclusions:

• The developed QSPR model provides a reliable method for predicting key properties of vegetable oils.
• This approach can be extended to other oil types, including fish oils, and serves as a foundation for further research.
• The study highlights the utility of QSPR in analyzing complex chemical mixtures relevant to food science.