Geographical origin discrimination of lentils (Lens culinaris Medik.) using 1H NMR fingerprinting and multivariate statistical analyses
View abstract on PubMed
Summary
This summary is machine-generated.Nuclear Magnetic Resonance (NMR) fingerprinting and chemometrics successfully classified lentil samples by geographical origin. Principal Component Analysis with Linear Discriminant Analysis (PCA-LDA) achieved 100% recognition and 96.7% prediction accuracy.
Area Of Science
- Analytical Chemistry
- Food Science
- Chemometrics
Background
- Geographical origin is a critical factor influencing lentil quality and traceability.
- Untargeted 1H NMR fingerprinting offers a powerful, non-destructive analytical approach for complex biological matrices.
- Chemometric methods are essential for extracting meaningful patterns from complex NMR data.
Purpose Of The Study
- To develop and compare chemometric models for classifying lentil samples based on their country of origin (Italy vs. Canada).
- To evaluate the performance of various classification algorithms, including SIMCA, k-NN, PCA-LDA, and PLS-DA.
- To identify key metabolites contributing to the discrimination of lentil samples by origin.
Main Methods
- Analysis of lentil samples from Italy and Canada using untargeted 1H NMR fingerprinting.
- Application of chemometric techniques: Soft Independent Modelling of Class Analogy (SIMCA), k-Nearest Neighbor (k-NN), Principal Component Analysis followed by Linear Discriminant Analysis (PCA-LDA), and Partial Least Squares-Discriminant Analysis (PLS-DA).
- Validation of models using test sets and Monte Carlo Cross Validation.
Main Results
- The PCA-LDA model demonstrated superior performance, achieving 100% average recognition and 96.7% cross-validation prediction ability.
- All applied statistical models showed high prediction abilities (>95%), confirming their suitability for geographical classification.
- Specific metabolites contributing to the discrimination between Italian and Canadian lentil samples were identified.
Conclusions
- Untargeted 1H NMR fingerprinting combined with chemometrics is a robust and effective strategy for determining the geographical origin of lentils.
- PCA-LDA is a highly accurate method for classifying lentil samples by origin, offering excellent recognition and prediction capabilities.
- This approach supports lentil authenticity verification and supply chain traceability.