Predicting the composition of red wine blends using an array of multicomponent Peptide-based sensors

Area of Science:

• Analytical Chemistry
• Biomimetic Sensors
• Supramolecular Chemistry

Background:

• Mammalian senses of taste and smell provide a model for differential sensing.
• Synthetic receptors offer a powerful method for analyzing complex mixtures.

Purpose of the Study:

• To evaluate a peptide-based sensing array for differentiating and predicting red wine blend compositions.
• To correlate sensor responses with wine sensory attributes.

Main Methods:

• Development of a cross-reactive, supramolecular, peptide-based sensing array.
• Utilized Linear Discriminant Analysis (LDA) for blend differentiation.
• Employed Partial Least Squares (PLS) Regression for predictive modeling.

Main Results:

• Clear differentiation of wine blends based on tannin concentration and composition.
• A predictive model accurately estimated varietal percentages in blends with a 15% average error.
• Receptor responses strongly correlated with perceived astringency, indicating polyphenol binding.

Conclusions:

• Peptide-based sensing arrays are effective for analyzing complex wine mixtures.
• The developed array can predict blend composition and correlates with key sensory perceptions.
• This approach holds potential for objective wine analysis and quality control.