Classification of edible bird’s nest samples using a logistic regression model through the mineral ratio approach

Origin tracing and adulteration identification of bird's nest by high- and low-field NMR combined with pattern recognition

Edible bird's nest (EBN) is a high-value health food with various nutrients and bioactive components. With increasing demand for EBN, they are often adulterated with cheaper ingredients or falsely labeled by the origin information, thus harming consumer interests. In this study, high- and low-field nuclear magnetic resonance (HF/LF-NMR) technology combined with multivariate statistical analysis was used to identify the geographical marker of EBN from different origins and authenticate the adulterated EBN with various adulterants at different adulteration rates. Authentic EBN samples from Malaysia were used to simulate adulteration using gelatin (GL), agar (AG) and starch (ST) at 10 %, 20 %, 40 %, 60 %, 80 %, and 100 % w/w, respectively. The results showed significant differences in composition among EBN from different origins, with isocaproate and citric acid serving as geographical markers for Malaysia and Vietnam, respectively. Leucine, glutamic acid, and N-acetylglycoprotein serving as geographical markers for Indonesia. In addition, PLS model further verified the accuracy of origin identification of EBN. The LF-NMR results of adulteration EBN showed a linear correlation between the transverse relaxation (T2, S2) and the adulterated ratio. The OPLS-DA based on T2 spectra could accurately identify authentic EBN from adulterated with GL, AG and ST at 40 %, 20 %, and 20 %, respectively. Fisher discrimination model was able to differentiate at 20 %, 20 %, and 40 %, respectively. These results show that the 1H NMR combined with multivariate statistical analysis method could be a potential tool for the detection of origin and adulteration of EBN.