A system for analyzing medicinal effects of a Mongolian medicine mei flower

Abstract

The application discloses a Mongolian medicine Meihuacao efficacy test analysis system and belongs to the technical field of Mongolian medicine efficacy analysis. The system comprises the following modules: a characteristic acquisition module, which is used for acquiring multi-source physicochemical characteristics of a Mongolian medicine Meihuacao sample, and obtaining chromatographic characteristic data, mass spectrometric characteristic data and spectral characteristic data; a spatial mapping module, which is used for performing characteristic spatial mapping on the characteristic data, and generating a multi-dimensional physicochemical characteristic space point set; a clustering analysis module, which is used for performing dynamic clustering analysis on the multi-dimensional physicochemical characteristic space point set, and obtaining a physicochemical characteristic dynamic clustering result; a path selection module, which is used for performing multi-level efficacy test path selection on the Mongolian medicine Meihuacao sample based on the physicochemical characteristic dynamic clustering result, and determining a target efficacy test path; an efficacy determination module, which is used for determining an efficacy index according to the target efficacy test path, and obtaining a multi-dimensional efficacy response spectrum; and a report generation module, which is used for generating an efficacy analysis report based on the multi-dimensional efficacy response spectrum.

Description

Technical Field

[0001] This invention relates to the field of Mongolian medicine efficacy analysis technology, specifically a Mongolian medicine plum blossom herb efficacy test analysis system. Background Technology

[0002] The efficacy testing of the Mongolian medicinal herb *Meihuacao* is a necessary step in evaluating its biological activity. Current techniques typically employ fixed-time-point efficacy assays, collecting samples at several predetermined discrete time points for cell activity detection. This approach relies on empirically set sampling intervals and cannot adjust the experimental strategy based on the dynamic evolution of the sample's physicochemical characteristics.

[0003] Existing technologies suffer from the following drawbacks: The high-dimensional heterogeneous data generated by multi-source physicochemical detection technologies are isolated from each other. Chromatographic, mass spectrometric, and spectral data each reflect different dimensions of material information, lacking an effective integration mechanism to construct a unified characteristic description that comprehensively characterizes the chemical composition of the sample. This results in subsequent analyses being unable to simultaneously utilize the complementary information from multimodal data. Furthermore, there is a lack of dynamic linkage between the determination of the efficacy testing pathway and changes in the physicochemical characteristics of the sample. Fixed testing protocols cannot respond to differences in component dissolution or chemical transformation rates among different samples, leading to the omission of key efficacy change nodes or the waste of experimental resources during ineffective time periods.

[0004] This invention addresses how to map the chromatographic, mass spectrometric, and spectral multi-source physicochemical characteristics of Mongolian medicinal herb Plum Blossom Grass samples to a unified multidimensional space for collaborative characterization, and how to adaptively determine multi-level efficacy test paths based on the dynamic clustering evolution of the sample's physicochemical characteristics over time. Summary of the Invention

[0005] The purpose of this invention is to provide a Mongolian medicine Meihuacao herbal efficacy test analysis system, which maps the chromatographic, mass spectrometric, and spectral multi-source physicochemical characteristics of Mongolian medicine Meihuacao samples to a unified multi-dimensional space for collaborative characterization, and adaptively determines multi-level efficacy test paths based on the dynamic clustering evolution of the sample physicochemical characteristics on the time axis.

[0006] To achieve the above objectives, the present invention provides the following technical solutions: The present invention provides a Mongolian medicine Parnassia herb pharmacodynamic test analysis system, which includes a feature acquisition module, a spatial mapping module, a cluster analysis module, a path selection module, a pharmacodynamic determination module, and a report generation module. The feature acquisition module is used to perform multi-source physicochemical feature acquisition on Mongolian medicine Parnassia herb samples to obtain chromatographic feature data, mass spectrometric feature data, and spectral feature data. By performing ultrasonic-assisted extraction on the samples and using an ultra-high performance liquid chromatograph, a quadrupole time-of-flight mass spectrometer and a near-infrared spectrometer for synchronous acquisition, the physicochemical properties of Mongolian medicine Parnassia herb can be comprehensively characterized from multiple dimensions such as chromatographic retention behavior, precise mass of parent ions, fragment ion abundance distribution and near-infrared spectral absorption, ensuring the completeness and complementarity of the feature information.

[0007] The spatial mapping module receives the chromatographic feature data, mass spectrometric feature data and spectral feature data, performs feature spatial mapping, and generates a multi-dimensional physicochemical feature spatial point set of the Mongolian medicine Parnassia herb samples. As a preferred technical solution of the present invention, in this mapping process, the spatial mapping module extracts the retention time data of chromatographic peaks as the first spatial dimension coordinate, extracts the chromatographic peak area data as the second spatial dimension coordinate, extracts the precise mass number of the parent ion in the first-stage mass spectrometry as the third spatial dimension coordinate, extracts the mass-to-charge ratio corresponding to the fragment with the maximum abundance among the fragment ions in the second-stage mass spectrometry as the fourth spatial dimension coordinate, and extracts the absorption intensity value of the spectral feature data at the preset characteristic wavenumber position as the fifth spatial dimension coordinate, combines the above coordinates into a five-dimensional physicochemical feature vector at each sampling time point, and arranges them in chronological order to generate the multi-dimensional physicochemical feature spatial point set. This mapping method unifies the instrument analysis signals from different sources into the same high-dimensional feature space, realizes a complete digital description of the evolution of the physicochemical state of the Mongolian medicine Parnassia herb samples over time, and provides an information basis for subsequent clustering and path decision-making.

[0008] The cluster analysis module performs dynamic cluster analysis on the multi-dimensional physicochemical feature spatial point set to obtain a dynamic clustering result of physicochemical features. Preferably, this module adopts an initial clustering center selection strategy and iterative update mechanism based on the maximum-minimum distance. After randomly selecting the first clustering center from the point set, it continuously selects the vector with the maximum distance from the existing clustering centers as a new clustering center until the preset number of clusters is met, and then performs iterative assignment and mean update until the offset of the clustering centers converges. This process can automatically discover the natural grouping structure of the physicochemical features of the samples during the test, merge the time points with similar physicochemical properties into the same cluster, so as to reveal the steady state or transition state presented by Mongolian medicine Parnassia herb in different test stages, and provide a reliable basis for the accurate selection of subsequent pharmacodynamic test paths.

[0009] The path selection module, guided by the dynamic clustering results of physicochemical characteristics, performs multi-level efficacy test path selection for Mongolian medicine *Meihuacao* samples to determine the target efficacy test path. Specifically, this module parses the cluster label sequence in the dynamic clustering results, identifies steady-state feature vectors that appear continuously for a duration exceeding a preset threshold, and uses their corresponding sampling time points as key test time nodes. A preset efficacy test library stores multiple historical *Meihuacao* efficacy test paths and their preset test node timestamps. The path selection module aligns the extracted key test time nodes with the preset test nodes of each candidate path along the time axis, calculates the cumulative time deviation, and selects the candidate path with the smallest cumulative time deviation as the target efficacy test path. This scheme intelligently matches the dynamic evolution of the sample's own physicochemical characteristics with the standardized efficacy test process, avoiding the shortcomings of fixed test paths that may ignore individual sample differences. This makes the test node settings more closely match the actual reaction process of the sample, improving the timeliness and specificity of efficacy evaluation.

[0010] The efficacy assay module measures the efficacy indicators of *Melilotus sibiricum* samples according to the target efficacy test path, obtaining a multidimensional efficacy response spectrum. As a technical solution of this invention, at each preset test node specified in the target path, the module extracts an equal volume of sample solution from the *Melilotus sibiricum* sample solution and adds it to different wells of a pre-cultured cell line plate containing human liver cancer cells, human breast cancer cells, and normal human liver cells. After culture, cell viability detection reagents are added and absorbance values ​​are read. The absorbance values ​​at all test time points are arranged according to time sequence and cell line type to construct a three-dimensional data matrix of time-cell line-absorbance, forming a multidimensional efficacy response spectrum. This spectrum simultaneously includes information on the inhibitory effect on the proliferation of different tumor cell lines and the cytotoxicity to normal cell lines, providing intuitive and multidimensional quantitative data for evaluating the efficacy and selectivity of *Melilotus sibiricum*.

[0011] The report generation module generates a pharmacodynamic analysis report for Mongolian herb *Meihuacao* based on a multidimensional pharmacodynamic response map. Preferably, this module extracts absorbance values ​​for human liver cancer cell lines, human breast cancer cell lines, and normal human liver cell lines at each experimental time point from the multidimensional pharmacodynamic response map, calculates the absorbance change rate sequence between adjacent time points, and generates a comparative change rate table. The values ​​in the comparative change rate table are compared with a reference range of standard pharmacodynamic change rates derived from historical data of the same batch of standard products. The experimental time points and cell line types corresponding to abnormal change rates exceeding the reference range are marked. The comparative change rate table and the marking results are then structured and arranged to generate a pharmacodynamic analysis report containing columns for abnormal markers. By automatically comparing the measured change rates with the standard range, the system can quickly identify key nodes where pharmacodynamics significantly deviate from expectations and sensitive cell types. The analysis results are clear and intuitive, facilitating researchers to quickly locate pharmacodynamic anomalies.

[0012] In a further optimized scheme, after marking the abnormal change rate, the report generation module extracts the chromatographic peak retention time data and the precise mass number of the primary mass ion from the five-dimensional physicochemical feature vector corresponding to the marked experimental time point. Using this as a search index, it queries the Mongolian medicine *Meihuacao* component database for the corresponding candidate chemical component names, and adds these candidate chemical component names as chemical component association tags to the efficacy analysis report. This establishes a direct correlation between abnormal efficacy responses and specific chemical components, providing direct clues for elucidating the efficacy material basis of Mongolian medicine *Meihuacao*.

[0013] As another preferred technical solution of the present invention, after generating the efficacy analysis report, the system further extracts all the rate of change values ​​exceeding the standard efficacy change rate reference range from the report, calculates the absolute value of the difference between each exceeding value and its reference range boundary value, calculates the average deviation amplitude by grouping by cell line type, and combines the distribution density of the experimental time points corresponding to the abnormal change rate to calculate the comprehensive efficacy activity score of the Mongolian herb *Melia azedarach* sample. The comprehensive efficacy activity score is compared with a preset activity grading threshold to determine the activity level of the sample and add it to the end of the efficacy analysis report. Through this comprehensive scoring mechanism, the system quantifies multidimensional efficacy response information into a single activity level index, realizing objective comparison and rapid grading of the efficacy activity of *Melia azedarach* samples from different batches or under different treatment conditions, providing a scientific basis for quality control and clinical application screening.

[0014] The technical effects and advantages provided by the present invention in the above technical solution are as follows: By constructing a five-dimensional physicochemical feature vector and generating a multi-dimensional physicochemical feature space point set, chromatographic peak retention time data, chromatographic peak area data, precise mass number of the precursor ion in primary mass spectrometry, maximum abundance fragment mass-to-charge ratio in secondary mass spectrometry fragment ion abundance distribution data, and absorption intensity values ​​of spectral feature data at preset feature wavenumber positions are combined into a unified feature expression. This collaborative characterization method breaks the limitation that chromatographic, mass spectrometric, and spectral data independently describe samples, enabling physicochemical information from different detection sources to be synchronously organized in the same five-dimensional coordinate system. The sample state at each sampling time point is completely described by a five-dimensional vector, and the complementary information of multimodal data can be jointly utilized in subsequent cluster analysis. The intrinsic correlation between the chromatographic retention behavior, molecular mass information, fragmentation patterns, and functional group spectral absorption characteristics of the sample chemical components is preserved in a unified feature space.

[0015] Steady-state feature vectors are extracted from the cluster label sequences of dynamic clustering results based on physicochemical characteristics. Key experimental time nodes are determined from these steady-state feature vectors. The target pharmacochemical test path is selected based on minimizing the cumulative time axis alignment deviation between the key experimental time nodes and the preset experimental nodes in the candidate pharmacochemical test path. This path selection mechanism drives the setting of pharmacochemical test time nodes by the dynamic clustering evolution of the sample's own physicochemical characteristics over time. Steady-state feature vectors that continuously appear for a duration exceeding a preset duration threshold in the clustering results mark the time segments where the sample's physicochemical state is relatively stable. Key experimental time nodes are extracted from these steady-state segments, avoiding the uncertainty of measurement results caused by setting test points in the non-steady-state stage where the sample's physicochemical characteristics fluctuate drastically. The calculation of the cumulative time deviation ensures the maximum consistency between the target path and the actual physicochemical evolution rhythm of the sample. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0017] Figure 1 This is a schematic diagram of the structure of the Mongolian medicine plum blossom herb efficacy test and analysis system; Figure 2 This is a flowchart of the multi-source physicochemical feature collection and multi-dimensional feature space construction of Mongolian medicine plum blossom grass samples; Figure 3 This is a flowchart of a method for generating pharmacodynamic analysis reports based on multidimensional pharmacodynamic response maps; Figure 4 This is a graph showing the retention time of the chromatographic peaks and the precise mass number of the parent ion in a primary mass spectrometry test for a Mongolian medicinal herb, Prunus mume. Figure 5 This is a schematic diagram of the dynamic cluster label sequence and steady-state interval of the physicochemical characteristics of Mongolian medicinal herb Plum Blossom Grass samples. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, 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.

[0019] See Figure 1This invention provides a system for analyzing the efficacy of Mongolian medicinal herbs, specifically *Meihua* grass. The system includes a feature acquisition module, a spatial mapping module, a cluster analysis module, a path selection module, an efficacy determination module, and a report generation module. The feature acquisition module collects multi-source physicochemical features from *Meihua* grass samples, obtaining chromatographic, mass spectrometric, and spectral feature data. The spatial mapping module performs feature spatial mapping on the chromatographic, mass spectrometric, and spectral feature data, generating a multi-dimensional physicochemical feature spatial point set for the *Meihua* grass samples. The cluster analysis module performs dynamic cluster analysis on the multi-dimensional physicochemical feature spatial point set, obtaining dynamic clustering results for the physicochemical features of the *Meihua* grass samples. Guided by the dynamic clustering results of the physicochemical features, the path selection module performs multi-level efficacy test path selection for the *Meihua* grass samples, determining the target efficacy test path. The efficacy determination module measures the efficacy indicators of the *Meihua* grass samples according to the target efficacy test path, obtaining a multi-dimensional efficacy response spectrum of the *Meihua* grass samples. The report generation module generates a pharmacodynamic analysis report for a Mongolian herb *Meihuacao* sample based on a multidimensional pharmacodynamic response map.

[0020] Example 1: In specific implementation, refer to Figure 2 The multi-source physicochemical characteristic collection process of Mongolian medicinal herb *Prunus mume* samples is as follows: The *Prunus mume* sample refers to *Prunus mume* powder that has been dried, pulverized, and sieved. A methanol-water mixture with a volume ratio of 70:30 was used as the extraction solvent. The *Prunus mume* sample and extraction solvent were mixed at a mass ratio of 1:10. This 1:10 ratio was determined based on pre-solubility tests, which showed that flavonoids and phenolic acids in *Prunus mume* reached over 95% of the saturated extraction yield at this ratio; further increasing the solvent volume would not significantly improve extraction efficiency. After mixing, the mixture was placed in an ultrasonic cleaner and ultrasonically extracted for 30 minutes at a frequency of 40kHz, a power of 300W, and a water temperature of 25℃. After extraction, the mixture was filtered through a 0.22μm microporous membrane to obtain the test solution of the *Prunus mume* sample.

[0021] The test solution was injected into an ultra-high performance liquid chromatograph (UHPLC) using a C18 reversed-phase column. Mobile phase A was 0.1% formic acid aqueous solution, and mobile phase B was acetonitrile. The gradient elution program was set as follows: from 0 to 2 minutes, the volume fraction of mobile phase B linearly increased from 5% to 20%; from 2 to 8 minutes, the volume fraction of mobile phase B linearly increased from 20% to 45%; from 8 to 12 minutes, the volume fraction of mobile phase B linearly increased from 45% to 80%; from 12 to 14 minutes, the volume fraction of mobile phase B remained at 80%; from 14 to 15 minutes, the volume fraction of mobile phase B linearly decreased from 80% to 5%, and equilibrated at 5% for 3 minutes. The flow rate was set to 0.3 mL / min, and the column temperature was maintained at 35℃. During elution, a UV detector was used to acquire chromatographic signals in real time at a wavelength of 280 nm to obtain the retention time and peak area data of the Mongolian medicinal herb *Meihuacao* sample at different elution time points. The retention time data and peak area data of a chromatographic peak collected at each elution time point are combined into a set of two-dimensional data records according to the correspondence. The two-dimensional data records of all elution time points are arranged in chronological order to generate chromatographic feature data.

[0022] Simultaneously, the analyte solution was injected into the ultra-high performance liquid chromatograph (UHPLC) and then introduced into a quadrupole time-of-flight mass spectrometer (QFS) at a split ratio of 1:4 for full-scan acquisition. The QFS employed an electrospray ionization source operating in positive ion mode, with capillary voltage set to 3.5 kV, cone voltage to 40 V, desolvation gas temperature to 350 °C, desolvation gas flow rate to 600 L / h, and a scan mass range of m / z 50 to 1200. During the full-scan acquisition, the precise mass numbers of the primary mass spectrometry precursor ions for each component in the analyte solution were obtained. In data-dependent mode, the three precursor ions with the highest responses were automatically selected for collision-induced dissociation, with the collision energy ramp set to 20 eV to 40 eV, yielding secondary mass spectrometry fragment ion abundance distribution data. The precise mass number of a primary mass spectrometry precursor ion acquired at each sampling point was combined with the corresponding secondary mass spectrometry fragment ion abundance distribution data into a single record. All records from all sampling points were arranged chronologically to generate mass spectrometry characteristic data.

[0023] The test solution was transferred to a quartz cuvette and placed in a near-infrared spectrometer equipped with a transmission accessory. The preset wavenumber range was set to 4000 cm⁻¹ to 10000 cm⁻¹. This wavenumber range covers the combination and overtone absorption regions of hydrogen-containing groups such as C-H, O-H, and N-H bonds, and can reflect the overall information of all hydrogen-containing organic components in the Mongolian medicine *Meihuacao*. The spectrometer scanning parameters were set to: resolution 8 cm⁻¹, 64 scans, and scanning against an air background. The spectral absorption intensity values ​​of the test solution at each wavenumber point within the preset wavenumber range were obtained. These values ​​were arranged in ascending order of wavenumber to form a spectral absorption intensity sequence data, which was used as the spectral characteristic data.

[0024] The process of generating a multidimensional physicochemical feature point set of Mongolian medicine *Meihuacao* sample by feature space mapping of chromatographic feature data, mass spectrometry feature data, and spectral feature data is as follows: From the chromatographic feature data, extract the chromatographic peak retention time data and chromatographic peak area data corresponding to each sampling time point. Use the chromatographic peak retention time data as the first spatial dimension coordinate and the chromatographic peak area data as the second spatial dimension coordinate. The chromatographic peak retention time data reflects the retention behavior of the component in the chromatographic column, and the chromatographic peak area data reflects the component content. From the mass spectrometry feature data, extract the precise mass number of the primary mass spectrometry precursor ion corresponding to each sampling time point as the third spatial dimension coordinate. Select the fragment ion with the highest abundance from the secondary mass spectrometry fragment ion abundance distribution data corresponding to the same sampling time point, and use the mass-to-charge ratio of this fragment as the fourth spatial dimension coordinate. The precise mass number of the primary mass spectrometry precursor ion provides the molecular weight information of the precursor ion, and the mass-to-charge ratio of the most abundant fragment ion reflects the characteristic fragment information of the most easily fragmented type. From the spectral feature data, extract the absorption intensity value of the spectral absorption intensity sequence data at the preset characteristic wavenumber position of 4800 cm⁻¹ as the fifth spatial dimension coordinate. The selection of the preset characteristic wavenumber position 4800 cm⁻¹ is based on the fact that this wavenumber corresponds to the first overtone absorption of the stretching vibration of hydrocarbon groups in aromatic compounds. In the near-infrared spectra of multiple batches of Mongolian medicinal herb Plum Blossom Grass, this position has the largest coefficient of variation in absorption intensity and is most sensitive to changes in physicochemical state.

[0025] By combining the five spatial dimensions mentioned above, a five-dimensional physicochemical feature vector is constructed for the Mongolian medicinal herb *Prunus mume* sample at each sampling time point. In specific implementation, the five-dimensional physicochemical feature vector is expressed in the following form: ; in, The Mongolian medicinal herb *Prunus mume* sample was in the first... Five-dimensional physicochemical feature vectors at each sampling time point; This is the sequence number of the sampling time point, with a value of [value]. , This represents the total number of all sampling time points; Indicates the first Chromatographic peak retention time data at each sampling time point, in minutes; Indicates the first Chromatographic peak area data at each sampling time point, in milliabsorbance units per minute; Indicates the first The precise mass number of the primary mass ion at each sampling time point, in Daltons; Indicates the first The mass-to-charge ratio of the largest abundance fragment in the secondary mass spectrometry fragment ion abundance distribution data at each sampling time point is a dimensionless ratio value. Indicates the first The absorption intensity values ​​of the spectral characteristic data at the preset characteristic wavenumber position of 4800 cm⁻¹ at each sampling time point are expressed in units of absorbance.

[0026] The five-dimensional physicochemical feature vectors of Mongolian medicinal plum blossom grass samples at all sampling time points are arranged according to the sampling time sequence. Arrange them in ascending order to generate a multidimensional physicochemical feature space point set, denoted as... This set of points serves as the input data for the subsequent cluster analysis module.

[0027] See Figure 4 The horizontal axis of the figure represents elution time (in minutes), ranging from 0 to 15 minutes. The left vertical axis represents UV absorption intensity (in milliabsorbance units, mAU), and the right vertical axis represents ion abundance (in counts, on the order of 10^6) corresponding to the exact mass number of the primary ion in the first-order mass spectrometry. In the legend, solid lines represent chromatographic peak retention time data, and dashed lines represent the baseline ion current at the exact mass number of the primary ion in the first-order mass spectrometry.

[0028] The solid line curve in the figure shows the change in UV detector signal intensity of the Mongolian medicinal herb *Meihuacao* sample at a wavelength of 280 nm with elution time, displaying five distinct chromatographic peaks located at approximately 2.0, 4.3, 7.3, 10.3, and 13.0 minutes. The peak heights reach approximately 120, 350, 200, 280, and 150 mAU, respectively, indicating that the absorption intensities of each component at this wavelength differ, reflecting good chromatographic separation and differences in the content of each component.

[0029] The dashed curve corresponds to the base peak ion abundance at the precise mass number of the primary ion in the first-order mass spectrometry, coinciding with the chromatographic peak time point. The dashed curve shows a peak value at the aforementioned chromatographic peak time point, highly overlapping with the retention time of the chromatographic peak, indicating that chromatographic and mass spectrometric detection were achieved simultaneously. The ion abundances of each peak are approximately 1.8 × 10^6, 1.5 × 10^6, 1.0 × 10^6, 1.2 × 10^6, and 0.7 × 10^6, respectively. The peaks are sharp and without significant tailing, reflecting that the stability and sensitivity of the mass spectrometry signal meet the requirements for sample component detection.

[0030] The overall curve exhibits a multi-peak distribution, with the baseline essentially returning to zero between peaks, demonstrating the effective separation of flavonoids and phenolic acids from the Mongolian herb *Meihuacao* sample using the C18 reversed-phase column and gradient elution program. The retention times and precursor ion characteristics of each peak are used in subsequent extraction of the five-dimensional physicochemical feature vector. and It provides an accurate numerical basis.

[0031] This figure effectively demonstrates the synchronicity and correlation between the retention time data of chromatographic peaks acquired by the ultra-high performance liquid chromatograph and the quadrupole time-of-flight mass spectrometer in Example 1 and the precise mass number of the precursor ion peak ion current data of the primary mass spectrometer. This provides reliable physicochemical feature data support for the subsequent construction of multidimensional physicochemical feature spatial point sets and dynamic cluster analysis.

[0032] Example 2: Dynamic clustering analysis was performed on a multidimensional physicochemical feature spatial point set to obtain the dynamic clustering results of the physicochemical features of Mongolian medicinal herb *Prunus mume* samples. The multidimensional physicochemical feature spatial point set is as follows: ,in This represents the total number of all sampling time points. The Mongolian medicinal herb *Prunus mume* sample was in the first... Five-dimensional physicochemical feature vectors at each sampling time point The value is The overall architecture of dynamic clustering analysis comprises four modules: a center initialization module, a distance calculation and candidate selection module, a cluster assignment module, and a center update and convergence judgment module. The output of the center initialization module is connected to the input of the distance calculation and candidate selection module, and the output of the distance calculation and candidate selection module is connected to the input of the center initialization module, forming a loop until a preset condition is met. Then, the final output of the center initialization module is connected to the input of the cluster assignment module, and the output of the cluster assignment module is connected to the input of the center update and convergence judgment module, forming an iterative loop until the convergence condition is met, at which point the dynamic clustering results based on physicochemical characteristics are output.

[0033] In specific implementation, the center initialization module performs the following steps: Initializes the cluster center set of the multidimensional physicochemical feature space point set as an empty set, denoted as... From the multidimensional physicochemical feature space point set A five-dimensional physicochemical eigenvector is randomly selected from the data, denoted as . ,in From 1 to A random integer value will be selected from the data. As the first cluster center ,Will Add to cluster center set In the middle, at this time .

[0034] The distance calculation and candidate selection module performs the following steps: First, it performs Z-score normalization on the point set of the multidimensional physicochemical feature space: Five-dimensional physicochemical eigenvectors Each dimension Calculate its arithmetic mean at all time points. and standard deviation And normalize according to the following formula: ; Calculate the five-dimensional physicochemical eigenvectors of each unselected point in the multidimensional physicochemical eigenspace. With the current set of cluster centers The Euclidean distance between each cluster center. The Euclidean distance is calculated as follows: ; in, Represents the five-dimensional physicochemical eigenvectors With cluster center The Euclidean distance between them; For the set of cluster centers The sequence number of the cluster center; This represents the index of the spatial dimension in the five-dimensional physicochemical eigenvectors. The value ranges from 1 to 5; express The first spatial dimension coordinates (normalized chromatographic peak retention time data); express The second spatial dimension coordinates (normalized chromatographic peak area data); express The third spatial dimension coordinates (normalized first-order mass spectrometry precursor ion precise mass number). express The fourth spatial dimension coordinates (the mass-to-charge ratio corresponding to the maximum abundance fragment after normalization). express The fifth spatial dimension coordinates (normalized preset characteristic wavenumber position 4800cm) -1 (absorption intensity value at the location). , , , , They represent the cluster centers respectively. The five spatial dimension coordinate values ​​(after normalization) (The dimensions are consistent).

[0035] Record each unselected five-dimensional physicochemical eigenvector The corresponding minimum Euclidean distance is denoted as , The calculation method is as follows: That is, from To the current cluster center set The minimum Euclidean distance is taken from all cluster centers. All unselected five-dimensional physicochemical eigenvectors are then sorted according to their respective minimum Euclidean distances. Sort the data from largest to smallest, and select the top-ranked, predetermined number of five-dimensional physicochemical feature vectors as candidate cluster center vectors. The predetermined number is set to 5. The rationale for setting this number to 5 is that, within the range of 100 to 300 points in the multidimensional physicochemical feature space, selecting 2% to 5% of the total points as the candidate size achieves a balance between initialization efficiency and the representativeness of the cluster centers. This was based on statistical analysis of the number of sampling time points from multiple batches of Mongolian medicinal herb *Prunus mume* samples, and the total number of all sampling time points... Typically, the number is between 120 and 180, with 5 being chosen as the candidate number, corresponding to a proportion of approximately 3% to 4%.

[0036] A five-dimensional physicochemical feature vector is randomly selected from the candidate cluster center vectors as a new cluster center, and the new cluster center is added to the cluster center set. The process of calculating distances and selecting candidates, and adding cluster centers in the center initialization module, is repeated until a set of cluster centers is reached. The number of cluster centers reached the preset number of clusters. The preset number of clusters was set to 5. The reason for setting the preset number of clusters to 5 is that the physicochemical characteristics of the Mongolian medicine Plum Blossom Grass sample will go through five typical phases in the time dimension: initial equilibrium period, active ingredient release period, synergistic effect period, degradation and transformation period, and terminal stability period. Setting the number of clusters to 5 can make the final cluster division correspond to these five physicochemical phases, which is convenient for the subsequent conversion of cluster labels into phase labels with physical meaning.

[0037] Cluster center set Once the number of cluster centers reaches 5, the center initialization module will finalize the 5 cluster centers. Output to the cluster allocation module.

[0038] In practical implementation, the cluster allocation module performs the following steps: It allocates the multidimensional physicochemical feature space point set... Each five-dimensional physicochemical eigenvector Assign it to the cluster represented by the cluster center that is closest to it in Euclidean distance. Calculate With 5 cluster centers to Their respective Euclidean distances, using the formula Complete the distance calculation, and Returning to the capital The one that gets the minimum value The corresponding cluster. For all After performing the above allocation operation on the five-dimensional physicochemical feature vectors, the initial cluster partitioning result is obtained. The initial cluster partitioning result consists of 5 clusters, each containing several five-dimensional physicochemical feature vectors. Each five-dimensional physicochemical feature vector is assigned a cluster label, and the value of the cluster label is an integer from 1 to 5.

[0039] The center update and convergence judgment module performs the following steps: Calculate the arithmetic mean of all five-dimensional physicochemical eigenvectors in each cluster. For clusters labeled as... clusters, among which The value ranges from 1 to 5, and the cluster contains A five-dimensional physicochemical eigenvector, denoted as Calculate the arithmetic mean of these vectors: ; in, Indicates cluster label as The cluster centers of the clusters are updated as a new five-dimensional vector; Indicates cluster label as The total number of five-dimensional physicochemical eigenvectors contained in the cluster; , represents the index of the five-dimensional physicochemical eigenvectors within the cluster. The value ranges from 1 to ; Indicates cluster label as The first in the cluster Five physicochemical feature vectors are generated. The arithmetic mean of these features is calculated sequentially for each of the five clusters to obtain five updated cluster centers. to The cluster centers will be updated. Compared with the previous cluster center of this cluster The two vectors are compared, and the offset between them is calculated using the Euclidean distance metric between the two five-dimensional vectors. The calculation method is as follows: ,in Indicates cluster label as The cluster center offset of the cluster. This indicates that the cluster centers are updated at the th... dimensional coordinates, This indicates that the last cluster center was in the [number]th ...]. The coordinates of the dimension. The offsets of the five clusters are summed to obtain the total offset. Compare the total offsets. The value of the preset convergence threshold is related to the value of the preset convergence threshold, which is set to 10. -4 Set the preset convergence threshold to 10. -4 The rationale is as follows: the dimensions of each dimension in the five-dimensional feature space differ significantly; the retention time of chromatographic peaks ranges from 1 to 15 minutes; the peak area ranges from 0 to 5000 mAbsorbance Units per minute; the precise mass number of the primary mass spectrometer precursor ion ranges from 50 to 1200 Daltons; the mass-to-charge ratio corresponding to the most abundant fragments ranges from 50 to 1200; and the absorbance value ranges from 0 to 1 absorbance. When clustering in the unnormalized raw space, the convergence threshold is set to 10. -4 It can make the average fluctuation of the cluster center position in each dimension lower than the lower limit of the effective resolution under the original dimension. After reaching this threshold, iteration will not change the cluster affiliation of each sample point.

[0040] If the total offset Greater than or equal to the preset convergence threshold of 10 -4 Then the cluster centers will be updated. As the new cluster centers for each cluster, the processing steps of the cluster assignment module and the center update and convergence judgment module are repeated for one iteration. If the total offset... Less than the preset convergence threshold 10 -4 The iteration terminates when the last iteration's cluster partitioning result is obtained, and this result is taken as the dynamic clustering result based on physicochemical features. Composed of cluster labels, arranged in chronological order of sampling time, denoted as... ,in The Mongolian medicinal herb *Prunus mume* sample was in the first... The cluster label to which each sampling time point belongs. The value can be an integer from 1 to 5.

[0041] Example 3: In specific implementation, guided by the dynamic clustering results of physicochemical characteristics, a multi-level efficacy test path selection is performed on the Mongolian medicine *Meihuacao* sample to determine the target efficacy test path for the *Meihuacao* sample. The multi-level efficacy test path selection architecture includes four modules: a steady-state identification module, a time alignment module, a deviation calculation module, and a path decision module. The input of the steady-state identification module receives the dynamic clustering results of physicochemical characteristics, and the output of the steady-state identification module is connected to the first input of the time alignment module. The second input of the time alignment module receives candidate efficacy test path data from a preset efficacy test library. The output of the time alignment module is connected to the input of the deviation calculation module, and the output of the deviation calculation module is connected to the input of the path decision module. The path decision module outputs the target efficacy test path.

[0042] The steady-state identification module performs the following steps: parses the dynamic clustering results of physicochemical features, and the dynamic clustering results of physicochemical features are denoted as... ,in This represents the total number of all sampling time points. The Mongolian medicinal herb *Prunus mume* sample was in the first... The cluster label to which each sampling time point belongs. The value can be an integer from 1 to 5. Begin by reading sequentially according to the sampling time. The value is used to count the number of consecutive occurrences of each cluster label in the cluster label sequence. When the sequence... The value from First switch to At that time, Values At the end of a continuous segment, record the cluster label. The duration of consecutive occurrences. The calculation method for the duration of consecutive occurrences is: duration of consecutive occurrences. Number of consecutive occurrences Sampling time interval. The sampling time interval is the time difference between two adjacent sampling time points when acquiring the multidimensional physicochemical feature spatial point set. The sampling time interval is set to 0.1 minutes. The basis for setting the sampling time interval to 0.1 minutes is as follows: the data acquisition frequency of the ultraviolet detector of the ultra-high performance liquid chromatograph is 10 Hz, the full scan acquisition frequency of the quadrupole time-of-flight mass spectrometer is 10 Hz, and the scanning speed of the near-infrared spectrometer is to complete a full wavenumber range scan every 0.1 minutes. The three instruments acquire data synchronously in the time dimension, and the synchronous trigger signal period is set to 0.1 minutes. Therefore, the time interval between adjacent sampling time points is 0.1 minutes.

[0043] The preset duration threshold is set to 2 minutes. The rationale for this setting is that the peak width of flavonoid glycosides in Mongolian medicinal herbs (such as *Meihuacao*) on the chromatographic column is typically between 0.5 and 1.5 minutes. Using the baseline separation requirements between adjacent peaks as a reference, 2 minutes covers the duration of a complete chromatographic peak from its start to its end, effectively distinguishing transient cluster tag jumps caused by chromatographic fluctuations from the true physicochemical stable state. Cluster tags whose consecutive occurrences exceed 2 minutes (i.e., the preset duration threshold) are marked as steady-state feature vectors. For a cluster tag whose consecutive occurrences exceed 2 minutes... To avoid data redundancy and improve path matching accuracy, only the start and end sampling timestamps of a segment are extracted as feature points (if the segment duration exceeds 5 minutes, the midpoint sampling time is also extracted). The sampling timestamps corresponding to the extracted feature points are used as key experimental time nodes. For example, if a steady-state interval is 0 to 4 minutes, then 0 minutes and 4 minutes are extracted as key nodes; if the steady-state interval is 8 to 15 minutes (7 minutes in length), then 8 minutes, 11.5 minutes (midpoint), and 15 minutes are extracted as key nodes.

[0044] The time alignment module performs the following steps: It acquires multiple candidate efficacy test paths from a pre-defined efficacy test library. This library is constructed by collecting historical efficacy test paths of Mongolian herbal medicine *Meihuacao* and annotating the timestamps of each pre-defined test node. In practice, the library is constructed as follows: It collects efficacy test records of *Meihuacao* completed within the past 24 months. Each record includes the name of each operational step during the test and the execution time of that step after the start of the test, in minutes, starting from the moment the *Meihuacao* sample solution is added to the cell culture plate to be tested, which is taken as time zero. The names of operational steps in the same test record are converted into pre-defined test node names. These pre-defined test node names follow standardized naming rules: "adding medicine" is unified as the node name "Medicine Contact Start Point"; "first culture medium change" is unified as the node name "first culture medium change point"; "first detection" is unified as the node name "first efficacy detection point"; "half culture medium changes" is unified as the node name "half culture medium change point"; and "final detection" is unified as the node name "final efficacy detection point". The standardized preset test node name is combined with the execution time of the operation step into a node record. All node records in the same test record are arranged in ascending order of execution time, forming a candidate drug efficacy test path. Each candidate drug efficacy test path contains multiple preset test nodes and execution order. The preset drug efficacy test library contains a total of 120 candidate drug efficacy test paths, with the number of preset test nodes in each path ranging from 3 to 7.

[0045] The time alignment module aligns the key experimental time nodes output by the steady-state identification module with the preset experimental nodes in each candidate pharmacodynamic test path in the preset pharmacodynamic test library. The time alignment method is as follows: the key experimental time node sequence of the Mongolian medicine *Meihuacao* sample is denoted as... ,in This represents the total number of key experimental time points. Indicates the first Key experimental time nodes The value ranges from 1 to For a candidate drug efficacy trial pathway, the timestamp sequence of its pre-set trial nodes is denoted as... ,in This represents the total number of pre-set test nodes in the candidate drug efficacy test pathway. Indicates the first The timestamps of the preset test nodes The value ranges from 1 to The dynamic time warping algorithm is used to align the two time series. The dynamic time warping algorithm constructs a... OK The cumulative distance matrix of the column, find a path from the cumulative distance matrix. arrive The regularized path minimizes the cumulative distance along the regularized path, and the regularized path determines... Each element in The correspondence between one or more elements. Due to the number of key experimental time nodes. The number of test nodes has been compressed to the preset number by feature point extraction (see the steady-state recognition module modification). For (3-7) nodes of similar magnitude, the normalized path of the dynamic time warping algorithm can more accurately reflect the correspondence between nodes, avoiding the influence of... Much larger The resulting "one-to-many" forced matching bias (e.g., original , At times, a single preset node may be forcibly matched with 10 key nodes, masking the true time correlation.

[0046] The deviation calculation module performs the following steps: Based on the correspondence determined by the regularized path output by the time alignment module, it calculates the absolute value of the time deviation between the key experimental time nodes and the corresponding preset experimental nodes in the candidate drug efficacy experimental path. For key experimental time nodes... The timestamp of the preset test node aligned with it in the regularized path is ,in Indicates the relationship between the regularized path and the first... The sequence number of the preset test node corresponding to each key test time node. The calculation method for the absolute value of the time deviation is as follows: .

[0047] The absolute values ​​of the time deviations are summed for each candidate pharmacodynamic testing pathway to obtain the cumulative time deviation for that pathway. The cumulative time deviation is calculated as follows: ; in, Indicates the first in the preset pharmacological efficacy test library Cumulative time deviation of each candidate drug efficacy test pathway, in minutes; This is the sequence number of the candidate drug efficacy trial pathway. The value ranges from 1 to 120; This refers to the total number of key experimental time points; These are the sequence numbers of key experimental time nodes. The value ranges from 1 to ; Indicates the first Key experimental time nodes; Indicates the first The regularized pathway of the candidate drug efficacy trial pathway is related to the first... The sequence number of the preset test nodes aligned with the key test time nodes; Indicates the first The candidate pharmacodynamic trial pathway numbered as follows: The timestamp of the preset test node.

[0048] The path decision module performs the following steps: Calculate the cumulative time deviation for all candidate pharmacodynamic test paths in the preset pharmacodynamic test library. Then, compare 120. The size of the value, select The candidate pharmacodynamic testing pathway with the smallest value is selected as the target pharmacodynamic testing pathway. If two or more candidate pharmacodynamic testing pathways exist... If the values ​​are equal and both are the minimum, then the preset number of test nodes will be selected. The one with fewer pathways is selected as the target efficacy trial pathway. The rationale for fewer paths is that, given the same cumulative time deviation, fewer preset test nodes mean a more streamlined experimental procedure, reducing human error introduced by frequent operations during actual efficacy determination. The target efficacy test path includes a preset sequence of test nodes, the execution order of each node, and its corresponding timestamp. The target efficacy test path is then sent to the efficacy determination module.

[0049] See Figure 5In the figure, the vertical axis represents the cluster label value range of 1 to 5, and the horizontal axis represents the sampling time point in minutes. The curve shows the dynamic clustering result of the Mongolian medicinal herb *Prunus mume* at each sampling time point in a stepped manner, displaying the change sequence of cluster labels in the clustering results. The curve shows several relatively stable step regions over time, indicating the dynamic changes in the physicochemical characteristics of the sample in different time periods. The steady-state interval is marked by light gray blocks in the figure, which is the time period when the cluster label appears continuously for more than 2 minutes, exceeding the preset threshold. The steady-state intervals are distributed in segments of approximately 0 to 4 minutes, 8 to 15 minutes, and 19 to 30 minutes, reflecting that the physicochemical characteristics of the sample remain relatively stable during these time periods.

[0050] The multiple vertical black dashed lines in the figure represent key experimental time points, distributed at approximately 1.5, 4.5, 8.5, 15.5, 22.5, and 27.5 minutes. These points mark the boundaries of the steady-state intervals of the physicochemical state and are important time points for selecting the efficacy test path. Sampling started at minute 0. The cluster tag sequence began in cluster 1, transitioning from cluster 1 to cluster 2 within approximately 4 minutes and briefly returning to cluster 1, subsequently entering the first steady-state interval. Afterward, the cluster tag rapidly jumped to cluster 3 after 4 minutes and maintained this state for 8 to 15 minutes, forming the second steady-state interval. After 15 minutes, the cluster tag gradually transitioned to cluster 4, maintaining this state for approximately 22 minutes. Finally, the cluster tag rose to cluster 5 and remained there until the end of sampling for 30 minutes, forming the third steady-state interval.

[0051] The dynamic changes in the cluster tag sequence conform to the analytical method of the steady-state identification module in Example 3 for the dynamic clustering results of physicochemical features. By statistically analyzing the duration of continuous occurrence of cluster tags, the steady-state feature vector is identified, and then key experimental time nodes are extracted. The distribution of steady-state intervals and key nodes in the figure corresponds to five typical phases of the physicochemical state of the Mongolian medicine *Meihuacao* sample: the initial equilibrium period, the active ingredient release period, the synergistic effect period, the degradation and transformation period, and the final stable period. This provides basic data support for the subsequent time alignment module to perform dynamic time warping based on key experimental time nodes and candidate paths in the preset pharmacodynamic test library, ensuring the scientificity and accuracy of the pharmacodynamic test path selection.

[0052] Example 4: In specific implementation, refer to Figure 3 The efficacy indicators of *Melia azedarach* samples were measured according to the target efficacy test path, and a multidimensional efficacy response spectrum of the *Melia azedarach* samples was obtained. The target efficacy test path was determined by the path decision module and sent to the efficacy measurement module. The target efficacy test path contains a preset sequence of test nodes, denoted as... ,in The total number of pre-set test nodes in the target efficacy test pathway, and each pre-set test node Each comes with a corresponding execution time. ,in The value ranges from 1 to , The time frame is calculated from the moment the Mongolian herb *Meihuacao* sample solution first comes into contact with the cell line to be tested, in minutes. In the efficacy assay module, the execution order of each preset test node in the target efficacy test pathway is followed sequentially at each preset test node. Corresponding test time points An equal volume of sample solution was drawn from the Mongolian medicinal herb *Prunus mume* sample. The *Prunus mume* sample was obtained from the same batch of *Prunus mume* powder used in the feature acquisition module, and the test solution was prepared using the same extraction method. The test solution was stored at 4°C in a light-protected environment during the efficacy determination.

[0053] At each test time point Using a pipette, sample solutions were drawn from the Mongolian medicinal herb *Meihuacao* sample and added to different wells of a pre-cultured cell line plate at each experimental time point. The cell line plates were 96-well plates containing human hepatocellular carcinoma cells, human breast cancer cells, and normal human liver cells. The HepG2 cell line was used for human hepatocellular carcinoma, the MCF-7 cell line for human breast cancer, and the L-02 cell line for normal human liver cells. The three cell lines were arranged in rows on the 96-well plate. Rows A to D were reserved for human hepatocellular carcinoma cells, rows E to G for human breast cancer cells, and row H for normal human liver cells. Twenty-four hours before the start of the experiment, the three cell lines to be tested were seeded into the corresponding wells at a density of 5,000 cells per well. 200 μL of RPMI-1640 medium containing 10% fetal bovine serum was added to each well. The wells were then incubated in a carbon dioxide incubator at 37°C and 5% CO2 for 24 hours to allow the cells to adhere and enter the logarithmic growth phase.

[0054] At each test time point When adding the sample solution, three replicates were selected from each of the wells of the human hepatocellular carcinoma cell line, human breast cancer cell line, and human normal hepatocyte line. Sample solution was added to each selected well. The volume of sample solution added to each well of the cell culture plate was 20 μL. The reason for setting the sample solution volume to 20 μL is that the total volume of the culture medium in the 96-well plate is 200 μL. Adding 20 μL of sample solution dilutes the sample solution 11 times. At this dilution, the final concentrations of total flavonoids and total phenolic acids in the Mongolian herbal extract of *Melanchoe pinnatifida* fall within the range of 50 μg / mL to 200 μg / mL. This concentration range is the concentration range determined in the preliminary experiment that can produce a measurable inhibitory effect on cell activity without causing immediate cell necrosis. Furthermore, the surface disturbance caused by the 20 μL addition will not cause cells to detach from the wells.

[0055] At each test time point After adding the sample solution, the cell culture plate was returned to the CO2 incubator and cultured at 37°C with 5% CO2. The starting point for the preset drug efficacy response time was the moment when the sample solution was added and the plate was returned to the incubator. The preset drug efficacy response time was set to 48 hours. The basis for setting the preset drug efficacy response time to 48 hours is that the doubling time of HepG2 human liver cancer cells and MCF-7 human breast cancer cells is approximately 22 hours and 28 hours, respectively. 48 hours covers 1.8 and 1.7 doubling cycles for these two cell lines, respectively, which can fully demonstrate the inhibitory effect of the active ingredients of Mongolian herbal medicine Prunella vulgaris on cell proliferation after the cells have completed a complete division cycle. At the same time, the survival rate of L-02 normal human liver cells remained above 85% at the end of 48 hours, which can effectively distinguish the selective killing effect of the drug on cancer cells and its toxic effect on normal cells.

[0056] After culture, remove the cell line culture plate from the CO2 incubator and add cell viability assay reagent to each well. The cell viability assay reagent used is MTT reagent, prepared at a concentration of 5 mg / mL using phosphate buffer. Add 20 μL of MTT reagent to each well and return the culture plate to the CO2 incubator for another 4 hours. After incubation, gently aspirate the supernatant from the wells and add 150 μL of dimethyl sulfoxide to each well. Place the plate on a plate shaker and shake in the dark for 10 minutes to fully dissolve the formazan crystals. Use a microplate reader to read the absorbance value of each well at the preset detection wavelength. The preset detection wavelength is set to 570 nm, and the reference wavelength is set to 630 nm. The difference between the absorbance value at 570 nm and the absorbance value at 630 nm is taken as the final absorbance value for that well.

[0057] Record each experimental time point The absorbance values ​​for each cell line to be tested are listed below. For human hepatocellular carcinoma cell lines, the arithmetic mean of the absorbance values ​​of three replicates in rows A to D is taken as the absorbance value of the human hepatocellular carcinoma cell line at that experimental time point, denoted as . For human breast cancer cell lines, the arithmetic mean of the absorbance values ​​of three replicates in rows E to G was taken as the absorbance value of the human breast cancer cell line at that experimental time point, denoted as . For the human normal hepatocyte cell line, the arithmetic mean of the absorbance values ​​of the three replicates in row H was taken as the absorbance value of the human normal hepatocyte cell line at that experimental time point, denoted as . The absorbance values ​​of the three cell lines tested at all experimental time points were arranged according to the experimental time sequence and cell line type to construct a three-dimensional data matrix of time-cell line-absorbance. This three-dimensional data matrix was used as a multidimensional pharmacodynamic response map. The dimensions of the multidimensional pharmacodynamic response map are as follows: the first dimension is the experimental time point, arranged according to... The first dimension is arranged in the order of cell line type; the second dimension is arranged in the order of human liver cancer cell line, human breast cancer cell line, and human normal liver cell line; the third dimension is the absorbance value.

[0058] A pharmacodynamic analysis report for Mongolian herb *Meihuacao* was generated based on a multidimensional pharmacodynamic response map. The absorbance values ​​of human liver cancer cell lines at each experimental time point were extracted from the multidimensional pharmacodynamic response map. Since all samples were cultured for 48 hours after drug administration, the measured absorbance values ​​reflect the terminal effect of cells after the same culture cycle at different drug administration times. Therefore, the relative difference rate of absorbance between different experimental groups was calculated, rather than the dynamic change rate. The relative difference rate was calculated as follows: ; in, Indicates the first The first experimental time point and the first The relative difference rate of absorbance of human liver cancer cell lines between the drug-treated groups at each experimental time point, in percentage form; The index represents the difference rate, with values ​​ranging from 1 to... ; Indicates the first The absorbance values ​​of the drug-treated groups at each experimental time point after 48 hours of incubation; Indicates the first The absorbance values ​​of the drug-treated groups at each experimental time point after 48 hours of incubation; This represents the mean absorbance of the control group (no drug administration) after 48 hours of culture. This formula is used to characterize the relative impact of different drug administration times on cell viability, avoiding misjudgment of dynamic change rates due to fixed culture duration.

[0059] Using the same calculation method, absorbance values ​​of human breast cancer cell lines at each experimental time point were extracted from the multidimensional drug response spectrum. The rate of change of absorbance values ​​between adjacent experimental time points was calculated to obtain the absorbance change rate sequence of human breast cancer cell lines, denoted as . The absorbance values ​​of human normal liver cell lines at each experimental time point were extracted from the multidimensional drug response spectrum. The rate of change of absorbance values ​​between adjacent experimental time points was calculated to obtain the absorbance change rate sequence of the human normal hepatocyte line, denoted as . The absorbance change rate sequence of human liver cancer cell lines. human breast cancer cell line absorbance change rate sequence Absorbance change rate sequence of normal human hepatocyte cell lines Align the data according to the experimental time points to generate a comparative rate of change table. The comparative rate of change table is one... The data table has 4 rows and 4 columns. The first column represents the experimental time interval, which is shown as follows: to The second column shows the rate of change in absorbance of human liver cancer cell lines. The third column shows the rate of change in absorbance of human breast cancer cell lines. The fourth column shows the rate of change in absorbance of normal human hepatocyte cell lines. .

[0060] Obtain the reference interval for the standard efficacy rate of change corresponding to the target efficacy test pathway. The reference interval for the standard efficacy rate of change is derived from historical measurement data of the same batch of Mongolian medicinal herb *Melia toosendan* standard under the same test conditions. In specific implementation, the same batch of Mongolian medicinal herb *Melia toosendan* standard refers to a representative control selected from the same batch of Mongolian medicinal herb collected, dried, and pulverized as the Mongolian medicinal herb sample. The standard has been calibrated, and its coefficient of variation for total flavonoid content and total phenolic acid content does not exceed 5%. Under the same test conditions, that is, using the same test node sequence and time arrangement as the target efficacy test pathway, the same cell line type, the same cell seeding density, the same culture medium composition, the same culture conditions, and the same cell viability detection method, three independent replicate efficacy tests are conducted on the standard. The closed interval formed by the minimum and maximum values ​​of the absorbance change rate for each cell line type in each test time interval of the three independent replicate tests is used as the reference interval for the standard efficacy rate of change. For human liver cancer cell lines, the first... The reference interval for the standard efficacy rate of change over the experimental time interval is denoted as . For human breast cancer cell lines, the first The reference interval for the standard efficacy rate of change over the experimental time interval is denoted as . For normal human liver cell lines, the first The reference interval for the standard efficacy rate of change over the experimental time interval is denoted as . .

[0061] Each rate of change in the comparison rate table was compared with the reference range for standard drug efficacy rate of change, and the experimental time points and cell line types corresponding to rate of change values ​​exceeding the reference range were marked. For the rate of change in human hepatocellular carcinoma cell lines... ,like or Then the first Each test time interval The cell line type "human liver cancer cell line" was marked as abnormal. The rate of change for human breast cancer cell lines... ,like or Then the first The experimental time interval and cell line type "human breast cancer cell line" were marked as abnormal. The rate of change for human normal liver cell lines... ,like or Then the first The experimental time interval and cell line type "human normal hepatocyte line" were marked as abnormal.

[0062] After anomaly labeling, the five-dimensional physicochemical feature vectors corresponding to the labeled experimental time points are extracted. The labeled experimental time point refers to the start time of the experimental time interval marked as an anomaly. and termination time Retrieve data from the multidimensional physicochemical feature space point set stored in the system. and The five-dimensional physicochemical feature vectors corresponding to time are established as follows: the sampling time points in the multi-dimensional physicochemical feature space point set are aligned with the test time points in the efficacy test based on a common zero moment. The zero moment is defined as the moment when the test solution is completed. After the zero moment, a five-dimensional physicochemical feature vector is collected every 0.1 minutes. When the test time point... When the absolute value of the difference between the numerical value and the acquisition time of the five-dimensional physicochemical feature vector is less than 0.05 minutes, it is considered that the five-dimensional physicochemical feature vector is related to the experimental time point. Corresponding. From the corresponding experimental time point Five-dimensional physicochemical eigenvectors Retention time data of chromatographic peaks separated from and precise mass number of precursor ions in primary mass spectrometry .

[0063] Based on chromatographic peak retention time data and precise mass number of precursor ions in primary mass spectrometry To retrieve the index, a search was conducted in the Mongolian medicinal herb *Melia azedarach* component database. The database was constructed as follows: Reported chemical component information of *Melia azedarach* was collected from Mongolian medicinal herb literature, phytochemistry databases, and measured data of standards. Each component record includes the component name, chromatographic peak retention time range, measured precise mass number of the parent ion from primary mass spectrometry, and a list of characteristic peaks of fragment ions from secondary mass spectrometry. During the search, the chromatographic peak retention time data was retrieved. Match the chromatographic peak retention time range with the records of each component in the database, if The chromatographic peak falls within the retention time range of a certain component's recorded peak, and the precise mass number of the primary mass spectrometer precursor ion is... If the absolute value of the difference between the component name and the measured precise mass number of the precursor ion recorded by mass spectrometry is less than 0.005 Daltons, the component name recorded by that component is considered a candidate chemical component name. If multiple component records are matched simultaneously, all component names are listed as candidate chemical component names. If no component record is matched, the candidate chemical component name is marked as "unknown component".

[0064] Candidate chemical component names are added next to the marking results in the pharmacodynamic analysis report to form chemical component association tags. The format of the chemical component association tag is as follows: an additional field is appended after the corresponding anomaly tag record, and the field content is the candidate chemical component name. If there are multiple candidate chemical component names, they are separated by semicolons.

[0065] The comparison tables of change rates and the labeling results are structured according to the experimental time sequence to generate a pharmacodynamic analysis report. A blank spreadsheet document is created using spreadsheet software to create a workbook object. In the first row of the first worksheet of the blank spreadsheet document, columns for experimental time points, human liver cancer cell line change rates, human breast cancer cell line change rates, human normal liver cell line change rates, and abnormal labeling are set sequentially. The results are then arranged in ascending order of experimental time points. From 1 to Each experimental time interval " - The text "" is filled in the first... The experimental time points are listed in the column. The absorbance change rate of the human liver cancer cell line corresponding to each experimental time interval is displayed. Enter the first line of the column for the change rate of human liver cancer cell lines. In the rows, the numerical values ​​are rounded to four decimal places. The numerical values ​​representing the rate of change in absorbance of human breast cancer cell lines are then presented. Enter the first item in the column for the rate of change of human breast cancer cell lines. In progress. The numerical value of the change rate of absorbance of normal human liver cell lines. Enter the first item in the column for the rate of change of human normal hepatocyte lines. In progress.

[0066] Determine each test time interval Does a labeled cell line type exist? If a labeled cell line type exists, then in the anomaly marker column... The row contains the name of the labeled cell line type and its corresponding out-of-direction identifier. The out-of-direction identifier is determined as follows: when... Greater than When the direction exceeds the specified value, it is indicated by "↑"; when... Less than When the direction exceeds the specified threshold, it is indicated by "↓". If there are chemical component-related markers within the experimental time interval, the chemical component name is appended after the content entered in the anomaly marker column. The single-line format of the anomaly marker column is "Human liver cancer cell line ↓, component: kaempferol-3-O-rutin", or "Human breast cancer cell line ↑, component: quercetin". If multiple labeled cell line types exist within the same experimental time interval, the anomaly marker content of different cell lines is separated by semicolons.

[0067] After completing the data entry for all experimental time points, save the blank table document as a report file in a preset format. The preset format is selected as OfficeOpenXML, and the file extension is .xlsx. This report file will serve as the pharmacodynamic analysis report. The report file will be provided to the operator for viewing or storage through the output interface of the report generation module.

[0068] Example 5: In specific implementation, after the pharmacodynamic analysis report is generated, all rate changes exceeding the standard pharmacodynamic change rate reference range are extracted from the report. The pharmacodynamic analysis report is a spreadsheet file containing columns for experimental time points, human liver cancer cell line change rates, human breast cancer cell line change rates, human normal liver cell line change rates, and anomaly markers. The spreadsheet file is parsed, and the contents of the anomaly marker column are read row by row. When a cell in the anomaly marker column is not empty, it indicates that within the experimental time interval represented by the corresponding row, one or more cell lines under test have absorbance change rates exceeding the standard pharmacodynamic change rate reference range. For each experimental time interval with anomalies, the values ​​in the human liver cancer cell line change rate, human breast cancer cell line change rate, and human normal liver cell line change rate columns in that row are extracted, and the cell line type name contained in the anomaly marker column is used to determine which rate change value is considered to be exceeding the range. The absorbance change rate value of the human liver cancer cell line marked as exceeding the range is denoted as... ,in The index is for values ​​that exceed the given value. The value is , The total number of human liver cancer cell lines marked as exceeding the rate of change; the number of human breast cancer cell lines marked as exceeding the rate of change is denoted as... ,in The value is The percentage change of a normal human liver cell line exceeding the threshold value is denoted as... ,in The value is .

[0069] Calculate the absolute value of the difference between each excess rate of change and the boundary value of the reference interval for the standard efficacy rate of change corresponding to that rate of change. For human liver cancer cell lines, if Greater than the upper limit of the reference range for the standard efficacy change rate corresponding to the experimental time interval. The absolute value of the difference is calculated as follows: ;like Less than the lower limit of the reference range for standard drug efficacy change rate The absolute value of the difference is calculated as follows: The absolute values ​​of the differences between human breast cancer cell lines and human normal liver cell lines were calculated in the same way, using the upper limit of the reference interval for standard drug efficacy change rate. , and lower limit , For reference.

[0070] The absolute values ​​of the differences were grouped according to their corresponding cell line types. All absolute values ​​of differences associated with human hepatocellular carcinoma cell lines were grouped into the first group, all absolute values ​​of differences associated with human breast cancer cell lines were grouped into the second group, and all absolute values ​​of differences associated with human normal hepatocyte cell lines were grouped into the third group. The arithmetic mean of the absolute values ​​of differences in each group was calculated to obtain the average deviation for each cell line type. The average deviation for human hepatocellular carcinoma cell lines is denoted as... The calculation method is as follows: ,in Indicates the first human liver cancer cell line The absolute value of the difference corresponding to the rate of change. The average deviation of human breast cancer cell lines is denoted as... The calculation method is as follows: The average deviation of a normal human hepatocyte cell line is denoted as... The calculation method is as follows: If no cell line type exceeds the rate of change, the average deviation of that cell line type is zero.

[0071] The comprehensive efficacy score of the Mongolian herb *Meihuacao* was calculated based on the distribution density of experimental time points corresponding to the average deviation and the rate of change exceeding the specified value. The distribution density of experimental time points corresponding to the rate of change exceeding the specified value was defined as: the ratio of the number of experimental time intervals in all Q-1 experimental time intervals where at least one cell line type was marked as exceeding the specified value to the total number of intervals, denoted as ρ, calculated as ρ = U / (Q-1), where U is the number of experimental time intervals where at least one cell line type was marked as exceeding the specified value. The comprehensive efficacy score was calculated as follows: ; in, The overall efficacy score of the Mongolian medicinal herb *Meihuacao* is a dimensionless value. It represents the average deviation of human liver cancer cell lines, and is a dimensionless ratio (the absolute value of the difference between the rate of change and the boundary of the standard interval, since the rate of change itself is a dimensionless quantity in the form of a percentage). This represents the average deviation of human breast cancer cell lines and is a dimensionless ratio. This represents the average deviation of a normal human liver cell line, and is a dimensionless ratio. The arithmetic mean of the average deviation of the three cell line types is given by the dimensionless ratio. The distribution density of the test time points exceeding the numerical value of the rate of change, with a value range between 0 and 1, and is dimensionless; The deviation magnitude weighting coefficient is set to 0.65 and is dimensionless. This is the distribution density weighting coefficient, set to a value of 0.35, which is dimensionless.

[0072] Deviation Amplitude Weighting Coefficient The value of 0.65 is set as follows: the deviation reflects the intensity of the difference between the metabolic response of the tested cell line caused by the Mongolian medicine *Meihuacao* sample and the standard. This intensity of difference directly corresponds to changes in the content or composition ratio of the active ingredient, and is a core indicator for measuring the degree of abnormality in efficacy. Therefore, it should be given a high weight in the activity score. Distribution density weighting coefficient. The value of 0.35 is set because distribution density reflects the persistence of abnormal drug activity over time. A high distribution density indicates that abnormal responses occur continuously across multiple experimental time intervals, and its importance is lower than the intensity of the abnormal response itself. Linear discriminant analysis was performed on multiple batches of Mongolian medicinal herb *Meihuacao* samples with known activity, aiming to maximize the rank correlation coefficient between activity level and score. and The time-rank correlation coefficient reaches its maximum value of 0.91. The two weighting coefficients satisfy... Normalization constraints.

[0073] Comprehensive pharmacological activity score The activity level of the Mongolian medicinal herb *Melia toosendan* sample was determined by comparing it with preset activity grading thresholds. The preset activity grading thresholds include a first grading threshold and a second grading threshold; the first grading threshold was set to 0.3, and the second grading threshold was set to 0.7. The basis for setting the first grading threshold to 0.3 was: performing 12 repeated measurements on the *Melia toosendan* standard sample and calculating the comprehensive efficacy activity score of the standard sample. The results obtained using the above calculation method all fall within the range of 0.02 to 0.15. Taking the upper limit of this range doubled and accounting for measurement error, 0.3 is used as the threshold value for distinguishing between low and medium activity. The basis for setting the second grading threshold to 0.7 is: Several batches of Mongolian medicinal herb *Meihuacao* samples with artificially added known active ingredients were tested. When the active ingredient content reached twice the content of the standard, the comprehensive efficacy activity score was... The concentration is between 0.65 and 0.75, with 0.7 taken as the boundary between medium and high activity. If... If so, the activity level is determined to be "low activity"; if If so, the activity level is determined to be "medium activity"; if If so, the activity level is determined to be "high activity".

[0074] Add the activity rating to the end of the pharmacodynamics analysis report. In the blank row following the last data row of the generated pharmacodynamics analysis report spreadsheet, enter the text "Activity Rating Assessment" in the cell for the "Experiment Time Point" column, and enter the comprehensive pharmacodynamics activity score in the cell for the "Human Hepatocellular Carcinoma Cell Change Rate" column. Enter the numerical value, rounded to three decimal places, and fill in the name of the activity level in the cell of the "Human Breast Cancer Cell Line Change Rate" column. After adding the value, save the efficacy analysis report to complete the complete process of the efficacy test analysis of Mongolian medicine Meihua herb.

[0075] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A system for testing and analyzing the efficacy of Mongolian medicine plum blossom herbs, characterized in that, The system includes: The feature acquisition module performs multi-source physicochemical feature acquisition on the Mongolian medicine plum blossom grass sample, and obtains the chromatographic feature data, mass spectrometry feature data and spectral feature data of the Mongolian medicine plum blossom grass sample; The spatial mapping module performs feature spatial mapping on the chromatographic feature data, the mass spectrometry feature data, and the spectral feature data to generate a multidimensional physicochemical feature spatial point set of the Mongolian medicine plum blossom grass sample; The clustering analysis module performs dynamic clustering analysis on the multidimensional physicochemical feature spatial point set to obtain the dynamic clustering results of the physicochemical features of the Mongolian medicine plum blossom grass sample; The path selection module, guided by the dynamic clustering results of the physicochemical characteristics, performs multi-level efficacy test path selection on the Mongolian medicine plum blossom grass sample to determine the target efficacy test path of the Mongolian medicine plum blossom grass sample. The efficacy determination module measures the efficacy indicators of the Mongolian medicine *Meihuacao* sample according to the target efficacy test path, and obtains a multidimensional efficacy response spectrum of the Mongolian medicine *Meihuacao* sample. The report generation module generates a pharmacodynamic analysis report for the Mongolian herb *Meihuacao* sample based on the multidimensional pharmacodynamic response spectrum.

2. The Mongolian medicine plum blossom herb efficacy testing and analysis system according to claim 1, characterized in that, Multi-source physicochemical characteristic sampling was performed on the Mongolian medicinal herb *Prunus mume*, obtaining chromatographic, mass spectrometric, and spectral characteristic data of the sample. Specifically: The Mongolian medicinal herb *Prunus mume* sample was mixed with an extraction solvent according to a preset mass ratio, and then subjected to ultrasonic-assisted extraction to obtain the test solution of the Mongolian medicinal herb *Prunus mume* sample. The test solution is injected into an ultra-high performance liquid chromatograph for gradient elution separation. The chromatographic peak retention time data and chromatographic peak area data of the test solution at different elution time points are collected in real time. The chromatographic peak retention time data and chromatographic peak area data are combined to generate the chromatographic feature data. The test solution is synchronously introduced into a quadrupole time-of-flight mass spectrometer for full-scan acquisition to obtain the precise mass number of the primary mass spectrometer precursor ion and the abundance distribution data of the secondary mass spectrometer fragment ions. The mass spectrometry feature data is generated by combining the precise mass number of the primary mass spectrometer precursor ion and the abundance distribution data of the secondary mass spectrometer fragment ions. The test solution is transferred to a near-infrared spectrometer for transmission spectral scanning to obtain spectral absorption intensity sequence data of the test solution within a preset wavenumber range, and the spectral absorption intensity sequence data is used as the spectral feature data.

3. The Mongolian medicine plum blossom herb efficacy testing and analysis system according to claim 2, characterized in that, The chromatographic feature data, the mass spectrometry feature data, and the spectral feature data are mapped in a feature space to generate a multidimensional physicochemical feature space point set of the Mongolian medicine *Prunus mume* sample, specifically as follows: Extract the chromatographic peak retention time data and chromatographic peak area data from the chromatographic feature data, and use the chromatographic peak retention time data as the first spatial dimension coordinate and the chromatographic peak area data as the second spatial dimension coordinate; The precise mass number of the primary mass ion in the mass spectrometry feature data is extracted as the third spatial dimension coordinate, and the mass-to-charge ratio of the maximum abundance fragment in the second mass spectrometry fragment ion abundance distribution data is extracted as the fourth spatial dimension coordinate. Extract the absorption intensity value of the spectral absorption intensity sequence data at a preset feature wavenumber position from the spectral feature data and use it as the fifth spatial dimension coordinate; The first spatial dimension coordinate, the second spatial dimension coordinate, the third spatial dimension coordinate, the fourth spatial dimension coordinate, and the fifth spatial dimension coordinate are combined to construct a five-dimensional physicochemical feature vector of the Mongolian medicine plum blossom grass sample at each sampling time point; The five-dimensional physicochemical feature vectors of the Mongolian medicinal herb Plum Blossom Grass sample at all sampling time points are arranged in the order of sampling time to generate the multi-dimensional physicochemical feature space point set.

4. The Mongolian medicine plum blossom herb efficacy testing and analysis system according to claim 3, characterized in that, Dynamic clustering analysis was performed on the multidimensional physicochemical feature spatial point set to obtain the dynamic clustering results of the physicochemical features of the Mongolian medicinal herb *Prunus mume* sample, specifically as follows: The cluster center set of the multidimensional physicochemical feature space point set is initialized to an empty set, and a five-dimensional physicochemical feature vector is randomly selected from the multidimensional physicochemical feature space point set as the first cluster center; Calculate the Euclidean distance between each unselected five-dimensional physicochemical feature vector in the multidimensional physicochemical feature space point set and each cluster center in the current cluster center set, and record the minimum Euclidean distance corresponding to each unselected five-dimensional physicochemical feature vector; Sort the unselected five-dimensional physicochemical feature vectors in descending order of their minimum Euclidean distance, and select a predetermined number of the top-ranked five-dimensional physicochemical feature vectors as candidate cluster center vectors. A five-dimensional physical and chemical feature vector is randomly selected from the candidate cluster center vectors and added to the cluster center set as a new cluster center; Repeat the cluster center addition step until the number of cluster centers in the cluster center set reaches the preset number of clusters; Each five-dimensional physicochemical feature vector in the multidimensional physicochemical feature space point set is assigned to the cluster represented by the cluster center that is closest to it in Euclidean distance, thus obtaining the initial cluster partitioning result; Calculate the arithmetic mean of all five-dimensional physicochemical eigenvectors in each cluster, and use the arithmetic mean as the updated cluster center of that cluster; The cluster allocation step and the cluster center update step are executed iteratively until the sum of the offsets between the updated cluster centers and the previous cluster centers is less than a preset convergence threshold. The cluster partitioning result obtained in the last iteration is used as the dynamic clustering result of the physicochemical features.

5. The Mongolian medicine plum blossom herb efficacy testing and analysis system according to claim 4, characterized in that, Guided by the dynamic clustering results of the aforementioned physicochemical characteristics, a multi-level efficacy test path selection was performed on the Mongolian medicine *Prunus mume* sample to determine the target efficacy test path for the *Prunus mume* sample, specifically as follows: The dynamic clustering results of the physicochemical characteristics are analyzed to obtain the cluster label sequence of the Mongolian medicine plum blossom grass sample at all sampling time points; The duration of continuous occurrence of each cluster label in the cluster label sequence is counted, and the five-dimensional physicochemical feature vectors corresponding to cluster labels whose continuous occurrence duration exceeds a preset duration threshold are marked as steady-state feature vectors. Extract the corresponding sampling time points from the steady-state feature vector as key experimental time nodes; Obtain multiple candidate pharmacodynamic test paths from the preset pharmacodynamic test library. Each candidate pharmacodynamic test path contains multiple preset test nodes and execution order. Align the key test time nodes with the preset test nodes in each candidate drug efficacy test path on the time axis, and calculate the absolute value of the time deviation between the key test time nodes and the preset test nodes at the corresponding positions in the candidate drug efficacy test path. The absolute values ​​of the time deviations are summed according to the candidate drug efficacy test paths to obtain the cumulative time deviation for each candidate drug efficacy test path; The candidate pharmacodynamic test path with the smallest cumulative time deviation is selected as the target pharmacodynamic test path.

6. The Mongolian medicine plum blossom herb efficacy testing and analysis system according to claim 5, characterized in that, The preset efficacy test library is constructed by collecting historical Mongolian medicine Meihua herbal efficacy test paths and marking the timestamps of each preset test node.

7. The Mongolian medicine plum blossom herb efficacy testing and analysis system according to claim 5, characterized in that, The efficacy indicators of the Mongolian medicinal herb *Prunus mume* were determined according to the target efficacy test pathway, and a multidimensional efficacy response spectrum of the *Prunus mume* sample was obtained. Specifically: According to the execution order of each preset test node in the target efficacy test path, an equal volume of sample solution is taken from the Mongolian medicine plum blossom grass sample at the test time point corresponding to each preset test node. The sample solution taken at each experimental time point was added to different wells of a pre-cultured cell line plate, including human liver cancer cell line, human breast cancer cell line and human normal liver cell line. After adding the sample solution at each experimental time point, the cell line culture plate to be tested was placed in a carbon dioxide incubator and cultured for the preset drug efficacy reaction time. After the culture is completed, add cell viability detection reagent to each well and use an ELISA reader to read the absorbance value of each well at the preset detection wavelength. Record the absorbance value of each cell line under test at each experimental time point, arrange the absorbance values ​​of all experimental time points according to the experimental time sequence and cell line type, and construct a three-dimensional data matrix of time-cell line-absorbance. Use the three-dimensional data matrix as the multidimensional drug efficacy response map.

8. The Mongolian medicine plum blossom herb efficacy testing and analysis system according to claim 7, characterized in that, The volume of sample solution added to each well of the cell line culture plate is 20 μL, and the preset drug effect reaction time is 48 hours.

9. The Mongolian medicine plum blossom herb efficacy testing and analysis system according to claim 7, characterized in that, Based on the multidimensional pharmacodynamic response map, a pharmacodynamic analysis report for the Mongolian herb *Meihuacao* sample was generated, specifically as follows: The absorbance values ​​of the human liver cancer cell line at each experimental time point were extracted from the multidimensional drug efficacy response spectrum, and the rate of change of absorbance values ​​between adjacent experimental time points was calculated to obtain the absorbance change rate sequence of the human liver cancer cell line. The absorbance values ​​of the human breast cancer cell line at each experimental time point were extracted from the multidimensional pharmacodynamic response spectrum, and the rate of change of absorbance values ​​between adjacent experimental time points was calculated to obtain the absorbance change rate sequence of the human breast cancer cell line. The absorbance values ​​of the human normal hepatocyte cell line at each experimental time point were extracted from the multidimensional pharmacodynamic response spectrum, and the rate of change of absorbance values ​​between adjacent experimental time points was calculated to obtain the absorbance change rate sequence of the human normal hepatocyte cell line. The absorbance change rate sequences of the human liver cancer cell line, the human breast cancer cell line, and the human normal liver cell line were aligned according to the experimental time points to generate a comparative change rate table. Obtain the standard efficacy change rate reference interval corresponding to the target efficacy test path, compare each change rate value in the comparison change rate table with the standard efficacy change rate reference interval, and mark the test time point and cell line type corresponding to the change rate value that exceeds the standard efficacy change rate reference interval; The comparative change rate table and labeling results are arranged in a structured manner according to the experimental time sequence to generate the pharmacodynamic analysis report.

10. The Mongolian medicine plum blossom herb efficacy testing and analysis system according to claim 9, characterized in that, The reference range for the standard efficacy change rate is derived from historical measurement data of the same batch of Mongolian herbal medicine Meihuacao standard under the same test conditions.

Images