Hypericum sampsonii hance medicinal material quality evaluation method based on multi-component quantification and chemical mode recognition of UPLC fingerprint spectrum

Abstract

The invention relates to a UPLC fingerprint spectrum-based hypericum sampsonii hance medicinal material quality evaluation method based on multi-component quantification and chemical pattern recognition, which comprises the following steps: preparing a mixed reference substance solution and a test solution, and carrying out full spectrum peak matching on multiple batches of mixed reference substance solutions and test solutions according to preset chromatographic conditions, obtaining UHPLC fingerprints and control fingerprints of multiple batches of samples; and marking common peaks by using the fingerprint spectrum, and carrying out chemical mode identification to obtain a quality evaluation result of the hypericum sampsonii hance medicinal material. According to the invention, a simple and reliable hypericum sampsonii hance quality evaluation method is established by combining the UHPLC fingerprint spectrum with a chemical mode recognition method, and a certain scientific basis and practical reference are provided for the quality control and resource development of hypericum sampsonii hance.

Description

Technical Field

[0001] This invention relates to the field of medicinal material quality evaluation technology, and in particular to a method for evaluating the quality of *Gynostemma pentaphyllum* medicinal materials based on multi-component quantification and chemical pattern recognition using UPLC fingerprinting. Background Technology

[0002] *Hypericum sampsonii* Hance, also known as opposite-leaf grass, is a plant belonging to the genus *Hypericum* in the family Clusiaceae. It is cold in nature and bitter in taste, and has the effects of regulating menstruation, promoting blood circulation, stopping bleeding, and detoxifying. It is used for hematemesis, epistaxis, dysentery, enteritis, high fever in children, irregular menstruation, traumatic injuries, external bleeding, burns, and snake bites. Studies have shown that the main chemical components of *Hypericum sampsonii* include polyisoprene-based phloroglucinol derivatives, flavonoids, naphthodianthrones, benzophenone derivatives, volatile terpenes, and anthraquinones. Modern pharmacological research has found that *Hypericum sampsonii* possesses antiviral, anti-inflammatory, antitumor, and antidepressant functions.

[0003] Research on the quality evaluation of traditional Chinese medicine (TCM) by combining its multi-component and multi-target characteristics is of great significance for the quality control and scientific use of TCM. However, local standards and literature reports only include the morphological identification, microscopic identification, and thin-layer chromatography of *Gnaphalium affine*, and the quality control indicators only cover a few components. There is a scarcity of literature on the qualitative and quantitative analysis of multiple components of *Gnaphalium affine*, as well as UHPLC fingerprint studies.

[0004] As a traditional Chinese medicinal plant, *Gnaphalium affine* has a long history of use in my country, widely used as both a medicinal herb and food ingredient. However, since it primarily originates from the wild, its quality standards are not yet fully developed. Existing research is mostly limited to a single production area or a limited sample from the harvesting period, and the quality evaluation indicators are relatively limited, failing to reflect the overall quality characteristics of *Gnaphalium affine*. Summary of the Invention

[0005] To address the problems existing in the prior art, the purpose of this invention is to provide a quality evaluation method for Yuanbao herbal medicine based on multi-component quantification and chemical pattern recognition using UPLC fingerprinting. Samples of different batches of Yuanbao herbal medicine from multiple major producing areas across China were collected. Based on the separation characteristics of ultra-high performance liquid chromatography (UHPLC), a characteristic HPLC fingerprint of Yuanbao herbal medicine was established, and the contents of six target components were determined. Combined with chemical pattern recognition, cluster analysis (CA) was used to classify the samples. Principal component analysis (PCA) and orthogonal partial least squares-discriminant analysis (OPLS-DA) were used to screen for quality difference markers and analyze the correlation patterns between key components. Finally, a three-in-one evaluation system of "multi-component quantification - fingerprinting - pattern recognition" was established, laying a methodological foundation for standardized quality control of Yuanbao herbal medicine and subsequent spectral-effect correlation research.

[0006] To achieve the above objectives, the present invention provides the following solution: A quality evaluation method for Yuanbao herbal medicine based on multi-component quantification and chemical pattern recognition using UPLC fingerprinting includes: Prepare a mixed reference solution and a test solution, and perform full-spectrum peak matching on multiple batches of the mixed reference solution and the test solution according to preset chromatographic conditions to obtain UHPLC fingerprint chromatograms and reference fingerprint chromatograms of multiple batches of samples; The common peaks were identified using fingerprint spectroscopy, and chemical pattern recognition was performed to obtain the quality evaluation results of Yuanbao herbal medicine.

[0007] Optionally, preparing the mixed reference solution includes: Obtain reference standards for the target content; the reference standards include: caffeic acid, rutin, hyperoside, quercetin, kaempferol, and emodin; A mixed reference solution containing a first target amount of caffeic acid, a second target amount of rutin, a third target amount of hyperoside, a fourth target amount of quercetin, a fifth target amount of kaempferol, and a sixth target amount of emodin was prepared using methanol and filtered through a microporous membrane.

[0008] Optionally, the preparation of the test solution includes: Powdered samples of *Gynostemma pentaphyllum* from different origins were obtained, and methanol was added to each sample. The samples were then sealed and weighed, and ultrasonic extraction and cooling were performed. The lost weight was replenished with methanol, and the samples were then filtered through filter paper and a microporous membrane to prepare the test solution.

[0009] Optionally, the preset chromatographic conditions include: An Agilent ZORBAX RR StableBond C18 column was set up, with acetonitrile-0.5% acetic acid solution as the mobile phase, and the target gradient elution, target detection wavelength, target column temperature, target injection volume, and target flow rate were set. Setting the target gradient elution includes setting different concentrations of acetonitrile and 0.5% acetic acid solution at different elution times to achieve the target gradient elution.

[0010] Optionally, the chemical pattern recognition includes: similarity evaluation, cluster analysis, principal component analysis, orthogonal partial least squares-discriminant analysis, and Pearson correlation analysis.

[0011] Optionally, the similarity evaluation includes: The fingerprint similarity of multiple batches of samples is calculated to characterize the differences in fingerprint similarity between different origins, but the overall similarity is stable, and to show the consistency of chemical composition in each batch of samples.

[0012] Optionally, the cluster analysis includes: Using the peak area of ​​the common peak of multiple batches of samples as a variable, the intergroup linkage method was adopted and the squared Euclidean distance was used as the classification basis. That is, when the Euclidean distance is greater than the first preset threshold, the multiple batches of samples are divided into two categories to obtain the clustering results. This is used to characterize the quality differences of medicinal materials from different origins and batches, and no significant regional clustering characteristics are shown, reflecting the overall quality stability of *Gynostemma pentaphyllum*.

[0013] Optionally, the cluster analysis includes: PCA processing is performed using the peak area of ​​the common peak of multiple batches of samples as a variable to obtain feature values. The main target component is extracted when the feature value is greater than a second preset threshold, and the corresponding variance contribution rate and cumulative variance contribution rate are calculated. The comprehensive score is calculated using the variance contribution rate of each principal objective component as a weighting coefficient, and then ranked to characterize the differences in sample processing and storage.

[0014] Optionally, the orthogonal partial least squares-discriminant analysis includes: Using the peak area of ​​the common peaks of multiple batches of samples as the dependent variable and the sample number as the independent variable, batches clustered into one class in principal component analysis were subjected to OPLS-DA processing. The contribution of each common peak was determined by variable importance projection, and multiple quality difference markers were screened out and ranked. The component content of each quality difference marker was then used to evaluate the drug quality.

[0015] Optionally, the Pearson correlation analysis includes: Pearson correlation analysis was performed based on the total content of multiple target components and the PCA comprehensive score of multiple batches of samples to obtain the correlation coefficient results. The target components that are significantly positively correlated with the total content are used as evaluation indicators of the quality of *Gynostemma pentaphyllum* to characterize the overall quality.

[0016] The beneficial effects of this invention are as follows: This invention establishes a UPLC fingerprint spectrum for *Gynostemma pentaphyllum*, identifying 18 common peaks. Combining chemometric methods such as cluster analysis, PCA, and OPLS-DA, it achieves comprehensive quality evaluation of different batches of samples, demonstrating good precision, repeatability, and stability. Simultaneously, this invention quantitatively analyzes six representative components, including caffeic acid and rutin, clarifying the content differences of each component. This forms a comprehensive evaluation strategy of "fingerprint spectroscopy - multi-component quantification - chemical pattern recognition," overcoming the shortcomings of traditional single-indicator methods and providing a reference for the quality evaluation of *Gynostemma pentaphyllum*. Further research can be conducted based on this, combining its chemical characteristics with pharmacological activity to elucidate the pharmacodynamic material basis and mechanism of action, providing theoretical support for the standardized application and development of *Gynostemma pentaphyllum*. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 The UPLC fingerprints and control fingerprints of 36 batches of samples in this embodiment of the invention are shown. Figure 2 This is an HPLC chromatogram of the mixed reference solution according to an embodiment of the present invention; Figure 3 This is a dendrogram of cluster analysis of 38 batches of *Gnaphalium affine* according to an embodiment of the present invention. Figure 4 The images shown are PCA-X score diagrams of fingerprint spectra of *Gnaphalium affine* from various batches according to embodiments of the present invention. Figure 5 The following are OPLS-DA score charts and corresponding permutation test charts for 36 batches of samples in this embodiment of the invention; (a) is the OPLS-DA score chart, and (b) is the permutation test chart. Figure 6 The VIP values ​​of 18 common components in 36 batches of samples in this embodiment of the invention; Figure 7 The following are chromatograms of the mixed reference solution and the test solution in embodiments of the present invention; (a) is the mixed reference solution, and (b) is the self-made test solution. Figure 8 This is a flowchart of the method for quality evaluation of Yuanbao herbal medicine based on multi-component quantification and chemical pattern recognition using UPLC fingerprinting, according to an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0021] like Figure 8As shown in the figure, this embodiment discloses a method for quality evaluation of Yuanbao herbal medicine based on multi-component quantification and chemical pattern recognition using UPLC fingerprinting. The method includes: preparing a mixed reference solution and a test solution; performing full-spectrum peak matching on multiple batches of the mixed reference solution and the test solution according to preset chromatographic conditions to obtain UHPLC fingerprints and reference fingerprints of multiple batches of samples; using the fingerprints to mark common peaks and performing chemical pattern recognition to obtain the quality evaluation results of Yuanbao herbal medicine.

[0022] Specifically, this embodiment discloses a method for quality evaluation of *Gnaphalium affine* based on multi-component quantification and chemical pattern recognition using UPLC fingerprinting, including: A fingerprint chromatogram of *Gnaphalium affine* was established using UHPLC fingerprinting, and the contents of six identified components—caffeic acid, rutin, hyperoside, quercetin, kaempferol, and emodin—were determined. The quality was also evaluated using chemical pattern recognition. The method employed an Athena UHPLC C18 column (150 mm × 2.1 mm, 1.8 µm) with gradient elution of acetonitrile-0.5% acetic acid solution as the mobile phase at a flow rate of 0.2 mL·min⁻¹ and a detection wavelength of 254 nm. Similarity evaluation, cluster analysis (CA), principal component analysis (PCA), partial least squares-discriminant analysis (PLS-DA), and Pearson correlation analysis were used to evaluate 36 batches of *Gnaphalium affine*.

[0023] A UPLC fingerprint of *Gnaphalium affine* was established, identifying 18 common chromatographic peaks. Peaks 4, 8, 9, 12, 13, and 17 were identified as caffeic acid, rutin, hyperoside, quercetin, kaempferol, and emodin, respectively. The similarity of 36 batches of *Gnaphalium affine* ranged from 0.888 to 0.987. CA clustering 36 batches of *Gnaphalium affine* into two groups; PCA and OPLS-DA analyses yielded consistent results, classifying the samples into two groups. Pearson correlation coefficient analysis further screened out three main component peaks influencing the differences among *Gnaphalium affine* samples, and a methodological investigation of multiple index components was conducted. The recoveries of each component ranged from 99.50% to 100.28%, and the relative standard deviations of the measured values ​​were all less than 2.3%.

[0024] Conclusion: There are certain quality differences in *Gnaphalium affine* from different origins and harvesting seasons. A simple and reliable quality evaluation method for *Gnaphalium affine* was established using UHPLC fingerprinting combined with chemical pattern recognition, providing a scientific basis and practical reference for the quality control and resource development of *Gnaphalium affine*.

[0025] In one embodiment, this embodiment discloses a method for quality evaluation of *Gnaphalium affine* based on multi-component quantification and chemical pattern recognition using UPLC fingerprinting, including: 1.1 Reagents and Medicines: Caffeic acid (batch number: 110709-201607; purity: 97.4%), rutin (batch number: 100080-201409; purity: 91.9%), hyperoside (batch number: 111521-201708; purity: 95.1%), kaempferol (batch number: 110861-200808; purity: 95.9%), quercetin (batch number: 100081-201408; purity: 99.1%), and emodin (batch number: 110756-200110; purity: 96.2%) were all purchased from the National Institutes for Food and Drug Control. Methanol and acetonitrile were of chromatographic grade. All other reagents used in the experiment were of analytical grade. The water used in the experiment was purified water.

[0026] The 36 batches of *Gynostemma pentaphyllum* medicinal materials were collected from multiple major producing areas across the country. Information on the collection locations is shown in Table 1. Associate Professor Li Li of the Department of Traditional Chinese Medicine Identification at Guangxi University of Traditional Chinese Medicine identified them as *Gynostemma pentaphyllum*, a plant belonging to the Clusiaceae family. Hypericum sampsonii The dried aerial parts of Hance.

[0027] Table 1 Information on 36 batches of *Gynostemma pentaphyllum* samples 1.2 Instruments: Agilent 1290Ⅱ ultra-high pressure liquid chromatograph; XSE205DU electronic balance; AS30T ultrasonic cleaner.

[0028] 2. Methods and Results: 2.1 Chromatographic conditions: Agilent ZORBAX RR StableBond C18 (4.6×150mm, 3.5µm) column; mobile phase: acetonitrile (A) - 0.5% acetic acid solution (B), gradient elution: 0.00–10.00 min, 5%–15% A, 95%–85% B; 10.00–30.00 min, 15%–30% A, 85%–70% B; 30.00–55.00 min, 30%–55% A, 70%–45% B; 55.00–55.10 min, 55%–5% A, 45%–95% B; 55.10–60.00 min, 5% A, 95% B; detection wavelength: 254 nm; column temperature: 30℃; injection volume: 3 μL; flow rate: 0.2 mL / min.

[0029] 2.2 Solution preparation: 2.2.1 Preparation of mixed reference solution: Take appropriate amounts of six reference standards—caffeic acid, rutin, hyperoside, quercetin, kaempferol, and emodin—accurately weigh them, and then prepare a solution containing 0.021 mg / mL of caffeic acid using methanol. −1 Rutin 0.0095 mg / mL −1 Hyperoside 0.0212 mg / mL −1Quercetin 0.0022 mg / mL −1 Kaempferol 0.013 mg / mL −1 and emodin 0.0203 mg·mL −1 The mixed reference solution was filtered through a 0.45 μm microporous membrane and determined according to chromatographic conditions.

[0030] 2.2.2 Preparation of test solution: Take 1g of *Gynostemma pentaphyllum* sample powder (passed through a No. 3 sieve), accurately add 25mL of methanol, seal tightly, weigh, extract by sonication for 30min, cool, replenish the lost weight with methanol, filter with filter paper, filter through a 0.45μm microporous membrane, and determine the filtrate according to chromatographic conditions.

[0031] 2.3 Methodological Examination of UPLC Fingerprinting: 2.3.1 Precision test: Take the test solution of sample S2 and repeat the injection and determination 6 times according to the chromatographic conditions in section “2.1”. The peak area RSDs of caffeic acid, rutin, hyperoside, quercetin, kaempferol and emodin were calculated to be 1.26%, 0.88%, 0.84%, 4.02%, 0.97% and 0.85%, respectively, indicating that the instrument has good precision.

[0032] 2.3.2 Repeatability test: Take sample S2 (batch number 20230803), prepare 6 test solutions in parallel according to the method under section "2.2.2", inject and determine, and calculate the peak area RSD of caffeic acid, rutin, hyperoside, quercetin, kaempferol and emodin as 6.18%, 2.71%, 1.93%, 2.65%, 2.51% and 7.51%, respectively. The results show that the method has good repeatability.

[0033] 2.3.2. Spiking recovery test: Six portions of sample S8 (batch number 20230703), each approximately 1g, were taken. Six target components with the same content as the target components in the sample were added to each portion. The samples were continuously injected and determined six times according to the chromatographic conditions in section “2.3”. The average spiking recoveries of caffeic acid, rutin, hyperoside, quercetin, kaempferol, and emodin were calculated to be 100.28%, 98.80%, 99.22%, 98.88%, 99.50%, and 99.71%, respectively, with RSDs of 2.19%, 0.75%, 0.66%, 0.55%, 1.13%, and 2.21%, respectively.

[0034] 2.4 Establishment and similarity evaluation of UHPLC fingerprint chromatograms: 2.4.1 Construction of fingerprint chromatograms: Thirty-six batches of *Gnaphalium affine* samples were injected and analyzed according to the chromatographic conditions described in section "2.1". The samples were then imported into the *Similarity Evaluation System for Chromatographic Fingerprints of Traditional Chinese Medicine* (2012A version) for full-spectrum peak matching, yielding UHPLC fingerprint chromatograms and reference fingerprint chromatograms (R) for all 36 batches of samples. After establishing the fingerprint chromatograms using multi-point calibration, an overlay plot was generated, as shown below. Figure 1 As shown. A total of 18 common peaks were labeled, and compared with the chromatograms of each reference standard ( Figure 2 The common peaks 4, 8, 9, 12, 13, and 17 were identified as caffeic acid, rutin, hyperoside, quercetin, kaempferol, and emodin, respectively. The retention times of the common peaks showed little difference, with an RSD of less than 0.08%.

[0035] 2.4.2 Similarity Evaluation: The similarity of 36 batches of *Gynostemma pentaphyllum* decoction pieces was calculated using the "Similarity Evaluation System for Chromatographic Fingerprints of Traditional Chinese Medicine" (2012A version), as shown in Table 2. The results showed that there were differences in fingerprint similarity among different producing areas. The similarity of the 36 batches of *Gynostemma pentaphyllum* decoction pieces ranged from 0.877 to 0.992, with 32 batches showing a similarity > 0.9. This indicates that although *Gynostemma pentaphyllum* decoction pieces from different producing areas have certain differences, they are generally stable, suggesting good consistency in the chemical composition of each batch of samples.

[0036] Table 2. Similarity results of UPLC fingerprints of Yuanbao medicinal materials 2.5 Comprehensive Sample Evaluation Based on Chemical Pattern Recognition: 2.5.1 CA: Using the peak area of ​​18 common peaks from 36 batches of *Gnaphalium affine* samples as variables, the data was imported into SPSS 25.0 software. The inter-group linkage method was employed, with squared Euclidean distance as the classification criterion. For example... Figure 3 As shown, when the Euclidean distance threshold is >20, the 36 batches of samples can be divided into two categories: Category I includes batches 8, 11, and 30, and Category II covers the remaining 33 batches. The clustering results indicate that although there are quality differences among medicinal materials from different origins and batches, no significant regional clustering characteristics are observed, reflecting the relative stability of the overall quality of *Gynostemma pentaphyllum*.

[0037] 2.5.2 PCA: To further compare the quality differences among different batches of *Gynostemma pentaphyllum* samples, the peak areas of 18 common peaks from 36 batches of *Gynostemma pentaphyllum* samples were used as variables, and PCA was performed using SPSS 25.0 software. Five principal components were extracted with eigenvalues ​​> 1 as a condition, and their cumulative contribution to the total variance was 84.575%. Their variance contribution rates and cumulative variance contribution rates are shown in Table 3.

[0038] Table 3. PCA Eigenvalues ​​and Variance Contribution Rate The principal component scores of 36 batches of *Gnaphalium affine* samples were calculated using software. The comprehensive score was calculated using the variance contribution rate of each principal component as a weighting coefficient, and the samples were then ranked. The results are shown in Table 4. Table 3 shows that these five principal components contain 84.575% of the information contained in the measurement indicators. Specifically, the first principal component contains information on rutin, quercetin, kaempferol, and emodin; the third principal component contains information on hyperoside; and the fourth principal component contains information on caffeic acid. The score coefficients of the five principal components (denoted by X) are shown in Table 4.

[0039] Table 4 PCA Factor Loading Matrix Table 4 shows that the loading values ​​of chromatographic peaks 3, 6, 7–9, 12, 13, and 17 in principal component 1 are all relatively high, indicating that the first principal component mainly reflects the overall differences of the above chromatographic peaks. Chromatographic peaks 5, 11, and 16–18 have high loading values ​​in the second principal component, indicating that the second principal component reflects their comprehensive variation. The third principal component mainly reflects the information of chromatographic peaks 1, 9, and 10, with peak 10 making a particularly significant contribution. The fourth principal component mainly reflects the information of chromatographic peaks 2, 4, and 14. Principal component 5 is almost entirely determined by peak 15, indicating that peak 15 is a highly specific chromatographic peak. To analyze the variation patterns in the quality of *Yuanbao* medicinal materials, PCA was performed on the peak areas of 18 common peaks in 36 batches of samples after standardization using SIMCA 14.1 software. Figure 4 As shown.

[0040] The PCA plot showed that all sample points were within the 95% confidence interval, indicating that the chemical composition of the samples was highly similar, which is basically consistent with the results of the similarity analysis.

[0041] Combining the PCA score coefficients (X1~X19) of the 18 common peaks, the scores of the five principal components (Y1, Y2, Y3, Y4, Y5) were calculated. The variance contribution rate of the five principal components was used as the weighting coefficient, and the comprehensive score (Y) was calculated using Y=(0.3918Y1+0.2083Y2+0.1070Y3+0.0772Y4+0.0608Y5) / 0.8457. The results are shown in Table 5.

[0042] Table 5. PCA composite scores of 36 batches of samples As shown in Table 5, the PCA composite score of the 36 batches of samples ranged from 2.8438 to 0.2658, indicating that there were significant differences in the quality of different batches of samples. Among them, the overall quality of Guangxi samples was significantly better than that of other production areas, but there were large internal differences. Some samples from Hunan were the next best. This difference may be related to the sample processing and storage methods.

[0043] 2.5.3 PLS-DA: PLS-DA combines the advantages of principal component analysis (PCA) dimensionality reduction and variable correlation to form an analytical method that is both interpretable and predictive, suitable for multi-index data and small sample sizes. This study used SIMCA 14.1 software, with the peak areas of 18 common peaks in the samples as the dependent variable and the sample number as the independent variable. 36 batches clustered into one class in PCA but still exhibiting internal differences were subjected to PLS-DA and permutation tests, as shown in the figure. Figure 5 (a) and Figure 5 (b) in the middle.

[0044] Depend on Figure 5 It can be seen that the 36 batches of samples were classified into three categories, which is largely consistent with the classification results obtained from cluster analysis. The model stability parameter obtained after 200 permutation tests is R. 2 X=0.754, R 2 The values ​​of Y=0.53 are all greater than 0.5, indicating good model stability. The intercepts of the R² and Q² regression lines on the Y-axis are 0.214 and -0.351, respectively. The intersection of the Q² regression line and the Y-axis is below zero, indicating that there is no overfitting. The model is valid and feasible, and can be used to discriminate the quality differences among 36 batches of samples.

[0045] The importance projection (VIP) value reflects the contribution of each common peak; the larger the VIP value of a common component, the greater its impact on the differences between sample groups. The VIP values ​​of 18 common components in 36 batches of samples are shown below. Figure 6 As shown.

[0046] like Figure 6 As shown, based on a VIP value greater than 1, seven quality differential markers were screened. Arranged from largest to smallest VIP value, these markers are: quercetin (corresponding to common peak 12, VIP value 1.5478), common peak 7 (VIP value 1.4423), kaempferol (corresponding to common peak 13, VIP value 1.4303), common peak 6 (VIP value 1.4284), hyperoside (corresponding to common peak 12, VIP value 1.3151), common peak 1 (VIP value 1.2377), common peak 3 (VIP value 1.1999), and rutin (corresponding to common peak 8, VIP value 1.0863). The peak area of ​​these eight common peaks accounts for 47% of the total peak area of ​​the 18 common components. These components may affect the overall efficacy through synergistic effects with the pharmacodynamic components; therefore, accurate determination of their content is of significant value for drug quality evaluation.

[0047] 2.6 Determination of the content of multiple components: 2.6.1 System suitability assessment: Chromatograms of the mixed reference solution and the test solution are shown below. Figure 7(a) Figure 7 (b) of the above; wherein, 1-rutin; 2-hyperoside; 3-caffeic acid; 4-quercetin; 5-kaempferol; 6-emodin.

[0048] The results showed that the separation degree of each target component in the mixed control solution was greater than 1.5, indicating good separation effect. The retention time of the six target components in the sample was basically the same as that of the control. These results indicate that the method has strong specificity and high accuracy, and can be used for the quality analysis of *Gynostemma pentaphyllum*.

[0049] 2.6.2 Standard Curve: Accurately pipette 0.2, 1.0, 1.5, 2, 3.0, and 4.0 mL of the mixed reference solution into separate 10 mL volumetric flasks. Dilute with methanol and bring to volume, mix well, and prepare a series of mixed reference solutions. Inject and determine the solutions according to chromatographic conditions. Perform linear regression with the mass concentration of each target component as the x-axis and the corresponding peak area as the y-axis to calculate the linear parameters and determine the effective concentration range for quantification, as shown in Table 6.

[0050] Table 6 Linear Parameters 2.6.3 Sample analysis: Take 36 batches of samples and perform chromatographic analysis according to the method in section “2.3”. Each sample was analyzed twice. The results are shown in Table 7.

[0051] Table 7 Sample Analysis Results Pearson correlation analysis was performed on the six target components, the total content of these six components, and the PCA composite score obtained in Section 2.5.4 from 36 batches of samples. The correlation coefficients are shown in Table 8. "This indicates a confidence level below 0.05, meaning there is a significant positive correlation between the two corresponding components or composite scores." A value of 0.01 indicates a confidence level of less than 0.01, meaning that the two components or the overall score show a highly significant positive correlation.

[0052] Table 8 Pearson Correlation Coefficient Table 8 shows that the six components are significantly positively correlated with the total content. Specifically, caffeic acid, rutin, hyperoside, quercetin, and kaempferol show extremely significant positive correlations; emodin shows a significant positive correlation with hyperoside. This indicates a reliable linear relationship between the changes in the content of each component and the total content. In summary, hyperoside, quercetin, kaempferol, and the total content of the six components can be used as evaluation indicators for the quality of *Gynostemma pentaphyllum*. The correlation coefficients between hyperoside, quercetin, and kaempferol and the total content are close to 1, meaning their content changes can approximately reflect the overall level. A single component can reflect the overall quality, providing a basis for simplifying quality testing.

[0053] 3.1 Chromatographic conditions and content determination: This invention investigated the effects of different mobile phase systems (methanol-0.1% formic acid aqueous solution, methanol-0.1% phosphoric acid aqueous solution, etc.), different detection wavelengths (PDA full wavelength scan 200-400 nm), different flow rates (0.2, 0.3, 0.4, 0.5 mL / min), and different column temperatures (20, 30, 40℃) on fingerprint chromatograms. Finally, the chromatographic conditions under “2.1” were determined to be the optimal separation conditions.

[0054] 3.2 Results and Discussion: In the determination of the content of *Gnaphalium affine* from multiple production areas and batches, the overall content of rutin, quercetin, and kaempferol in samples from Guangxi was significantly higher than that in other production areas; while samples from Hunan were rich in hyperoside. This regional difference may be due to the combined influence of the growing environment, including the synergistic effects of factors such as moisture, temperature, light, soil, microorganisms, and microorganisms. Traditional Chinese medicine fingerprinting presents the composition, distribution, and content of the main chemical components in the samples, thereby providing an overall quality evaluation of traditional Chinese medicine. This invention established UPLC fingerprints for 36 batches of *Gnaphalium affine* from 17 production areas across China, identifying six main characteristic components: caffeic acid, rutin, hyperoside, quercetin, kaempferol, and emodin. Chemometric analysis showed that the overall chemical composition of the 36 batches of samples was highly consistent, but no significant geographical differences were found. However, overall, there is still a certain correlation in the quality of *Gnaphalium affine*. Furthermore, the cluster analysis of the fingerprint spectrum showed that the 36 batches of *Gynostemma pentaphyllum* were clustered into two categories. Except for three batches that were clustered into one category, the remaining 33 batches of *Gynostemma pentaphyllum* were all clustered into one category, which also indicates that the overall quality of *Gynostemma pentaphyllum* from different origins and batches is relatively stable.

[0055] Based on the OPLS-DA model, six key differential chemical components were screened. The factor loading plot shows that all six identified chemical components in the fingerprint spectrum contributed significantly to the quality of *Gynostemma pentaphyllum*, indicating that they are reasonable control indicators for the overall quality of *Gynostemma pentaphyllum*. OPLS-DA analysis of the *Gynostemma pentaphyllum* fingerprint spectrum revealed that the main component peaks causing differences in chemical quality between different batches were peaks 5, 1, 3, 8, 18, 11, and 15. Hyperoside (peak), quercetin (peak), and kaempferol (peak) are the main active ingredients in *Gynostemma pentaphyllum*, demonstrating the representativeness and accuracy of the content determination of these six components. Comprehensive analysis shows that while there are differences in the chemical quality of *Gynostemma pentaphyllum* from different origins, no regional regularity has been established. This confirms that origin factors cannot be used as a sole evaluation criterion; other factors such as altitude, climate, and pre-processing techniques must be considered comprehensively.

[0056] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for quality evaluation of *Gnaphalium affine* based on multi-component quantitative analysis and chemical pattern recognition using UPLC fingerprinting, characterized in that... include: Prepare a mixed reference solution and a test solution, and perform full-spectrum peak matching on multiple batches of the mixed reference solution and the test solution according to preset chromatographic conditions to obtain UHPLC fingerprint chromatograms and reference fingerprint chromatograms of multiple batches of samples; The common peaks were identified using fingerprint spectroscopy, and chemical pattern recognition was performed to obtain the quality evaluation results of Yuanbao herbal medicine.

2. The method for quality evaluation of *Gnaphalium affine* based on multi-component quantification and chemical pattern recognition using UPLC fingerprinting as described in claim 1, characterized in that... The preparation of the mixed reference solution includes: Obtain reference standards for the target content; the reference standards include: caffeic acid, rutin, hyperoside, quercetin, kaempferol, and emodin; A mixed reference solution containing a first target amount of caffeic acid, a second target amount of rutin, a third target amount of hyperoside, a fourth target amount of quercetin, a fifth target amount of kaempferol, and a sixth target amount of emodin was prepared using methanol and filtered through a microporous membrane.

3. The method for quality evaluation of Yuanbao herbal medicine based on multi-component quantification and chemical pattern recognition using UPLC fingerprinting as described in claim 1, characterized in that, The preparation of the test solution includes: Powdered samples of *Gynostemma pentaphyllum* from different origins were obtained, and methanol was added to each sample. The samples were then sealed and weighed, and ultrasonic extraction and cooling were performed. The lost weight was replenished with methanol, and the samples were then filtered through filter paper and a microporous membrane to prepare the test solution.

4. The method for quality evaluation of Yuanbao herbal medicine based on multi-component quantification and chemical pattern recognition using UPLC fingerprinting as described in claim 1, characterized in that, The preset chromatographic conditions include: An Agilent ZORBAX RR StableBond C18 column was set up, with acetonitrile-0.5% acetic acid solution as the mobile phase, and the target gradient elution, target detection wavelength, target column temperature, target injection volume, and target flow rate were set. Setting the target gradient elution includes setting different concentrations of acetonitrile and 0.5% acetic acid solution at different elution times to achieve the target gradient elution.

5. The method for quality evaluation of Yuanbao herbal medicine based on multi-component quantification and chemical pattern recognition using UPLC fingerprinting as described in claim 1, characterized in that, The chemical pattern recognition includes: similarity evaluation, cluster analysis, principal component analysis, orthogonal partial least squares-discriminant analysis, and Pearson correlation analysis.

6. The method for quality evaluation of Yuanbao herbal medicine based on multi-component quantification and chemical pattern recognition using UPLC fingerprinting as described in claim 5, characterized in that, The similarity evaluation includes: The fingerprint similarity of multiple batches of samples is calculated to characterize the differences in fingerprint similarity between different origins, but the overall similarity is stable, and to show the consistency of chemical composition in each batch of samples.

7. The method for quality evaluation of Yuanbao herbal medicine based on multi-component quantification and chemical pattern recognition using UPLC fingerprinting as described in claim 5, characterized in that, The cluster analysis includes: Using the peak area of ​​the common peak of multiple batches of samples as a variable, the intergroup linkage method was adopted and the squared Euclidean distance was used as the classification basis. That is, when the Euclidean distance is greater than the first preset threshold, the multiple batches of samples are divided into two categories to obtain the clustering results. This is used to characterize the quality differences of medicinal materials from different origins and batches, and no significant regional clustering characteristics are shown, reflecting the overall quality stability of *Gynostemma pentaphyllum*.

8. The method for quality evaluation of Yuanbao herbal medicine based on multi-component quantification and chemical pattern recognition using UPLC fingerprinting as described in claim 5, characterized in that, The cluster analysis includes: PCA processing is performed using the peak area of ​​the common peak of multiple batches of samples as a variable to obtain feature values. The main target component is extracted when the feature value is greater than a second preset threshold, and the corresponding variance contribution rate and cumulative variance contribution rate are calculated. The comprehensive score is calculated using the variance contribution rate of each principal objective component as a weighting coefficient, and then ranked to characterize the differences in sample processing and storage.

9. The method for quality evaluation of Yuanbao herbal medicine based on multi-component quantification and chemical pattern recognition using UPLC fingerprinting as described in claim 5, characterized in that, The orthogonal partial least squares discriminant analysis includes: Using the peak area of ​​the common peaks of multiple batches of samples as the dependent variable and the sample number as the independent variable, batches clustered into one class in principal component analysis were subjected to OPLS-DA processing. The contribution of each common peak was determined by variable importance projection, and multiple quality difference markers were screened out and ranked. The component content of each quality difference marker was then used to evaluate the drug quality.

10. The method for quality evaluation of Yuanbao herbal medicine based on multi-component quantification and chemical pattern recognition using UPLC fingerprinting as described in claim 5, characterized in that, The Pearson correlation analysis includes: Pearson correlation analysis was performed based on the total content of multiple target components and the PCA comprehensive score of multiple batches of samples to obtain the correlation coefficient results. The target components that are significantly positively correlated with the total content are used as evaluation indicators of the quality of *Gynostemma pentaphyllum* to characterize the overall quality.

Images