ATR-FTIR as a green tool for rapid identity authentication of Gastrodia elata Blume under the influence of multi-biological variability

Rapid authentication of labelling information for medicinal and edible plants interfered by multiple biological variability (species, origin, growth pattern, etc.) has always been an important challenge for market supervision and management as well as consumers purchase orientation. Gastrodia elata Blume (G.elata) is used both as a food and as an herbal medicine for the treatment of migraine, hyperglycaemia and epilepsy. The advantages of Attenuated total reflection Fourier-transform infrared (ATR-FTIR) spectroscopy, which is a practical, fast and reliable spectroscopic technique used for solid samples, can be combined with chemometric models allows for fast confirmation of the desired label under the influence of different factors. A total of 344 G.elata samples from five geographic origins, three growth patterns and four species were discriminatively analyzed using ATR-FTIR spectra combined with the Orthogonal partial least squares discriminant analysis (OPLS-DA) model and Support Vector Machine (SVM) model. Spectra were preprocessed using S-G, FD, SD, MSC and SNV and their combinations to eliminate scattering and baseline drift, and the clustering results of PCA showed that S-G+SD preprocessing was the most effective. The models developed all had good accuracy with 97.09–100.00 % and Matthews correlation coefficient (Mcc) values of 0.80–1.00. The linear model PLSR and the nonlinear model SVR were used to predict the weight of dried individuals of G.elata for quality grade information, and the PLSR model with S-G+SD preprocessing could obtain a prediction accuracy of R²_P = 0.94 and RPD= 4.19. This study can provide a green and feasible method for the authenticity identification of G.elata labels under the influence of multiple biological variability, which is of great significance for the quality control and rapid detection of medicinal and edible plants in the market.