Rheum tanguticum spectrum effect analysis method and system based on infrared spectrum

Abstract

The invention discloses a rheum tanguticum spectrum effect analysis method and system based on infrared spectroscopy, and belongs to the field of pharmaceutical analys.The method comprises the steps that S1, a rheum tanguticum sample is collected, and a sample solution is prepared; s2, collecting a high performance liquid chromatography fingerprint spectrum of the sample solution; s3, collecting an NIR spectrum of the rheum tanguticum sample; s4, determining the antioxidant activity of the sample solution, and establishing an antioxidant activity quantitative model by using the spectral data collected in the step S3; and S5, based on the high performance liquid chromatography fingerprint spectrum in the step S2, screening the anti-oxidation medicinal component of the rheum tanguticum, and detecting the anti-oxidation medicinal component of the rheum tanguticum by using an anti-oxidation activity quantitative model. The invention investigates the spectrum-effect relationship between the high performance liquid chromatography fingerprint of rheum tanguticum and the in-vitro antioxidant activity, and screens the antioxidant medicinal components of rheum tanguticum. The infrared spectrum rapid evaluation of the antioxidant activity is realized, and a scientific reference is provided for the quality standard improvement and clinical application of medicinal materials.

Description

Technical Field

[0001] The present invention relates to the field of drug analysis, and in particular to a spectral-effect analysis method and system for Tangut rhubarb based on infrared spectroscopy. Background Art

[0002] Tanguticum rhubarb (Rheum tanguticum Maxim. ex Balf.) is a tall perennial herb in the genus Rheum (Rheum Linn.) of the Polygonaceae family. It is one of the three authentic rhubarb species listed in the Pharmacopoeia of the People's Republic of China, and its roots and rhizomes are used medicinally. Modern research has shown that rhubarb contains a variety of chemical components, including anthraquinones, anthrones, tannins, flavonoids, and polysaccharides. It exhibits pharmacological effects such as purgative, anti-inflammatory, antioxidant, anti-tumor, hepatoprotective, and cardiovascular protection. In 1956, Harman proposed the free radical theory of aging, positing that oxidation is the primary cause of biological aging and various diseases. Oxidative damage can lead to chronic diseases such as cardiovascular disease and cancer, and scavenging free radicals is a key approach to preventing oxidative damage. Current research on tanguticum rhubarb focuses on the extraction, isolation, and pharmacological activity of its active ingredients. However, understanding the material basis for its antioxidant activity and rapidly evaluating its antioxidant activity remain challenges.

[0003] Traditional activity evaluation relies primarily on the "extraction-isolation-purification-verification" model for screening single compounds, which is labor-intensive, costly, and time-consuming. Chromatographic fingerprinting, particularly high-performance liquid chromatography (HPLC) fingerprinting, has been widely used in the quality evaluation and identification of traditional Chinese medicines. Combining this with pharmacological indicators allows for the establishment of rational spectrum-activity models and analytical methods, systematically revealing the relationships between individual components and pharmacological effects, further elucidating pharmacologically active components and their compatibility. Existing techniques have revealed potential key components for the quality evaluation and identification of Phyllanthus emblica (Emblica officinalis) through spectrum-activity relationship studies. Spectrum-activity relationship and molecular docking were used to screen key components in Polygonum cuspidatum extract that inhibit melanin production. This demonstrates that the use of spectrum-activity relationship to screen drug quality markers aligns with the theoretical framework of Traditional Chinese Medicine (TCM) and addresses the shortcomings of chemical fingerprinting in controlling TCM quality.

[0004] Currently, research on the medicinal material Rhubarb (Rhubarb tangutica) primarily focuses on the extraction, isolation, and pharmacological activity of its active ingredients. Existing antioxidant research primarily focuses on determining the free radical scavenging capacity of crude rhubarb extracts and investigating their antioxidant pharmacological activity. However, there are still deficiencies in the basic analysis of antioxidant active substances and rapid evaluation methods. The spectrum-activity relationship between its active ingredient groups and antioxidant potency remains unclear. There is an urgent need to systematically investigate this spectrum-activity relationship and establish a rapid and accurate method for evaluating antioxidant activity. Traditional in vitro antioxidant assays for Chinese medicinal materials are limited by complex procedures, high technical requirements, high testing costs, and susceptibility to environmental influences. Summary of the Invention

[0005] The purpose of the present invention is to overcome the problems existing in the prior art and provide a spectrum-effect analysis method and system for Tangut rhubarb based on infrared spectroscopy. Based on the spectrum-effect correlation strategy, the material basis of the antioxidant activity of Tangut rhubarb is analyzed and revealed. At the same time, a rapid detection model for the antioxidant activity of Tangut rhubarb is constructed by combining near-infrared spectroscopy technology with different chemometric methods, which realizes rapid and accurate detection of antioxidant activity and provides technical support for the construction of an intelligent evaluation system for the quality of Chinese medicinal materials.

[0006] The object of the present invention is achieved through the following technical solutions:

[0007] In a first aspect, a method for analyzing the spectral efficacy of Rhubarb Tangut based on infrared spectroscopy is provided, comprising the following steps:

[0008] S1. Collect Rheum tanguticum samples and prepare sample solutions;

[0009] S2. collecting a high performance liquid chromatography fingerprint of the sample solution;

[0010] S3, collecting the NIR spectrum of the Rhubarb Tangut sample;

[0011] S4, determining the antioxidant activity of the sample solution, and establishing a quantitative antioxidant activity model using the spectral data collected in step S3;

[0012] S5. Based on the high performance liquid chromatography fingerprint in step S2, the antioxidant active ingredients of Rheum tanguticum were screened, and the antioxidant active ingredients of Rheum tanguticum were detected using an antioxidant activity quantitative model.

[0013] In some embodiments, collecting the Rhubarb Tangut sample comprises:

[0014] Tangut rhubarb samples were collected from various places in Qinghai Province from September to October. The roots of the collected Tangut rhubarb samples were washed, sliced, dried in the shade, coarsely crushed, passed through an 80-mesh sieve, and placed in a desiccator for later use.

[0015] In some embodiments, preparing the sample solution comprises:

[0016] Take 0.5000±0.0001g of Tangut Rhubarb sample powder, place it in a stoppered conical flask, accurately add 25mL of methanol, heat and reflux for 1h, cool, transfer to a centrifuge tube and centrifuge at 4000rpm*15min, then use a volumetric flask to make up the volume, add methanol to make up the lost weight, shake well, and take the filtrate.

[0017] In some embodiments, collecting the high performance liquid chromatography fingerprint of the sample solution comprises:

[0018] Chromatographic conditions: Eclipse Plus C column 18 (4.6 mm × 250 mm, ID 5 μm), mobile phase: acetonitrile (A)-0.1% phosphoric acid aqueous solution (B), gradient elution program: 0-15 min, 2%→15%A; 15-25 min, 15%→20%A; 25-40 min, 20%→20%A; 40-70 min, 20%→30%A; 70-90 min, 30%→55%A; 90-100 min, 55%→75%A; injection volume: 10 μL; detection wavelength: 254 nm; column temperature: 30°C; volume flow rate: 0.8 mL / min;

[0019] Precision tests, repeatability tests, stability tests and fingerprint similarity evaluations were carried out under the above chromatographic conditions.

[0020] In some embodiments, the NIR spectrum acquisition comprises:

[0021] The NIR one-dimensional infrared spectrum of Tangut rhubarb sample was collected using the NIR fiber of the Fourier transform infrared spectrometer, with 32 scans and a resolution of 8 cm -1 , the spectrum acquisition range is 10000-4000cm -1 Each sample was scanned three times, the air background value was deducted before each scan, and three spectra of each sample and the average spectrum were taken.

[0022] In some embodiments, determining the antioxidant activity of the sample solution comprises:

[0023] ABTS value determination, DPPH value determination and FRAP value determination.

[0024] In some embodiments, establishing a quantitative model for antioxidant activity comprises:

[0025] Establish a quantitative detection model corresponding to different antioxidant activity values ​​under NIR spectroscopy.

[0026] In some embodiments, the screening of the antioxidant effective components of Rheum tanguticum contains the following steps:

[0027] The antioxidant effect of Rheum tanguticum is respectively subjected to cluster analysis, principal component analysis, partial least squares regression analysis and Pearson correlation analysis.

[0028] The detection of the antioxidant effective components of Rheum tanguticum by the antioxidant activity quantitative model contains the following steps:

[0029] The detection effects of the quantitative detection models under different spectra are compared.

[0030] In the second aspect, a spectrum-effect analysis system of Rheum tanguticum based on infrared spectrum is provided, which contains the following steps:

[0031] A sample preparation module is used to collect Rheum tanguticum samples and prepare sample solutions.

[0032] A fingerprint spectrum acquisition module is used to collect high performance liquid chromatography fingerprints of the sample solutions.

[0033] A spectrum acquisition module is used to collect NIR spectra of the Rheum tanguticum samples.

[0034] A quantitative model establishment module is used to determine the antioxidant activities of the sample solutions and establish antioxidant activity quantitative models by using the collected spectrum data.

[0035] A spectrum-effect analysis module is used to screen antioxidant effective components of Rheum tanguticum based on the high performance liquid chromatography fingerprints and detect the antioxidant effective components of Rheum tanguticum by using the antioxidant activity quantitative models.

[0036] It should be further explained that the technical features of the above-mentioned embodiments can be combined or replaced with each other to form new technical solutions without conflict.

[0037] Compared with the prior art, the present application has the following advantages:

[0038] The present application preliminarily clarifies the characteristic chromatographic peaks and the chemical components thereof related to the antioxidant activity of Rheum tanguticum by studying the spectrum-effect relationship of Rheum tanguticum, so as to explore the pharmacodynamic material basis of the in vitro antioxidant activity of Rheum tanguticum. The infrared characteristic spectrum and the antioxidant activity of Rheum tanguticum are analyzed by means of near infrared (NIR) spectrum and chemometrics, a quantitative model between the infrared spectrum and the antioxidant activity is established, and the in vitro antioxidant activity of Rheum tanguticum samples is effectively predicted. The relationship between the antioxidant activity of Rheum tanguticum and the spectrum-substance and the spectrum-effect is constructed, and a scientific basis for the antioxidant quality control of Rheum tanguticum is provided. BRIEF DESCRIPTION OF DRAWINGS

[0039] Figure 1 A flow chart of a method for analyzing the spectral efficacy of Tangut Rhubarb based on infrared spectroscopy according to an embodiment of the present invention is shown;

[0040] Figure 2 The fingerprints of 18 batches of Tangut rhubarb samples shown in the embodiments of the present invention are as follows;

[0041] Figure 3 HPLC chromatograms of the reference substance (A) and the Rheum tanguticum extract sample (B) shown in the examples of the present invention;

[0042] Figure 4 A cluster analysis diagram of different sampling locations of Tangut rhubarb shown in an embodiment of the present invention;

[0043] Figure 5 Schematic diagram of the determination of ABTS, DPPH, and FRAP antioxidant activity values ​​of 18 batches of Rhubarb tangutica samples shown in the examples of the present invention;

[0044] Figure 6 The PCA score graphs of 18 batches of Rheum tanguticum extracts shown in the embodiments of the present invention are as follows;

[0045] Figure 7 Graphs showing the importance of common peaks for ABTS (a), DPPH (b), and FRAP (c) according to the examples of the present invention;

[0046] Figure 8 A Pearson correlation analysis diagram is shown for an embodiment of the present invention;

[0047] Figure 9 This is a comparative diagram showing the effects of the common components in Rheum tanguticum on the antioxidant activity shown in the examples of the present invention;

[0048] Figure 10 The original NIR spectrum of Rhubarb tanguticum shown in the embodiment of the present invention;

[0049] Figure 11 This is the outlier elimination result shown in the embodiment of the present invention;

[0050] Figure 12 VIP score diagram of NIR spectral data shown in an embodiment of the present invention;

[0051] Figure 13 This is a schematic diagram of the optimal modeling result shown in an embodiment of the present invention, wherein: Figure 13 (a), (b), and (c) correspond to the optimal results of Python software modeling. Figure 13 (d), (e), and (f) correspond to the optimal modeling results of TQ analyst software. DETAILED DESCRIPTION

[0052] The technical solutions of the present invention are described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings herein can be arranged and designed in various different configurations. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0053] It should be noted that the defects existing in the solutions in the above-mentioned prior art are the results obtained by the inventor after practice and careful research. Therefore, the discovery process of the above-mentioned problems and the solutions proposed in the embodiments of this application below for the above-mentioned problems should be the contributions made by the inventor to this application in the process of invention and creation, and should not be understood as technical contents known to technical personnel in this field.

[0054] In response to the technical problems pointed out in the background technology, the embodiments provided by the present invention are as follows:

[0055] Example 1

[0056] In an exemplary embodiment, a method for analyzing the spectrum effect of Tanggut Rhubarb based on infrared spectroscopy is provided, referring to Figure 1 , including the following steps:

[0057] S1. Collect Rheum tanguticum samples and prepare sample solutions;

[0058] S2. collecting a high performance liquid chromatography fingerprint of the sample solution;

[0059] S3, collecting the NIR spectrum of the Rhubarb Tangut sample;

[0060] S4, determining the antioxidant activity of the sample solution, and establishing a quantitative antioxidant activity model using the spectral data collected in step S3;

[0061] S5. Based on the high performance liquid chromatography fingerprint in step S2, the antioxidant active ingredients of Rheum tanguticum were screened, and the antioxidant active ingredients of Rheum tanguticum were detected using an antioxidant activity quantitative model.

[0062] Specifically, this invention investigates the spectrum-activity relationship between the high-performance liquid chromatography (HPLC) fingerprint of Rheum tanguticum and its in vitro antioxidant activity to screen for active antioxidant components in Rheum tanguticum. A quantitative infrared spectroscopy detection model for the antioxidant activity of Rheum tanguticum was constructed to provide a basis for the quality evaluation of the antioxidant efficacy of Rheum tanguticum. HPLC was used to collect fingerprints from each batch of Rheum tanguticum to identify common peaks. ABTS, DPPH, and FRAP values ​​were measured for the samples. Principal component analysis, partial least squares analysis, and Pearson correlation analysis were used to establish a spectrum-activity relationship for the antioxidant activity of the samples. Furthermore, infrared spectroscopy was used to construct quantitative models for the antioxidant activity of Rheum tanguticum, and the effectiveness of each model was evaluated. This method clarifies the material basis for the antioxidant activity of Rheum tanguticum and enables rapid infrared spectroscopy evaluation of antioxidant activity, providing a scientific basis for improving the quality standards of medicinal materials and their clinical application.

[0063] Example 2

[0064] Based on the inventive concept of Example 1, this example provides a specific experimental analysis process, which mainly includes the following parts:

[0065] 1. Test method

[0066] 1.1. Instruments and Reagents

[0067] Instruments: Fourier transform infrared spectrometer (iS50, ThermoNicolet), high performance liquid chromatograph (Infinity1260, Agilent Technologies, USA), microplate reader (Epoch2, BioTek Instruments, USA), grinder (Tianjin Test Co., Ltd.), electronic balance (ME104, 0.0001 g, Mettler Toledo, Switzerland), chromatographic column (Agilent, USA), ultrapure water machine (Milion-Q Integral3, Merck Chemical Technology Co., Ltd., Germany), drying oven (Shanghai Yiheng Scientific Instrument Co., Ltd., China).

[0068] Reagents: Total antioxidant capacity (T-AOC) assay kit (FRAP microplate method) (A30IR224565, Yuanye Biotechnology), total antioxidant capacity (T-AOC) assay kit (ABTS microplate method) (A10IR222662, Yuanye Biotechnology), acetonitrile (chromatographic grade, Supelco®, USA), methanol (analytical grade, Futon, China), phosphoric acid (chromatographic grade, Aladdin Reagent (Shanghai) Co., Ltd., China). All other reagents were of analytical grade.

[0069] Reference substances: aloe-emodin (batch number: II20683), chrysophanol (batch number: MP00111), rhein (batch number: MP00560), emodin (batch number: 09H18Q), physcion (batch number: SA10924) were purchased from Henan Standard Material Research Center (China). Sennoside A (batch number: N2303165401), sennoside B (batch number: N2303165508) were purchased from Sichuan Hengcheng Zhiyuan Biological Technology Co., Ltd. (China). Sennoside C (batch number: AFCL0602), emodin-8-O-glucoside (batch number: AZDD0351) were purchased from Chengdu Elfa Biological Technology Co., Ltd. (China). Rhein-8-O-β-D-glucoside (batch number: MUST-12111602) was purchased from Chengdu Man Sit Biological Technology Co., Ltd. (China). Chrysophanol-8-O-glucoside (batch number: 141208), physcion-8-O-glucoside (batch number: 131012) were purchased from Chengdu Keluoma Biological Technology Co., Ltd. (China). Aloe-emodin-8-O-glucoside (batch number: DR010436) was purchased from Dingrui Chemical (Shanghai) Co., Ltd.

[0070] 1.2. Sample source

[0071] The samples of Rheum tanguticum Maxim. ex Ralf. used in the test were collected from various places in Qinghai Province from September to October. After cleaning, slicing, air-drying, and coarse crushing, the roots of the samples were sieved through an 80-mesh sieve and placed in a desiccator for use. The original plant samples were identified by professional researchers as Rheum tanguticum Maxim. ex Ralf. The specific sample information is shown in Table 1.

[0072]

[0073] 1.3. Preparation of test solution

[0074] Take 0.5000±0.0001 g of the powder, place it in a conical flask with a stopper, precisely add 25 mL of methanol, heat it to reflux for 1 h, cool it down, transfer it to a centrifuge tube, centrifuge it at 4000 rpm for 15 min, then use a volumetric flask to make up the volume, use methanol to make up the weight loss, shake it well, and take the filtrate, which is obtained.

[0075] 1.4. Preparation of reference solution

[0076] Dissolve the reference substances in methanol at appropriate concentrations to prepare the reference mother liquor, and dilute to prepare the mixed reference.

[0077] 1.5. HPLC fingerprint collection

[0078] 1.5.1. Chromatographic conditions

[0079] Chromatographic column Eclipse Plus C18 The column was eluted with a 4.6 mm × 250 mm diameter column (4.6 mm × 250 mm, ID 5 μm). The mobile phase was acetonitrile (A)-0.1% aqueous phosphoric acid (B). Gradient elution program: 0–15 min, 2% → 15% A; 15–25 min, 15% → 20% A; 25–40 min, 20% → 20% A; 40–70 min, 20% → 30% A; 70–90 min, 30% → 55% A; 90–100 min, 55% → 75% A. Injection volume: 10 μL; detection wavelength: 254 nm; column temperature: 30°C; flow rate: 0.8 mL / min.

[0080] 1.5.2. Methodological Review

[0081] 1.5.2.1. Precision test

[0082] The same sample was injected five times continuously according to the chromatographic conditions in 1.4.1., and the relative standard deviation (RSD) of the retention time and content percentage of the common peak was calculated respectively.

[0083] 1.5.2.2. Repeatability test

[0084] Extract the same sample five times and inject 10 μL of each extract under the chromatographic conditions in 1.4.1. Calculate the RSD values ​​of the retention time and content percentage of the common peak to determine the repeatability of the method.

[0085] 1.5.2.3. Stability test

[0086] Inject 10 μL of the same sample at 0h, 2h, 8h, 12h, and 24h after extraction using the chromatographic conditions in 1.4.1. Calculate the RSD values ​​of the retention time and content percentage of the common peak to determine the stability of the method.

[0087] 1.5.3. Fingerprint similarity evaluation

[0088] The fingerprints of 18 batches of sample chromatograms were processed using the "Chinese Medicine Chromatographic Fingerprint Similarity Evaluation System 2012 Edition" software.

[0089] 1.6. Near-infrared spectroscopy acquisition

[0090] The NIR fiber module of the Fourier transform infrared spectrometer was used to collect the NIR one-dimensional infrared spectrum of the Tangut rhubarb extract sample, with 32 scans and a resolution of 8 cm -1 , the spectrum acquisition range is 10000 – 4000 cm -1 Each batch of samples was divided into three equal parts and scanned three times. The air background value was deducted before each scan, and three spectra of each sample were taken and the average spectrum was obtained.

[0091] 1.7. Antioxidant activity assay

[0092] ABTS value determination

[0093] Take methanol reflux extraction and dilute the test solution 10 times as the sample solution, and operate according to the kit method. Mark different sample wells as A0 and A1, react in the dark for 5 minutes, and measure the absorbance A value at 734nm. Prepare 280μLABTS working solution and 7μL methanol mixed solution, measure the absorbance value as A0; prepare 280μLABTS working solution and 7μL sample solution mixed solution, measure the absorbance value as A 1。 The ABTS free radical scavenging rate formula is as follows:

[0094] S=(A0-A1) / A0×100%

[0095] 1.7.2. DPPH value determination

[0096] Prepare 0.47 mg / L DPPH solution, extract the test solution with methanol by reflux and dilute it 200 times as the sample solution. Mark different sample wells as A0, A1, and A2 respectively. React in the dark for 30 minutes, and measure the absorbance A value at 517 nm. Prepare 100 μL of methanol and 100 μL of DPPH mixed solution to measure the absorbance value as A0; prepare 100 μL of sample and 100 μL of DPPH mixed solution to measure the absorbance value as A1; prepare 100 μL of sample and 100 μL of anhydrous ethanol mixed solution to measure the absorbance value as A2. The formula for DPPH free radical scavenging rate is as follows:

[0097] S=(A0-A1-A2) / A0×100%

[0098] FRAP value determination

[0099] Take methanol reflux extraction solution and dilute it 20 times as sample solution, and operate according to the kit method. 2- The standard (No. 1 to No. 8) is used to draw a standard curve with the concentration of ferrous ions (mM) as the horizontal axis and the corresponding absorbance as the vertical axis. 2- The total antioxidant capacity can be expressed by the concentration. Therefore, the corresponding Fe 2- The antioxidant capacity of the sample can be known by the concentration.

[0100] 1.8. Correlation analysis of antioxidant spectrum and efficacy

[0101] 1.8.1. Grey relational analysis

[0102] Gray correlation analysis was used to conduct a sensitivity analysis of the common peak area data from the fingerprints of Tangut rhubarb samples and the antioxidant activity data of ABTS, DPPH, and FRAP. By calculating the correlation between the subsequences (sample fingerprint data) and the parent sequence (antioxidant activity index data), the influence of each component on antioxidant activity was determined, providing a theoretical basis for related research.

[0103] The specific steps are as follows: the three antioxidant index values ​​are used as the parent sequence X0(k), k=1, 2, …n (n is the number of samples); the chromatographic fingerprint data of the sample are used as subsequences, respectively recorded as X1(k), X2(k), X m (k); k = 1, 2, …n (m is the common peak number). The original data is transformed into a mean value, that is, each sequence element is divided by the average value of the corresponding sequence. The calculation process is shown in the following formula (1), thereby obtaining a homogenized sequence. Y i = , i=0, 1, …m.

[0104] ( 1 )

[0105] The correlation coefficient is calculated using formula (2):

[0106] ( 2 )

[0107] The correlation is calculated by formula (3)

[0108] ( 3 )

[0109] Where ρ is the resolution coefficient, 0≤ρ≤1, generally ρ=0.5;

[0110]

[0111] The correlation degree γ0, i ≥ 0.9 indicates that the subsequence has a significant impact on the parent sequence; 0.8 ≤ γ0, i < 0.9 indicates a relatively significant impact; 0.7 ≤ γ0, i < 0.8 indicates a significant impact; 0.6 ≤ γ0, i < 0.7 indicates a small impact; γ0, i < 0.6 indicates a very small impact.

[0112] 1.8.2. Partial Least Squares Regression Analysis

[0113] To investigate the quantitative relationship between the antioxidant activity and chemical composition of Rheum tanguticum, a partial least squares (PLS) regression analysis was performed using ABTS, DPPH, and FRAP values ​​as independent variables (indicating antioxidant activity) and the common peak area as the dependent variable (indicating chemical composition). PLS effectively addresses multicollinearity among independent variables and constructs association models. According to the principles of partial least squares regression analysis, a higher VIP value indicates a greater influence of the independent variable on the dependent variable. It is generally considered that an independent variable has a significant influence on the dependent variable when the VIP is greater than 1.

[0114] The three antioxidant activity indicators of ABTS, DPPH and FRAP were used as reference series X0(k), k=1, 2, ...n, and the m relative peak area values ​​of 18 batches of Tangut rhubarb were used as comparison series X0(k). i (k), k=1, 2, …18, i=0, 1, …m. All variables are dimensionless to eliminate order of magnitude errors. The mean value is used, that is, the variable value in each series is divided by the average value of the corresponding series to obtain the dimensionless series. The formula is referenced to (1). This study has a total of 18 dependent variables {y1, …, y n} and m independent variables {x1,…, x m}, partial least squares regression is used to extract the principal components t1 and u1 from x and y respectively, so that they can represent x and y in the original data table as much as possible, and the correlation between t1 and u1 is large enough. Partial least squares regression will continue to extract principal components and regress x on t and y on u until a good accuracy is obtained, and finally expressed as y n About the original variables {x1,…, x m This process can be implemented using a variety of software. In this study, SIMCA was used for partial least squares regression analysis. The partial regression coefficients and VIP values ​​were obtained through analysis to determine the explanatory power of the independent variables on the dependent variables.

[0115] 1.8.3. Pearson Correlation Analysis

[0116] This study used Origin software to perform Pearson correlation analysis, aiming to investigate the linear correlation and significance level between the common peaks of Rheum tanguticum and its antioxidant activity. Peak area data for the common peaks of Rheum tanguticum were used as independent variables, and data for ABTS, DPPH, and FRAP antioxidant activity indices were used as dependent variables. The software calculated the Pearson correlation coefficient for each independent and dependent variable using the formula. A two-sided test was used to determine the significance level of the correlation, calculating the corresponding p-value. If p > 0.05, there was insufficient evidence to reject the null hypothesis, indicating that the correlation was not significant. p ≤ 0.05 indicated a statistically significant difference (with a confidence level of ≥ 95%); p ≤ 0.01 indicated a statistically significant difference (with a confidence level of ≥ 99%); and p ≤ 0.001 indicated an extremely significant difference (with a confidence level of ≥ 99.9%).

[0117] Principal Component Analysis

[0118] The fingerprint data of Tangut Rhubarb has a high dimensionality, with multiple common peak area data and complex correlations between these data. Principal Component Analysis (PCA) is a commonly used multivariate statistical analysis method that aims to transform multiple related variables in the original data into a set of new, uncorrelated principal components through linear transformation, thereby simplifying the data structure and reducing data complexity while retaining most of the key information. . PCA calculates the covariance matrix of the data and extracts its eigenvalues ​​and eigenvectors, selecting the first few principal components with larger eigenvalues ​​to retain the data's variation information to the greatest extent. This method not only effectively reduces dimensionality but also removes noise from the data. In this study, PCA was used to analyze the characteristic peak data of the chemical components of Tangut rhubarb to extract the main variation information and reveal the intrinsic correlation between different components.

[0119] 1.9. Modeling of antioxidant activity using infrared spectroscopy

[0120] 1.9.1.TQanalyst Software Modeling

[0121] Utilizing 18 batches of Tangut rhubarb NIR spectra, totaling 54 spectral data, the samples were divided into a modeling set and an external validation set at a 4:1 ratio: 45 spectral data were used as the modeling set, and 9 data were used as the external validation set. Anomalous spectra were removed from the modeling set using Mahalanobis distance (MD) and PCA. Three antioxidant activity values ​​and their corresponding modeling set spectra were imported into the TQ analyst system in descending order of antioxidant activity. The modeling set ratio was optimized using calibration set:prediction set ratios of 2:1, 3:1, 4:1, and 5:1, to establish NIR spectral quantitative detection models for ABTS, DPPH, and FRAP. The modeling methods used included PLS and principal component regression (PCR). Preprocessing methods included: no preprocessing (Constant), multivariate scattering correction (MSC), standard normal variation (SNV), first derivative spectrum (1D), second derivative spectrum (2D), Savitzky-Golay smoothing (SG smoothing), and Norris derivative filter smoothing. A three-factor, three-level table was designed, as shown in Table 2. Modeling was performed for each combination as in a single-factor experiment. Norris smoothing, as a derivative filter, was performed on the first derivative spectrum, while the remaining preprocessing was performed on the original spectrum. SIMCA software was used to screen NIR spectral VIP modeling bands, and modeling band optimization was performed using bands with VIP > 1 and the entire spectrum. The NIR spectrum of the external validation sample was substituted into the optimal model to obtain the model-calculated value. The difference between the calculated value given by the model and the actual value was compared, and the accuracy of the model's prediction results for external validation was judged by the model prediction rate. The calculation formula is: prediction rate = 1-(|predicted value-measured value| / measured value×100%).

[0122] Finally, RMSE c (root mean square error of the calibration set), RMSE p (root mean square error of the prediction set), R cal (Correlation coefficient of calibration set), R val The prediction set correlation coefficient, RPD (residual prediction deviation), and external validation prediction rate were used as indicators to evaluate the modeling performance of each model. The smaller the root mean square error, the higher the correlation coefficient, external validation prediction rate, and RPD, the better the predictive ability of the model.

[0123]

[0124] 1.9.2. Python Software Modeling

[0125] There are five modeling methods in Python algorithm that are commonly used for quantitative regression and have good results, namely Bayesian Ridge Regression (BRR), Elastic Net Regression (ENR), Gaussian Process Regression (GPR), PLS, and Support Vector Machine Regression (SVR). Python software was used to divide the NIR spectrum of Tangut rhubarb samples. The same modeling set and external validation set as TQ software were used. The modeling set ratio was also optimized with a calibration set: prediction set ratio of 2:1, 3:1, 4:1, and 5:1. The modeling band was optimized using bands with VIP > 1 and all bands. The spectra were optimized using Norris smoothing, MSC, and D1 preprocessing methods. The NIR spectrum and antioxidant activity value data were modeled in sequence using the above five modeling methods. Finally, the RMSE was used to calculate the optimal value. c , RMSE p 、R cal 、R val , RPD, and external validation prediction rate were used as indicators to evaluate the modeling effect of each model.

[0126] Data Processing

[0127] The 2012 version of the "Traditional Chinese Medicine Chromatographic Fingerprint Evaluation System" was used to evaluate the similarity of fingerprints, TQanalyst and Python software were used for modeling, SIMCA 14.1 software was used to screen VIP>1 modeling bands, IBM SPSS Statistics 27.0.1 was used for principal component analysis and spectral effect analysis, and Origin 2021 was used for cluster analysis.

[0128] 2. Results and Analysis

[0129] 2.1. Fingerprint establishment and feature analysis

[0130] The results of the methodological tests are shown in Table 3. The RSDs of the retention times and content percentages of the common peaks in each verification were all less than 2%, indicating that the instrument precision was good, the repeatability and stability of the method were good, the common peak selection was reliable, and the established HPLC fingerprint determination method for Rhubarb tanguticum samples was feasible.

[0131]

[0132] According to the above chromatographic conditions, the HPLC fingerprint of Tanggut Rhubarb sample was established, as shown in Figure 2 As shown in the figure, the HPLC fingerprints of 18 batches of Tangut rhubarb samples were evaluated for similarity using the "Chinese Medicine Chromatographic Fingerprint Similarity Evaluation System (2012 Edition)" developed by the Chinese Pharmacopoeia Committee. The chromatographic fingerprints of the 18 batches of samples were imported into the system, a reference spectrum was set, and the median method was used to generate a reference fingerprint R. The similarity between the sample fingerprints and the reference fingerprint was calculated. The similarity between the samples was between 0.884 and 0.974. The generated reference fingerprint is shown in Figure 1 A total of 18 common peaks were identified, and the correlation coefficients between the common peaks and the reference fingerprints were all greater than 0.900, indicating that the common peaks can represent the characteristics of the sample components.

[0133] Figure 3 The chromatographic separation diagrams of the reference and sample were compared. A total of 13 common peaks corresponding to the chemical components were identified, among which peak 1 was aloe-emodin-8-O-glucoside, peak 2 was rhein-8-O-glucoside, peak 3 was sennoside B, peak 5 was sennoside C, peak 6 was sennoside A, peak 8 was chrysophanol-8-O-glucoside, peak 9 was rhein-8-O-glucoside, peak 11 was physophanol methyl ether-8-O-glucoside, peak 13 was aloe-emodin, peak 14 was rhein, peak 16 was emodin, peak 17 was chrysophanol, and peak 18 was physophanol methyl ether.

[0134] Cluster analysis was used to group samples from different origins based on the similarity of their chemical compositions. This visually demonstrates the differences and similarities in the chemical composition of Rheum tanguticum from different origins, providing a deeper understanding of the key quality characteristics of this medicinal material from different origins. Using the common peak area as the variable and the Euclidean distance as the interval, the within-group uniform linkage method was used to cluster 18 batches of Rheum tanguticum samples from different origins. When the distance is 0.12, the samples are clustered into two major groups, among which S1, S2, S3, S5, S6, S7, and S8 are clustered into one category, and the main production areas are Dari County, Golog Prefecture, Banma County, Golog Prefecture, Zeku County, Huangnan Prefecture, and Tongren County, Huangnan Prefecture; the remaining 11 production areas (S4, S9, S11, S12, S13, S15, S16, S17, S18, S10, and S14) are clustered into one category, and the main production areas are Qilian County, Haibei Prefecture, and Huangzhong County, Xining City. The clustering results are as follows: Figure 4 The two major groups are geographically distributed in a north-south direction, with the Laji Mountains in Qinghai Province as the dividing line, and there are certain regional differences.

[0135] 2.2. Antioxidant activity assay

[0136] The results of antioxidant activity assay were as follows Figure 5 As shown in the figure, the ABTS free radical scavenging rates of R. tanguticum ranged from 61.69% to 92.51%, the DPPH free radical scavenging rates ranged from 69.19% to 86.62%, and the FRAP antioxidant capacity values ​​ranged from 0.743 to 1.242. The ABTS and DPPH free radical scavenging rates were both above 60%, and the FRAP antioxidant capacity was above 0.700. Samples from different batches exhibited good antioxidant activity, and rhubarb from different origins showed roughly similar trends in all three antioxidant activity indicators. However, it was observed that the antioxidant capacity of R. tanguticum samples from different origins and growth environments varied to some extent. Samples from batches 1, 3, and 8 (i.e., Dari County, Golog Prefecture, Banma County, Golog Prefecture, and Tongren City, Huangnan Prefecture) had higher in vitro antioxidant activity than those from other regions. All three regions are located south of the Laji Mountains and have a relatively high average altitude. It is speculated that environmental factors such as altitude and temperature may contribute to the varying content of antioxidant-related components in Tangut rhubarb from these regions, leading to its stronger antioxidant capacity. Studies have shown significant geographic variation in the active components of Tangut rhubarb, with high temperatures, high sunshine, and low precipitation in cooler regions favoring the formation and accumulation of combined anthraquinones and polyphenols. Combined anthraquinones and polyphenols exhibit significant antioxidant activity, and these findings are consistent with the antioxidant activity test results.

[0137] The antioxidant activity of samples from different origins was analyzed by one-way ANOVA, and the results are shown in Table 4. The results showed that the antioxidant activity of samples from different origins was significantly different (P < 0.001), further indicating that the content of the main antioxidant components in samples from different origins was different. Therefore, it is necessary to conduct antioxidant quality control on Tangut rhubarb.

[0138]

[0139] 2.3. Analysis of the antioxidant spectrum-effect relationship of Tangut rhubarb in vitro

[0140] 2.3.1 Principal Component Analysis

[0141] The peak areas of 18 common peaks in 18 batches of Tangut rhubarb samples were imported into SPSS software for principal component analysis. After rotation, the initial eigenvalues ​​and the sum of squares of the rotation loads were obtained. The results of the eigenvalues ​​and variance contribution rates are shown in Table 5. The first three principal components were extracted, and their cumulative contribution rate was 82.69%, which is in line with the principle of providing the most information represented by the original data with a cumulative contribution rate greater than 80% and using the least number of principal components. The common peak area data were classified according to the cluster analysis results and imported into OMANIC software for PCA analysis. The results are shown in Table 5. Figure 6The 18 batches of Rheum tanguticum extracts could be divided into two categories, which was consistent with the results of cluster analysis.

[0142]

[0143] As shown in Table 5, the first principal component had the highest contribution rate, reaching 40.37%. Peak 5 (sennoside C), peak 7, peak 9 (emodin-8-O-glucoside), peak 10, and peak 12 had relatively high positive loading factors in this principal component. Among them, peak 5 (sennoside C) and peak 9 (emodin-8-O-glucoside) were both conjugated anthraquinone derivatives (glycosides). This type of component is the main active substance in rhubarb that exerts a purgative effect, indicating that this principal component is associated with the pharmacological activity of rhubarb, such as purgative effect. It can be seen that the first principal component mainly reflects the expression level of conjugated anthraquinone / dianthrone components in rhubarb that are related to the purgative effect. The contribution rate of the second principal component reached 24.62%. Peak 2 (rhein-8-O-glucoside) and peak 4 had high positive loads, both of which were bound anthraquinone active ingredients; while peak 13 (aloe-emodin), peak 14 (rhein), peak 16 (emodin), peak 17 (chrysophanol), and peak 18 (physcion methyl ether) had high negative loads. These five components were all free anthraquinone chemical components. It can be seen that the second principal component mainly reflects the difference in the ratio of bound anthraquinone to free anthraquinone in Rhubarb tanguticum. Studies have shown that the laxative effect of bound anthraquinone is better than that of free anthraquinone. ] This difference in ratio is closely related to the processing method and the pharmacological activity of bound and free anthraquinones, and may distinguish the clinical application tendency of the medicinal material (such as mild or severe laxative). Within the third principal component, peaks 2 and 3 (sennoside B), 4 and 14 (rhein), and 16 (emodin) showed high scores. These components have strong antibacterial and antioxidant effects, reflecting the medicinal material's tendency towards antibacterial and antioxidant efficacy.

[0144] 2.3.2. Partial Least Squares Regression Analysis

[0145] PLS analysis was performed on the peak area data of 18 common peaks and the data of three antioxidant indicators. The results are shown in Figure 7By analyzing the VIP values ​​and partial regression coefficients of three antioxidant activity indices for 18 batches of Tangut rhubarb fingerprint data and comparing the PLS results of the three antioxidant indices, we found that the VIP values ​​of peak 2 (rhein-8-O-glucoside), peak 4, peak 17 (chrysophanol), and peak 18 (physcion methyl ether) were all greater than 1, indicating that the substances corresponding to these four common peaks have a significant impact on the in vitro antioxidant activity of Tangut rhubarb. Combining the correlation coefficients and VIP values, we found that peaks 2 and 4 were positively correlated with the three antioxidant indices, and their VIP values ​​were all greater than 1, indicating that these two active ingredients play a significant role in promoting the antioxidant activity of Tangut rhubarb.

[0146] Pearson correlation analysis

[0147] Pearson correlation analysis was performed using Origin 2021 software, and the results are as follows Figure 8 Peak 2 (rhein-8-O-glucoside) showed a highly significant positive correlation with ABTS free radical scavenging activity, DPPH free radical scavenging activity, and FRAP antioxidant capacity (p ≤ 0.01), while peak 4 showed a highly significant positive correlation with DPPH free radical scavenging activity and FRAP antioxidant capacity (p ≤ 0.01). Peak 18 (physcion methyl ether) showed a significant negative correlation with FRAP antioxidant capacity (p ≤ 0.05). These results are consistent with the results of PLS ​​analysis. The chemical structure of peak 2 rhein-8-O-glucoside contains four phenolic hydroxyl groups and one carboxyl group, which may contribute to its significant antioxidant contribution. Peak 18 physcion methyl ether showed a significant negative correlation with FRAP, which may be related to the ability of iron ions to form complexes with physcion methyl ether.

[0148] 2.3.4. Grey relational analysis

[0149] To further explore the correlation between the common peaks and antioxidant indicators, the peak area data of 18 common peaks and the data of three antioxidant indicators were subjected to grey correlation analysis. The results are shown in Table 6.

[0150]

[0151] The correlation of each common peak with ABTS, DPPH and FRAP was greater than 0.720, indicating that each common peak had a greater correlation with antioxidant activity, which explained that the antioxidant activity of R. tanguticum was a synergistic effect of multiple components rather than a single dominant substance. The ABTS scavenging rate of the components corresponding to each common peak of R. tanguticum was ranked in descending order as follows: peak 2> peak 8> peak 1> peak 7> peak 11> peak 16> peak 3> peak 6> peak 14> peak 5> peak 13> peak 9> peak 15> peak 18> peak 17> peak 10> peak 4> peak 12. Among them, peak 2 (rheic acid-8-O-glucoside), peak 8 (chrysophanol-8-O-glucoside), peak 1 (aloe emodin-8-O-glucoside), peak 7, peak 11 (physcion-8-O-glucoside), peak 16 (emodin), peak 3 (sennoside B), peak 6 (sennoside A), peak 14 (rheic acid), peak 5 (sennoside C), and peak 13 (aloe emodin) had a relatively significant influence (0.8≤γ0, i<0.9) on the ABTS scavenging rate. Peak 9 (emodin-8-O-glucoside), peak 15, peak 18 (physcion), peak 17 (chrysophanol), peak 10, peak 4, and peak 12 had a significant influence (0.7≤γ0, i<0.8) on the ABTS scavenging rate.

[0152] The effects of the components corresponding to the common peaks of Rheum tanguticum on the DPPH clearance rate were ranked from large to small as follows: peak 8 > peak 2 > peak 1 > peak 7 > peak 11 > peak 3 > peak 6 > peak 16 > peak 14 > peak 5 > peak 9 > peak 13 > peak 15 > peak 17 > peak 18 > peak 10 > peak 4 > peak 12. Among them, peak 8 (chrysophanol-8-O-glucoside), peak 2 (rhein-8-O-glucoside), peak 1 (aloe-emodin-8-O-glucoside), peak 7, peak 11 (physcion-8-O-glucoside), peak 3 (sennoside B), peak 6 (sennoside A), peak 16 (emodin), peak 14 (rhein), peak 5 (sennoside C), peak 9 (emodin-8-O-glucoside), peak 13 (aloe-emodin), and peak 15 had relatively significant effects on DPPH scavenging rate (0.8≤γ0, i<0.9); components peak 17 (chrysophanol), peak 18 (physcion), peak 10, peak 4, and peak 12 had significant effects on DPPH scavenging rate (0.7≤γ0, i<0.8).

[0153] The influence of the components corresponding to each common peak on the FRAP antioxidant activity was ranked from large to small as follows: peak 2 > peak 1 > peak 8 > peak 7 > peak 11 > peak 3 > peak 16 > peak 6 > peak 13 > peak 5 > peak 14 > peak 15 > peak 9 > peak 17 > peak 18 > 4 > peak 10 > peak 12. Among them, peak 2 (rhein-8-O-glucoside), peak 1 (aloe-emodin-8-O-glucoside), peak 8 (chrysophanol-8-O-glucoside), peak 7, peak 11, peak 3 (sennoside B), peak 16 (rhein), peak 6 (sennoside A), peak 13 (aloe-emodin), peak 5 (sennoside C), and peak 14 (rhein) had relatively significant effects on the FRAP antioxidant capacity (0.8≤γ0, i<0.9); peak 15, peak 9 (rhein-8-O-glucoside), peak 17 (chrysophanol), peak 18 (physcionol methyl ether), peak 4, peak 10, and peak 12 had significant effects on the FRAP antioxidant capacity (0.7≤γ0, i<0.8).

[0154] Comprehensive comparison of the three antioxidant activity correlation results ( Figure 9 ), the changing trends and inflection points of the three antioxidant indices were largely consistent, indicating that the common components in Rhubarb tangutica have a similar impact on the three antioxidant activity indicators. Peaks 2 and 8 have the greatest antioxidant effects, while peaks 4, 10, and 12 have lesser effects. Furthermore, bound anthraquinones, such as chrysophanol-8-O-glucoside, rhein-8-O-glucoside, aloe-emodin-8-O-glucoside, and physcion-8-O-glucoside, dominate the antioxidant activity. This is consistent with related research findings showing a significant positive correlation between total anthraquinone glycosides and antioxidant activity. Furthermore, some studies have shown that bound anthraquinones have a superior purgative effect to free anthraquinones. This suggests that an appropriate amount of glycosides must be retained during processing of Rhubarb tangutica to maintain its purgative pharmacological activity.

[0155] 2.4. Establishment of an infrared antioxidant model for Rhubarb and rapid evaluation of its antioxidant activity based on near-infrared spectroscopy

[0156] 2.4.1. Infrared spectral feature analysis

[0157] NIR spectra of 18 batches of Tangut rhubarb samples Figure 10 , a total of 6 absorption peaks were obtained, namely 8370 cm -1 、6337cm -1 、5886 cm -1 4825cm -1 , 4400 cm -1 and 4277 cm -1 . Assign the common peaks [ , 8370 cm -1 Nearby is the second harmonic absorption peak of CH2 (anthraquinone nucleus and glycosyl methyl / methylene); 6337 cm -1 Nearby is the first harmonic absorption peak of OH (glycoside hydroxyl and free anthraquinone phenol hydroxyl); 5886 cm -1 It is the first harmonic absorption peak of CH of anthraquinone aromatic ring; 4825cm -1 Nearby is the combined absorption peak of OH (glycoside and polysaccharide hydroxyl); 4400 cm -1 Nearby is the OH and CO stretching combination frequency (characteristic of the combined anthraquinone glycosidic bond); 4277 cm -1 The peaks at the bottom of the spectrum are the second harmonics of the fundamental frequency of the CH bending vibration of polysaccharides (cellulose / starch components). The above characteristic peaks can comprehensively reflect the chemical structure of the anthraquinone components and the distribution characteristics of polysaccharides in rhubarb.

[0158] 2.4.2. Establishment of near-infrared spectroscopy evaluation model

[0159] 2.4.2.1. Outlier Removal

[0160] The abnormal spectra corresponding to the three antioxidant index models were removed, and the Mahalanobis distance and principal component score of the spectral data were shown in Figure 2. Figure 11 The MD method showed that the MD values ​​of all modeled spectra were less than 1.8, indicating that there were no anomalous spectra using this method. The PCA score diagnostic results showed that all samples were within the 95% confidence interval. Therefore, the near-infrared spectra of all samples were normal using both methods and no exclusion was required.

[0161] 2.4.2.2. Modeling band selection

[0162] The original spectrum used for model building usually includes all measured wavelengths, but the full spectrum model contains some redundant information, which may have a negative impact on the model's predictive ability. Wavelength selection can improve the performance of the calibration model. Therefore, the VIP method is used to select the bands of the modeling set, and the resulting score map is shown in Figure 12 .

[0163] When modeling, first perform full-band (10000~4000 cm -1 ) model, and then select the band 5296~4000 cm with a VIP value greater than 1 -1 、6399~6244 cm -1 Modeling band optimization was performed. Using different software, each indicator was optimized within the modeling range using the specified preprocessing and modeling methods. The results are shown in Table 7. As can be seen, ABTS and DPPH achieved better modeling results across the entire band than those in the band with VIP > 1 using both different analysis software, while FRAP achieved the best modeling results in the band with VIP > 1.

[0164]

[0165] 2.4.2.3. Sample set partitioning

[0166] TQ Analyst software was used to partition the modeling set using a concentration gradient method, while Python software was used to randomly partition the modeling set. The calibration set:prediction set ratio was optimized for each of the two methods. The results are shown in Table 8. The results show that using TQ analyst software to partition the sample using a gradient method, the three antioxidant index models all achieved the best modeling results when the sample set was partitioned at a 5:1 ratio. Python software randomly partitioned the modeling set based on the ratio, and the results showed that the ABTS and DPPH models both performed best when the sample set was partitioned at a 3:1 ratio, while the FRAP model performed best when the sample set was partitioned at a 4:1 ratio.

[0167]

[0168] 2.4.3. Selection of spectral pretreatment and modeling method

[0169] The NIR spectra and ABTS, DPPH, FRAP content data were input into TQ analyst and Python software, and the optimal modeling wavelength and modeling ratio were used for modeling. The modeling results under different modeling software, pretreatment methods, and modeling methods were compared, and the optimal modeling results were selected to determine the best pretreatment and modeling method. The results are shown in Table 9.

[0170]

[0171] The model established by PLS on the NIR spectrum pretreated by 1D+Norris smoothing+MSC method under Python software showed the best effect for ABTS, as shown in Figure 13 (a). The model showed low RMSE c (0.016) and RMSE p (0.043), and high R cal (0.982) and R val (0.937). The RPD value was 2.190, and the external validation prediction rate reached 95.22%, indicating that the model had good prediction ability and robustness.

[0172] Similarly, under Python software, the model established by PLS method on the spectrum pretreated by 1D+Norris smoothing+MSC showed the best effect for DPPH antioxidant index, as shown in Figure 13 (b). The model had low RMSE c (0.009) and RMSE p (0.023), and high R cal (0.985) and R val (0.871). The RPD value was 2.284, and the external validation prediction rate was 94.83%, further verifying the accuracy and reliability of the model.

[0173] The model established by GPR method on the spectrum pretreated by 1D+Norris smoothing+MSC showed the best effect for FRAP antioxidant index under Python software, as shown in Figure 13 (c). The model showed low RMSE c (0.011) and RMSE p (0.071), and high R cal (0.997) and R valAlthough the RPD value was slightly lower (1.803), the external validation prediction rate still reached 88.55%.

[0174] The average prediction rate for all external validation indicators exceeded 85%, with the ABTS and DPPH models achieving prediction rates exceeding 90%, demonstrating that the developed models can provide excellent predictions for unknown samples. Because NIR spectroscopy provides a more comprehensive representation of the sample's chemical composition and structural information, it can accurately predict antioxidant activity. The free radical scavenging mechanism of ABTS / DPPH may rely more on linear additive effects. Because band importance screening (PLS) effectively balances information retention with model complexity, it offers more robust predictions. FRAP's reducibility may be related to the synergistic or antagonistic effects of various functional groups (such as phenolic hydroxyl groups and thiol groups), and its spectral response may exhibit nonlinear characteristics. Therefore, GPR modeling is most effective, as global band noise may interfere with its performance. The Python software's partitioning of the modeling and calibration sets does not proceed in descending order of antioxidant activity, but rather randomly, which may be the primary reason for the discrepancy between the two software models.

[0175] 2.5 Summary

[0176] This example used cluster analysis, PCA, PLS, Pearson correlation analysis, and infrared spectroscopy to analyze 18 batches of rhubarb samples from different geographic locations, aiming to further explore the relationship between their main components, infrared spectra, and antioxidant activity. Both cluster analysis and PCA successfully divided the samples into two groups: samples from Guoluo Prefecture, Haibei Prefecture, and Huangnan Prefecture were clustered into one group, while samples from Huangzhong County, Xining City, were clustered into a separate group.

[0177] PLS analysis combined with Pearson correlation analysis further revealed the material basis of the antioxidant activity of Rheum tanguticum. PLS results showed that common peaks 2, 4, 17, and 18 had high variable importance values ​​(VIP>1) in the PLS analysis. Pearson correlation analysis further confirmed the significant correlations between these common peaks and antioxidant activity indicators. Common peak 2 (rhein-8-O-glucoside) showed extremely significant positive correlations with ABTS, DPPH, and FRAP (p≤0.01), common peak 4 showed extremely significant positive correlations with DPPH and FRAP (p≤0.01), and common peak 18 (physcion methyl ether) showed a significant negative correlation with FRAP (p≤0.05).

[0178] Furthermore, near-infrared (NIR) spectroscopy revealed the main chemical components of Rheum tanguticum, including anthraquinones, polysaccharides, and tannins, and confirmed their primary functional groups. Based on the IR spectral data, this study established quantitative detection models for the antioxidant activities of ABTS, DPPH, and FRAP. The models developed using a single NIR spectrum demonstrated excellent predictive performance, with externally validated prediction rates of 95.22%, 94.83%, and 91.53%, respectively.

[0179] Discussion

[0180] This example uses HPLC to establish liquid phase fingerprints of Rheum tanguticum from different origins, and attributes the common peaks to clarify the chemical components represented by each common peak. Cluster analysis results show that the sampling locations are distributed north-south, with the Laji Mountains as the dividing line, Qilian County and Huangzhong County in the north, and Banma County, Zeku County, and Tongren County in the south. Spectrum-effect relationship analysis reveals that its antioxidant activity is a multi-component combined effect. The grey correlation values ​​of peak 8 (chrysophanol-8-O-glucoside), peak 2 (rhein-8-O-glucoside), peak 1 (aloe-emodin-8-O-glucoside), peak 7, peak 11 (physcion-8-O-glucoside), peak 3 (sennoside B), peak 6 (sennoside A), peak 16 (emodin), peak 14 (rhein), peak 5 (sennoside C), peak 9 (emodin-8-O-glucoside), and peak 13 (aloe-emodin) were high, indicating that the synergistic effect of bound anthraquinone glycosides and free anthraquinones was the main antioxidant force. PLS and Pearson correlation analysis showed that peak 2 (rhein-8-O-glucoside), peak 4, peak 17 (chrysophanol), and peak 18 (physophanol methyl ether) all had VIP values ​​greater than 1. Peak 2 (rhein-8-O-glucoside) showed extremely significant positive correlations with ABTS free radical scavenging activity, DPPH free radical scavenging activity, and FRAP antioxidant capacity (p ≤ 0.01), peak 4 showed extremely significant positive correlations with DPPH free radical scavenging activity and FRAP antioxidant capacity (p ≤ 0.01), and peak 18 (physophanol methyl ether) showed a significant negative correlation with FRAP antioxidant capacity (p ≤ 0.05). This suggests that these compounds have a high explanatory weight for antioxidant activity and can be used as priority markers for quality control.

[0181] Among the constructed NIR quantitative antioxidant activity models, Python software performed best for all three antioxidant indicators. The optimal models for ABTS (RPD: 2.190), DPPH (RPD: 2.284), and FRAP (RPD: 1.803) all enabled rapid and accurate evaluation of the in vitro antioxidant activity of Rheum tanguticum. This study provides a scientific basis for improving the quality standards and clinical application of Rheum tanguticum. Further research will be needed to deepen the study of its chemical composition and the molecular mechanisms of its antioxidant activity, optimize modeling strategies, and fully explore the antioxidant potential of Rheum tanguticum.

[0182] Example 3

[0183] Based on the same inventive concept as Example 1, a spectral efficacy analysis system for Tangut Rhubarb based on infrared spectroscopy is provided, comprising:

[0184] The sample preparation module is used to collect Rheum tanguticum samples and prepare sample solutions;

[0185] A fingerprint acquisition module, used for collecting the high performance liquid chromatography fingerprint of the sample solution;

[0186] A spectrum acquisition module, used for acquiring the NIR spectrum of the Tangut rhubarb sample;

[0187] a quantitative model building module, for determining the antioxidant activity of the sample solution and establishing a quantitative model of antioxidant activity using the collected spectral data;

[0188] The spectrum-effect analysis module is used to screen the antioxidant active ingredients of Rhubarb tanguticum based on the high-performance liquid chromatography fingerprint, and to detect the antioxidant active ingredients of Rhubarb tanguticum using the antioxidant activity quantitative model.

[0189] It should be noted that each module of the system utilizes the means and methods used in each step of the aforementioned method to achieve corresponding functions, which will not be elaborated here.

[0190] The above specific implementation methods are detailed descriptions of the present invention. It cannot be considered that the specific implementation methods of the present invention are limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, they can make several simple deductions and substitutions without departing from the concept of the present invention, which should be regarded as falling within the scope of protection of the present invention.

Claims

1. A spectral analysis method for Tangut Rhubarb based on infrared spectroscopy, characterized in that: The following steps are involved: S1. Collect Rheum tanguticum samples and prepare sample solutions; S2. collecting a high performance liquid chromatography fingerprint of the sample solution; S3, collecting the NIR spectrum of the Rhubarb Tangut sample; S4, determining the antioxidant activity of the sample solution, and establishing a quantitative antioxidant activity model using the spectral data collected in step S3; S5. Based on the high performance liquid chromatography fingerprint in step S2, the antioxidant active ingredients of Rheum tanguticum were screened, and the antioxidant active ingredients of Rheum tanguticum were detected using an antioxidant activity quantitative model.

2. The infrared spectroscopy-based spectral analysis method for Rhubarb Tangut according to claim 1, wherein: The method of collecting Tangut rhubarb samples comprises: Tangut rhubarb samples were collected from various places in Qinghai Province from September to October. The roots of the collected Tangut rhubarb samples were washed, sliced, dried in the shade, coarsely crushed, passed through an 80-mesh sieve, and placed in a desiccator for later use.

3. The infrared spectroscopy-based spectral analysis method for Rhubarb Tangut according to claim 1, wherein: The preparation of the sample solution comprises: Take 0.5000±0.0001g of Tangut Rhubarb sample powder, place it in a stoppered conical flask, accurately add 25mL of methanol, heat and reflux for 1h, cool, transfer to a centrifuge tube and centrifuge at 4000rpm*15min, then use a volumetric flask to make up the volume, add methanol to make up the lost weight, shake well, and take the filtrate.

4. The infrared spectroscopy-based spectral analysis method for Rhubarb Tangut according to claim 1, wherein: The collecting of the high performance liquid chromatography fingerprint of the sample solution comprises: Chromatographic conditions: Eclipse Plus C column 18 (4.6 mm × 250 mm, ID 5 μm), mobile phase: acetonitrile (A)-0.1% phosphoric acid aqueous solution (B), gradient elution program: 0-15 min, 2%→15%A; 15-25 min, 15%→20%A; 25-40 min, 20%→20%A; 40-70 min, 20%→30%A; 70-90 min, 30%→55%A; 90-100 min, 55%→75%A; injection volume: 10 μL; detection wavelength: 254 nm; column temperature: 30°C; volume flow rate: 0.8 mL / min; Precision tests, repeatability tests, stability tests and fingerprint similarity evaluations were carried out under the above chromatographic conditions.

5. The infrared spectroscopy-based spectral analysis method for Rhubarb Tangut according to claim 1, wherein: The NIR spectrum acquisition includes: The NIR one-dimensional infrared spectrum of Tangut rhubarb sample was collected using the NIR fiber of the Fourier transform infrared spectrometer, with 32 scans and a resolution of 8 cm -1 , the spectrum acquisition range is 10000-4000cm -1 Each sample was scanned three times, the air background value was deducted before each scan, and three spectra of each sample were taken and the average spectrum was obtained.

6. The infrared spectroscopy-based spectral analysis method for Rhubarb Tangut according to claim 1, wherein: Determining the antioxidant activity of the sample solution comprises: ABTS value determination, DPPH value determination and FRAP value determination.

7. The infrared spectroscopy-based spectral analysis method for Rhubarb Tangut according to claim 1, wherein: The method of establishing a quantitative model for antioxidant activity comprises: Establish a quantitative detection model corresponding to different antioxidant activity values ​​under NIR spectroscopy.

8. The infrared spectroscopy-based spectral analysis method for Rhubarb Tangut according to claim 7, wherein: The screening of the antioxidant medicinal components of Rhubarb tangutica comprises: Cluster analysis, principal component analysis, partial least squares regression analysis and Pearson correlation analysis were performed on the antioxidant efficacy of Rhubarb tanguticus.

9. A spectral analysis system for Tangut rhubarb based on infrared spectroscopy, characterized in that: include: The sample preparation module is used to collect Rheum tanguticum samples and prepare sample solutions; A fingerprint acquisition module, used for collecting the high performance liquid chromatography fingerprint of the sample solution; A spectrum acquisition module, used for acquiring the NIR spectrum of the Tangut rhubarb sample; a quantitative model building module, for determining the antioxidant activity of the sample solution and establishing a quantitative model of antioxidant activity using the collected spectral data; The spectrum-effect analysis module is used to screen the antioxidant active ingredients of Rhubarb tanguticum based on the high-performance liquid chromatography fingerprint, and to detect the antioxidant active ingredients of Rhubarb tanguticum using the antioxidant activity quantitative model.

Images