CN-121978225-A - Method for rapid classification and multicomponent quantitative analysis of different producing areas of parietal operation

CN121978225ACN 121978225 ACN121978225 ACN 121978225ACN-121978225-A

Abstract

The invention discloses a method for rapidly classifying different places of origin of a white skill and quantitatively analyzing multiple components. The method comprises the steps of collecting a large-area bighead atractylodes rhizome sample, carrying out pretreatment such as crushing and drying, accurately measuring the actual content of each component by utilizing High Performance Liquid Chromatography (HPLC), synchronously collecting sample NIR spectrum data, carrying out pretreatment and characteristic wave band screening on the spectrum by combining a chemometric method, constructing a training set based on an HPLC quantitative result and the NIR spectrum data, respectively training a partial least squares discriminant analysis (PLS-DA) model to realize the classification of the area, and realizing multicomponent synchronous quantification by a Partial Least Squares Regression (PLSR) model. The method provided by the invention does not need complex pretreatment, is lossless and rapid, is suitable for high-throughput quality evaluation of a large number of samples, and provides a reliable technical means for tracing the origin of the traditional Chinese medicinal materials and monitoring the quality of multiple indexes.

Inventors

WANG DIANLEI
XIA CHENGKAI
FANG WEI
Wan Dian
JIN HAOTIAN
Sun Nianxia
HAN ZHILI
HUANG PENG
HUANG HEPING
LIU YAOWU

Assignees

安徽中医药大学

Dates

Publication Date: 20260505
Application Date: 20251128

Claims (10)

1.A method for rapid classification and multicomponent quantitative analysis of different origin areas of a white operation, comprising: s1, collecting a plurality of bighead atractylodes rhizome samples of production places, and carrying out sample pretreatment; s2, performing component quantitative analysis on various components contained in the bighead atractylodes rhizome sample based on HPLC test to obtain the actual content of each component in the bighead atractylodes rhizome sample; S3, acquiring an NIR spectrum data graph of a bighead atractylodes rhizome sample based on NIR spectrum; S4, constructing a training set based on the actual content of the bighead atractylodes rhizome sample and the NIR spectrum data graph, and training a classification model PLS-DA of the producing area and a component quantitative model PLSR of the producing area; S5, performing the origin classification and the component quantitative analysis on the bighead atractylodes rhizome based on the trained PLS-DA model and PLSR model.
2. The method for rapid classification and multicomponent quantitative analysis of different origin of a white atractylodes rhizome according to claim 1, wherein the sample pretreatment in S1 comprises: S11, removing impurities on the surface of a bighead atractylodes rhizome sample; s12, drying the processed bighead atractylodes rhizome sample to constant weight in a 35 ℃ environment; S13, crushing the dried sample by using a crusher, and sieving by a No.3 sieve; s14, storing the sieved bighead atractylodes rhizome sample powder in a dryer for standby.
3. A method for rapid classification and multicomponent quantitative analysis of different origin of parisons according to claim 1, wherein S2 comprises the steps of: S21, preparing mixed standard substance solutions and bighead atractylodes rhizome sample solutions of different components; S22, setting HPLC detection parameters; s23, respectively feeding mixed standard substances and sample solutions of different components, calculating the actual content of each component of the bighead atractylodes rhizome sample, and obtaining HPLC data.
4. A method for rapid classification and multicomponent quantitative analysis of different origin of a white atractylodes according to claim 1, wherein S3 comprises the steps of: s31, setting parameters of an NIR spectrum data instrument, and collecting a near infrared NIR spectrum data graph of a built-in golden mirror background of a bighead atractylodes rhizome sample; s32, preprocessing the near infrared NIR spectrum data graph to obtain a preprocessed near infrared NIR spectrum data graph; S33, screening characteristic wavelengths of the preprocessed near infrared and NIR spectrum data graphs to obtain the screened NIR spectrum data graphs.
5. A method for rapid classification and multicomponent quantitative analysis of different origin areas of a white atractylodes rhizome according to claim 1, wherein: when the PLS-DA model is trained, an NIR spectrum data graph is used as an independent variable, and the type of a sample place of production is used as a dependent variable; in the PLSR model training, the NIR spectrum data graph is taken as an independent variable, and the actual content of each component tested by HPLC is taken as a dependent variable.
6. The method for rapid classification and multicomponent quantitative analysis of different origin of white skill in claim 4, wherein the preprocessing of the near infrared NIR spectrum data map in S32 comprises S-G smoothing, first derivative, second derivative, multiple scatter correction, standard normal variable transformation and/or data centering.
7. The method for rapid classification and multicomponent quantitative analysis of different origin of white spirit according to claim 4, wherein the step of performing characteristic wavelength screening in S33 comprises: based on IPLS, dividing continuous characteristic variables into a plurality of sub-wavelength intervals, establishing a partial least square model of different sub-wavelength interval combinations, evaluating performance, screening out a key sub-wavelength interval with the greatest contribution to the model as an effective characteristic, and screening an NIR spectrum data graph by using the effective characteristic.
8. The method of rapid classification and multicomponent quantitative analysis of different origin of a white atractylodes rhizome according to claim 1, further comprising qualitative origin and quantitative component assessment of PLS-DA model and PLSR model; the qualitative evaluation of the production place comprises cross verification accuracy, precision, recall rate and F1 score; The component quantitative evaluation includes residual prediction bias, correction decision coefficient, prediction decision coefficient, cross-validation decision coefficient, correction root mean square error, prediction root mean square error, and cross-validation root mean square error.
9. The method for rapid classification and multicomponent quantitative analysis of different origin of parities according to claim 8, wherein the cross-validation accuracy rating evaluation formula is expressed as: Wherein k is the cross-validation fold number, acc i is the accuracy in the ith cross-validation, Is the cross-validation accuracy.
10. The method for rapid classification and multicomponent quantitative analysis of different origin of parities according to claim 8, wherein said residual predictive bias quantitative evaluation formula is expressed as: Wherein RPD is the residual prediction error, SD is the standard deviation of the measured value, and RMSEP is the root mean square error.

Description

Method for rapid classification and multicomponent quantitative analysis of different producing areas of parietal operation Technical Field The invention relates to the technical field of classification of bighead atractylodes rhizome production places and component quantification, in particular to a method for rapidly classifying different production places of white atractylodes rhizome and performing multicomponent quantitative analysis. Background Atractylodis rhizoma (Atractylodes macrocephala Koidz) has effects of invigorating spleen, invigorating qi, eliminating dampness and promoting diuresis, and can be used as medicine for Compositae plant. As the demand of bighead atractylodes rhizome in the traditional Chinese medicine industry continues to increase, and bighead atractylodes rhizome producing areas continue to expand and shift, the differences in appearance, chemical components and the like are larger, and natural condition differences such as climate, soil and the like are caused. In addition, the ingredient content of the bighead atractylodes rhizome shows obvious regional difference due to different producing areas, so that the medicine effect is different, and therefore, the establishment of a rapid and accurate bighead atractylodes rhizome producing area identification and chemical ingredient quantitative analysis method has important significance for guaranteeing the quality and clinical curative effect of the bighead atractylodes rhizome. Traditional Chinese medicine identification methods include morphological examination, microscopic identification and chemical identification, however, these methods are limited by subjective dependence of the judgment and are not amenable to standardization. In modern times, high Performance Liquid Chromatography (HPLC), gas Chromatography (GC), thin Layer Chromatography (TLC) and other methods are often used, and the information of peak shape, retention time, peak area and the like of chemical components in a traditional Chinese medicine sample is combined to realize the distinction of the traditional Chinese medicine production places and the quantitative determination of the chemical components. However, during chromatography, gradient conditions are complicated to fumbly, sample integrity cannot be reduced, and batch processing is time consuming. In contrast, near-Infrared (NIR) has the advantages of rapidness, greenness, no damage and economy, and is increasingly popular in the field of traditional Chinese medicines. Near infrared spectrum is the spectrum between visible light (VIS) and mid-infrared light (MIR), the wave number range is about 12000-4000cm -1, the frequency multiplication and the frequency combination absorption of vibration of the hydrogen-containing group X-H (x= C, N, O) are mainly performed, and meanwhile, the frequency multiplication signal of partial carbon-containing functional group (c= O, C-O) can be detected, and the molecular structure information of most organic compounds is included. Chemometrics can convert complex signals of NIR into interpretable qualitative and quantitative output, and in the tasks of discrimination and classification, methods such as partial least squares discriminant analysis (PLS-DA) and Support Vector Machines (SVM) are widely applied to true and false discrimination or origin tracing, and in the aspect of content prediction, partial Least Squares Regression (PLSR) has robustness on multiple collinearity and is one of the common regression methods for mapping chemical component concentration of NIR spectrum data. By combining NIR with chemometrics, key information can be extracted from complex spectra, so that quick identification of traditional Chinese medicine production places and prediction of internal components can be realized. In the prior art, research on bighead atractylodes rhizome is mainly focused on separation and measurement of atractylone, lactone components and polysaccharide, but the measurement of phenolic acid, fatty acid and other volatile components in bighead atractylodes rhizome is less. Thus, there is a need for a rapid qualitative and quantitative analysis method for establishing bighead atractylodes rhizome of different origin based on NIR and HPLC combined with chemometric methods. On one hand, a PLS-DA model is established based on NIR spectral data to judge the white atractylodes rhizome in different producing areas, and on the other hand, a PLSR model is adopted to predict the contents of 11 chemical components (neochlorogenic acid, chlorogenic acid, cryptochlorogenic acid, atractylenolide III, atractylenolide II, atractylenolide I, beta-eucalyptol, linoleic acid, beta-elemene and oleic acid) in the white atractylodes rhizome, so that the model performance is good, a reference basis is provided for rapid content measurement and comprehensive quality control of white atractylodes rhizome medicinal materials, and an important tool is provided for manufacturers, consume