We developed an electronic nose method for classifying garlic cultivars. Each garlic cultivar gave different semiconductor gas sensor array response patterns, which we analysed using two-dimensional principal component analysis (PCA). The method was able to detect the differences between cultivars using a PCAoptimized set of five sensors without any significant loss of performance compared to the original eight sensors. The performance of the electronic nose system was confirmed using two alternative methods, namely (1) the cluster analysis of the genetic relationships between garlic cultivars using amplified fragment length polymorphism (AFLP) markers, and (2) measuring the concentrations of sulfur-containing compounds in each garlic cultivar by gas chromatography-mass spectrometry (GC–MS). Four Thai garlic cultivars were consistently characterized and placed in three groups using the AFLP, GC–MS, and electronic nose methods. These results suggest that garlic cultivars can be classified simply and quickly using a low-cost electronic nose system.
Garlic Electronic nose Semiconductor gas sensor Gas chromatography–mass spectrometry Principal component analysis Amplified fragment length polymorphism