Background: The identification of the geographical origin of Polygonatum cyrtonema Hua is of particular importance because the quality and market value of Polygonatum cyrtonema Hua from different production areas are highly variable due to differences in the growing environment and climatic conditions.
Objective: This study utilized near-infrared spectra (NIR) of Polygonatum cyrtonema Hua (n = 400) to develop qualitative models for effective differentiation of Polygonatum cyrtonema Hua from various regions.
Methods: The models were produced under different conditions to distinguish the origins distinctly. Ten pre-processing methods have been used to pre-process the original spectra (OS) and to select the most optimal spectral pre-processing method. Principal component analysis (PCA), partial least squares discriminant analysis (PLS-DA), and orthogonal partial least squares discriminant analysis (OPLS-DA) were employed to determine appropriate models. For simplicity, the pretreated full spectrum was calculated by different wavelength selection methods, and the four most significant variables were selected as discriminant indicator variables.
Results: The results show that Polygonatum cyrtonema Hua from different regions can be effectively distinguished using spectra from a series of samples analyzed by OPLS-DA. The accuracy of the OPLS-DA model is also satisfactory, with a good differentiation rate.
Conclusion: The study findings indicate the feasibility of using spectroscopy in combination with multivariate analysis to identify the geographical origins of Polygonatum cyrtonema Hua.
Highlights: The utilization of near-infrared spectroscopy combined with chemometrics exhibits high efficacy in discerning the provenance of herbal medicines and foods, thereby facilitating quality assurance measures.
Keywords: Polygonatum cyrtonema Hua; OPLS-DA; discriminate analysis; near-infrared spectroscopy.
© The Author(s) 2024. Published by Oxford University Press on behalf of AOAC INTERNATIONAL. All rights reserved. For permissions, please email: journals.permissions@oup.com.