[Study on identification of different producing areas of Gastrodia elata using multivariable selection and two-dimensional correlation spectroscopy of near infrared spectroscopy]
Ying Hou,Tian-Jun Yuan,Juan Xu,Xiang-Yi Liu
DOI: https://doi.org/10.19540/j.cnki.cjcmm.2019.0016
Abstract:The efficacy and quality of Gastrodia elata from different producing areas are significant difference. The identification of producing area is helpful for the scientific and reasonable usage of medicinal material. Application near infrared spectroscopy( NIR) together with variance spectra,principal component analysis( PCA),interval partial least squares( i PLS) and genetic algorithm( GA),the characteristic spectra of G. elata from inside and outside of Yunnan province( except Zhaotong) and Zhaotong were extracted,and the pattern recognition models of i PLS-DA and GA-SVM were built,and the suitability of the models were also validated. The results showed that the prediction accuracy of i PLS-DA model was 96. 15%,the values of R2,RMSECV and RMSEP were 0. 893,0. 224 and0. 321,respectively. The prediction accuracy of GA-SVM model was 100% and RMSECV was 0. 719 4. Both methods could identify G. elata from different producing area preferably. Further,two-dimensional spectroscopy analysis was conducted for the characteristic spectra extracted by i PLS and GA. The results showed that the differential spectra of G. elata from outside and inside Yunnan were mainly located in the absorption areas of the stretch,bending and double frequency of C-H,C-N,O-H and N-H bonds of polysaccharides,aromatic hydrocarbon,amides and starch. The differential spectra of G. elata from outside Yunnan and Zhaotong were mainly located in the absorption areas of the stretch,bending and double frequency of C-H,O-H and N-H bonds of vegetable protein,aromatic hydrocarbon,polysaccharides,and alcohols. The differential spectra of G. elata from inside Yunnan and Zhaotong were mainly located in the absorption areas of the stretch,bending,transformation and double frequency of CHO,N-H,C-H,O-H and HOH bonds of lignin,aromatic hydrocarbon,alcohols,polysaccharides and aliphatics. Effective recognition of different producing areas and exploration the difference of component of G. elata could be conducted through NIR combined with multivariable selection and two-dimensional spectroscopy,which provided the basis for the reasonable development and efficient utilization of G. elata.