A traditional chinese medicine prescription efficacy quantification method and system

Abstract

The application belongs to the technical field of traditional Chinese medicine, and particularly relates to a traditional Chinese prescription efficacy quantitative method and system, wherein a chemical component spectrum, a pharmacodynamic substance content and a biological activity index of a traditional Chinese medicine in a prescription composition are obtained; a multi-target regulation network is constructed according to the chemical component spectrum, the pharmacodynamic substance content and the biological activity index, combined with a traditional Chinese medicine-target interaction database; a Louvain algorithm is used to identify a core efficacy pathway cluster; a pathway target membership weighted model is used to calculate a pathway activation intensity value; according to the core efficacy pathway cluster, a mapping relationship between the core efficacy pathway cluster and a traditional Chinese efficacy term is established through ontology and semantic mapping; according to the pathway activation intensity value, a prescription efficacy intensity vector is generated through a weighted aggregation model; and according to the prescription efficacy intensity vector, a network weight is verified and corrected through an in-vitro cell model, and a calibration coefficient is combined with a clinical syndrome to generate a prescription efficacy quantification. Thus, the problems of single data dimension and insufficient quantification result accuracy in the prior art are solved.

Description

Technical Field

[0001] This invention belongs to the field of traditional Chinese medicine technology, specifically relating to a method and system for quantifying the efficacy of traditional Chinese medicine prescriptions. Background Technology

[0002] Traditional Chinese medicine (TCM) formulas are an important tool in TCM clinical treatment, and accurate evaluation of their efficacy is crucial for rational drug use and improving clinical outcomes. However, traditional TCM assessment of formula efficacy relies primarily on the experience and subjective judgment of TCM practitioners, lacking objective and quantitative evaluation methods.

[0003] While some quantitative attempts have been made in TCM research, most have focused on a single or a few attribute dimensions, failing to comprehensively consider the multidimensional attributes of TCM and their interactions. Furthermore, existing quantitative methods are largely based on static models, lacking deep integration with modern statistical methods. They are unable to effectively handle multi-source heterogeneous data or dynamically optimize based on clinical data feedback, resulting in insufficient accuracy of quantitative results. This makes the quantitative evaluation of the efficacy of TCM prescriptions insufficient to meet the needs of modern TCM development, especially in scenarios such as precision medicine, new drug development, and quality standardization control. There is an urgent need for a comprehensive, dynamic quantitative method that can comprehensively consider the multidimensional attribute relationships of TCM, integrate cutting-edge statistical analysis techniques, and continuously optimize through clinical feedback to promote the scientific and modern process of TCM research. Summary of the Invention

[0004] This application provides a method and system for quantifying the efficacy of traditional Chinese medicine prescriptions, in order to solve the problems of lack of dynamic optimization mechanism and insufficient accuracy of quantitative results in the prior art.

[0005] The first aspect of this application provides a method for quantifying the efficacy of traditional Chinese medicine (TCM) formulas, comprising the following steps: obtaining the chemical composition spectrum, active ingredient content, and bioactivity indicators of the TCM components of the formula; constructing a multi-target regulatory network for the formula's effects based on the chemical composition spectrum, active ingredient content, and bioactivity indicators, combined with a TCM-target interaction database; identifying core efficacy pathway clusters using the Louvain algorithm; calculating pathway activation intensity values ​​using a pathway target membership weighted model; establishing a mapping relationship between the core efficacy pathway clusters and TCM efficacy terms using ontology mapping and semantic mapping; generating a formula efficacy intensity vector using a weighted aggregation model based on the pathway activation intensity values; verifying target regulation efficiency on an in vitro cell model based on the formula efficacy intensity vector to correct network weights; and using the clinical syndrome integral change rate to feed back the efficacy vector calibration coefficient, thereby generating a quantitative representation of the formula's efficacy.

[0006] Preferably, based on the chemical composition spectrum, pharmacodynamic substance content, and bioactivity indicators, a multi-target regulatory network for the effects of the prescription is constructed in conjunction with a traditional Chinese medicine-target interaction database. This includes: constructing a standardized basic data layer; directly matching known component-target mapping relationships in the database based on the standardized basic data layer; inferring missing associations based on molecular structure and activity data using a similarity algorithm; mining literature to supplement unrecorded information; verifying the accuracy of unrecorded component-target mapping relationships by combining component content and bioactivity data; determining the level of evidence based on the component-target mapping relationship by the mapping source; calculating mapping weights using a linear weighted model based on the level of evidence and the component content and bioactivity; establishing a weight system by verifying credibility and removing outliers; and constructing a binary network for single-herb traditional Chinese medicine by defining chemical components and targets as network nodes and mapping weights as edges, and merging multiple binary networks of single-herb traditional Chinese medicine into a multi-target regulatory network for the prescription by accumulating or averaging the weights of the same components and edges.

[0007] Preferably, the Louvain algorithm is used to identify core functional pathway clusters, including: constructing the Louvain algorithm; dividing the multi-target regulatory network into modules according to the Louvain algorithm, and selecting high-confidence modules as candidate core pathway clusters by combining module degree and internal density indices; and identifying core functional pathway clusters by performing enrichment analysis on the gene ontology and KEGG pathway of the target in each pathway cluster according to the candidate core pathway clusters and combining the network topology index screening method.

[0008] Preferably, the Louvain algorithm formula is:

[0009] in, Modularity; The sum of the weights of all edges in the network; For any two target points in the network; Target With target Edge weights between them; Target The sum of the weights of all connected edges; Target The sum of the weights of all connected edges; Target The module tag it belongs to; Target The module tag to which it belongs. This is an indicator function.

[0010] Preferably, the formula for the pathway target membership weighted model is:

[0011]

[0012]

[0013] in, For pathway The activation strength value; For pathways Membership degree; Target The activity weight; target Activation direction; path The total number of targets included in the calculation; Target The original activity value; It is the minimum value among all the original values ​​of the target points; The maximum value among all original target values; This represents the upper limit of the summation; For index variables; Target With indicators The correlation between them.

[0014] Preferably, based on the core efficacy pathway clusters, a mapping relationship between the core efficacy pathway clusters and TCM efficacy terms is established through ontology mapping and semantic mapping. Based on the pathway activation intensity values, a formula efficacy intensity vector is generated through a weighted aggregation model, including: constructing a basic ontology library; obtaining the initial mapping relationship and weights between the core efficacy pathway clusters and TCM efficacy terms through structured rule matching based on the basic ontology library; calculating semantic similarity using a cosine similarity algorithm based on the initial mapping relationship and weights, and simultaneously optimizing and correcting the mapping weights using a weighted linear combination algorithm; and calculating the intensity value of each efficacy in the formula through a weighted aggregation model based on the corrected mapping weights and the pathway activation intensity values ​​to generate a formula efficacy intensity vector.

[0015] Preferably, the weighted aggregation model formula is:

[0016]

[0017]

[0018] in, For the first The original activation intensity of each pathway; This represents the standardized value after processing; It is the minimum value of the activation values ​​of all pathways; The maximum value of all pathway activation values; For the first The intensity of each efficacy dimension; For the first The weight of each path; The number of pathways participating in the calculation of this dimension; This represents the vector of efficacy intensity of the prescription. This represents the total number of efficacy dimensions.

[0019] A second aspect of this application provides a system for quantifying the efficacy of traditional Chinese medicine (TCM) formulas, comprising: an acquisition module for acquiring the chemical composition spectrum, active ingredient content, and bioactivity indicators of the TCM components of the formula; a construction module for constructing a multi-target regulatory network of the formula's effects based on the chemical composition spectrum, active ingredient content, and bioactivity indicators, combined with a TCM-target interaction database, identifying core efficacy pathway clusters using a modular algorithm, and calculating pathway activation intensity values ​​using a pathway target membership weighted model; an association module for establishing a mapping relationship between the core efficacy pathway clusters and TCM efficacy terms through ontology mapping and semantic mapping, and generating a formula efficacy intensity vector using a weighted aggregation model based on the pathway activation intensity values; and a verification module for verifying target regulation efficiency on an in vitro cell model based on the formula efficacy intensity vector to correct network weights, and feeding back the efficacy vector calibration coefficients with the clinical syndrome integral change rate to generate quantitative formula efficacy.

[0020] A third aspect of this application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the program to implement the method for quantifying the efficacy of traditional Chinese medicine prescriptions as described in the above embodiments.

[0021] A fourth aspect of this application provides a computer-readable storage medium having a computer program stored thereon, which is executed by a processor to implement the method for quantifying the efficacy of traditional Chinese medicine prescriptions as described in the above embodiments.

[0022] Therefore, this application has the following beneficial effects: This application constructs a standardized basic data layer covering chemical composition spectra, pharmacodynamic substance content, and bioactivity indicators. Combined with a database of TCM-target interactions, it uses similarity algorithms for inference and literature mining to supplement missing associations. A mapping weight system is established using a weighted model of evidence level, component content, and bioactivity, forming a multi-dimensional attribute-integrated multi-target regulatory network for prescriptions, overcoming the limitations of single-dimensional quantification. Secondly, the Louvain algorithm, combined with modularity, gene ontology, and KEGG pathway enrichment analysis, is used to screen core efficacy pathway clusters from both network topology and biological function dimensions. Cosine similarity and weighted linear combination algorithms are used to optimize the mapping weights of TCM efficacy terms, achieving statistical modeling of multi-source heterogeneous data. Finally, in vitro cell models are used to verify target regulation efficiency and correct network weights. The clinical syndrome integral change rate is used to feed back into the efficacy vector calibration coefficient, improving the accuracy of efficacy quantification results and meeting the modern needs of clinical precision medicine and new drug development. Thus, it solves the problems of single data dimensions, static models, and insufficient accuracy of quantification results in existing technologies.

[0023] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0024] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a flowchart of a method for quantifying the efficacy of a traditional Chinese medicine prescription according to an embodiment of this application; Figure 2 An example diagram illustrating the establishment of a multi-target regulatory network for a traditional Chinese medicine formula, Yiqi Tang, according to an embodiment of this application; Figure 3 This is an example diagram illustrating the identification of core functional pathway clusters in the traditional Chinese medicine formula Ma Xing Shi Gan Tang according to an embodiment of this application; Figure 4 This is an example diagram illustrating the calculation of pathway activation intensity values ​​using the traditional Chinese medicine formula Longdan Xiegan Tang according to one embodiment of this application; Figure 5 This is an example diagram illustrating the calculation of efficacy mapping weights for the traditional Chinese medicine formula Xuefu Zhuyu Decoction according to an embodiment of this application; Figure 6 This is an example diagram illustrating the calculation of the efficacy intensity vector of the traditional Chinese medicine formula Shenling Baizhu San according to an embodiment of this application; Figure 7 This is a flowchart illustrating a method for quantifying the efficacy of the traditional Chinese medicine formula Yinqiao San according to an embodiment of this application; Figure 8This is a schematic diagram of a quantitative system for the efficacy of traditional Chinese medicine prescriptions provided in accordance with an embodiment of this application. Figure 9 This is a schematic diagram of the structure of an electronic device provided according to an embodiment of this application. Detailed Implementation

[0025] The embodiments of this application are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0026] The following describes a method and system for quantifying the efficacy of traditional Chinese medicine (TCM) formulas according to embodiments of this application, with reference to the accompanying drawings. Addressing the issue of insufficient accuracy in quantification results mentioned in the background section, this application provides a method for quantifying the efficacy of TCM formulas. This method constructs a standardized basic data layer encompassing chemical composition spectra, pharmacodynamic substance content, and bioactivity indicators. Combined with a TCM-target interaction database, it uses similarity algorithms for inference and literature mining to supplement missing associations. A mapping weight system is established using a weighted model of evidence level, component content, and bioactivity, forming a multi-target regulatory network of the formula that integrates multi-dimensional attributes, overcoming the limitations of single-dimensional quantification. Secondly, the Louvain algorithm, combined with modularity, gene ontology, and KEGG pathway enrichment analysis, is used to screen core efficacy pathway clusters from both network topology and biological function dimensions. Cosine similarity and weighted linear combination algorithms are used to optimize the mapping weights of TCM efficacy terms, achieving statistical modeling of multi-source heterogeneous data. Finally, the network weights are corrected by verifying target regulation efficiency using in vitro cell models. The efficacy vector calibration coefficient is then fed back by the rate of change in clinical syndrome scores, improving the accuracy of the efficacy quantification results and meeting the modern needs of clinical precision medicine and new drug development. This solves the problems of single data dimension, static model, and insufficient accuracy of quantification results in existing technologies.

[0027] Specifically, Figure 1 This is a flowchart illustrating the method for quantifying the efficacy of traditional Chinese medicine prescriptions provided in the embodiments of this application.

[0028] like Figure 1 As shown, this method for quantifying the efficacy of traditional Chinese medicine prescriptions includes the following steps: In step S101, the chemical composition spectrum, content of active ingredients and bioactivity indicators of the Chinese herbal medicines that make up the prescription are obtained.

[0029] Among them, chemical composition spectrum refers to information such as the types, structures and relative proportions of chemical components contained in each component of Chinese medicine in the prescription; pharmacologically active substance content refers to the specific content of chemical components with clear pharmacological activity in the prescription; and bioactivity index refers to quantitative parameters that reflect the influence of the prescription on the physiological and pathological processes of biological organisms.

[0030] It is understood that the embodiments of this application can clarify the types, structures and proportions of chemical components in traditional Chinese medicine prescriptions through chemical composition, providing a basis for screening active ingredients and constructing component-target mapping; the content of active ingredients can be quantified through the content of active substances, assessing the intensity of the effect of ingredients on targets and establishing dose-effect relationships; the bioactivity indicators can reflect the quantitative parameters of the effect of prescriptions on the physiological and pathological processes of biological organisms, verifying the accuracy of component-target mapping and providing a basis for calculating pathway activation intensity; together, they form the data foundation for the quantitative analysis of the efficacy of traditional Chinese medicine prescriptions.

[0031] For example, in the traditional Chinese medicine formula Baihu Tang, chemical composition analysis identified more than 20 compounds, including gypsum, anemarrhena, and licorice. Among these, anemarrhena saponin AⅢ was found to contain approximately 1.5 mg / g, mangiferin approximately 0.8 mg / g, and glycyrrhizic acid approximately 2.0 mg / g. In vitro bioactivity assays showed that when the concentration of anemarrhena saponin AⅢ was 10 μM, it reduced the expression of tumor necrosis factor-α (TNF-α) in inflammatory cells by 40%; and when the concentration of mangiferin was 20 μg / mL, it inhibited Coxsackievirus B3 by 50%. These data clarified the relationship between the types and contents of components and their biological effects, providing a basis for screening antiviral and anti-inflammatory active ingredients in Baihu Tang, constructing component-target mapping relationships, assessing the intensity of component-target effects, and establishing dose-response relationships. Furthermore, they verified the accuracy of the component-target mapping and provided a basis for calculating pathway activation intensity, collectively laying a data foundation for the quantitative analysis of the efficacy of traditional Chinese medicine formulas.

[0032] In step S102, based on the chemical composition spectrum, the content of pharmacodynamic substances and bioactivity indicators, a multi-target regulatory network for the efficacy of the prescription is constructed in conjunction with the traditional Chinese medicine-target interaction database. The Louvain algorithm is used to identify the core efficacy pathway clusters, and the pathway activation intensity value is calculated through the pathway target membership weighted model.

[0033] Among them, the pathway activation intensity value refers to the key indicator used to quantify the degree of activation of biological pathway functions under the action of traditional Chinese medicine prescriptions. Through mathematical modeling, the correlation strength between targets and pathways, target activity and direction of action are integrated to form a quantifiable value that reflects the functional activity level of pathways under the regulation of prescriptions.

[0034] It is understood that the embodiments of this application quantify the degree of regulation of biological pathways by traditional Chinese medicine formulas by integrating the correlation between targets and pathways, target activity and activation direction, connect the molecular mechanism of formulas with TCM efficacy terminology, generate efficacy intensity vectors through weighted aggregation, and optimize efficacy quantification by using in vitro cell models for verification and clinical syndrome integration, providing key quantitative support for the scientific evaluation, mechanism analysis and precision research of formula efficacy.

[0035] For example, the traditional Chinese medicine formula Si Ni Tang contains 15 components, including aconitine (0.25 mg / g), hypoaconitine (0.18 mg / g), and 6-gingerol (2.1 mg / g). In vitro experiments showed that 5 μmol / L aconitine increased the activity of Ca²⁺-ATPase in cardiomyocytes by 20%, and 10 μg / mL 6-gingerol inhibited IL-1β expression in LPS-induced RAW264.7 cells by 35%. Based on these data, a multi-target regulatory network was constructed through component-target mapping relationships, and the Louvain algorithm was used to identify core functional pathway clusters such as the Na⁺ channel regulatory pathway and the TRPV1 inflammatory pathway. Using a pathway target membership weighted model, the activation intensity value of the Na⁺ channel pathway was calculated to be 0.73, and the activation intensity value of the TRPV1 inflammatory pathway was 0.67. This intensity value integrates component content, bioactivity, and target correlation. Through ontology mapping, it establishes weighted correlations with efficacy terms such as "restoring yang and rescuing from collapse" and "warming the interior and dispelling cold." After weighted aggregation, it generates an efficacy intensity vector. Combined with the action potential time history verification in cardiomyocytes in an in vitro cell model, the efficacy intensities of restoring yang and rescuing from collapse and warming the interior and dispelling cold of Si Ni Tang are finally quantified as 0.561 and 0.443, respectively. This highlights the key role of pathway activation intensity value in connecting molecular mechanisms and TCM efficacy in the quantitative analysis of formula efficacy.

[0036] In this embodiment, a multi-target regulatory network for the effects of traditional Chinese medicine (TCM) is constructed based on the chemical composition spectrum, pharmacodynamic substance content, and bioactivity indicators, combined with a TCM-target interaction database. This includes: constructing a standardized basic data layer; directly matching known component-target mapping relationships in the database based on the standardized basic data layer; inferring missing associations based on molecular structure and activity data using a similarity algorithm; mining literature to supplement unrecorded information; verifying the accuracy of unrecorded component-target mapping relationships using component content and bioactivity data; determining the level of evidence based on the component-target mapping relationship through the mapping source; calculating mapping weights using a linear weighted model based on the level of evidence, combined with component content and bioactivity; establishing a weight system by verifying credibility and removing outliers; and constructing a binary network for single-herb TCM by defining chemical components and targets as network nodes and mapping weights as edges, and merging multiple binary networks of single-herb TCM into a multi-target regulatory network for the TCM through the accumulation or averaging of weights for the same components and edges.

[0037] The similarity algorithm refers to an algorithm that predicts the potential interaction between a chemical component and a target by calculating the similarity of chemical molecular structures or biological activity characteristics based on molecular structure and activity data. The weighting coefficients are set by expert experience and domain knowledge, and the formula is as follows:

[0038]

[0039]

[0040] in, For ingredients Similarity coefficient in molecular structure with known active ingredients; For ingredients Molecular fingerprints; A molecular fingerprint of a known component with specific biological activity; For ingredients With target Similarity coefficient in biological activity; The number of times a bioactivity index is measured; For ingredients In the Measurement values ​​of a bioactivity index; For ingredients The average value measured across all bioactivity indicators; Target The relevant first Biological effect indicators; Target The average value measured across all biological effect indicators; For ingredients With target The overall similarity coefficient; Weighting coefficients for molecular structural similarity; This represents the weighting coefficient for biological activity similarity.

[0041] It is understood that the embodiments of this application infer the missing associations of components and targets based on molecular structure and activity data, supplement information not included in the database, expand the data coverage of the multi-target regulatory network, comprehensively evaluate potential effects by calculating molecular structure similarity and biological activity correlation, improve the reliability of association inference, provide data support, and make the multi-target regulatory network more in line with actual biological mechanisms.

[0042] For example, in the traditional Chinese medicine formula Huanglian Jiedu Decoction, the berberine content is 3.2 mg / g. In vitro experiments showed that 20 μg / mL could increase the expression level of p53 protein in colorectal cancer cells by 42%. Paclitaxel is known to target microtubules. The molecular structural similarity between berberine and paclitaxel was calculated using a similarity algorithm, and the structural similarity, combined with their molecular fingerprints, was 0.68. Simultaneously, the regulatory effect of berberine on p53 was correlated with the biological activity of cell cycle pathways related to microtubule targets at a correlation of 0.79. Setting the weights for molecular structure and biological activity to 0.3 and 0.7 respectively, and the overall similarity to 0.75, this component-target mapping relationship was incorporated into a database and used as edge weights along with other component-target associations to construct a multi-target regulatory network for Huanglian Jiedu Decoction containing multiple components and targets, providing data support for subsequent identification of core functional pathway clusters.

[0043] It should be noted that a linear weighted model is a mathematical model that calculates results by assigning appropriate weights to different factors and performing linear combinations. The linear weighted formula is as follows:

[0044] in, For mapping weights; The number of factors involved in the weighted calculation; Various properties related to the ingredients; The weight coefficient corresponding to the attribute; This is the index variable used in the summation operation.

[0045] The linear weighted model also includes the analytic hierarchy process (AHP) formula:

[0046] in, This is the weight vector; The sum of the number of influencing factors considered; This is the starting index for the product operation; To consider from a specific perspective, the components Relative to ingredients Relative importance; To consider from a specific perspective, component k is relative to component Relative importance; It is the largest eigenvalue; To determine the matrix With weight vector The product; for The One component; for The One component; Consistency ratio; It is a random consistency indicator.

[0047] It is understood that the embodiments of this application quantify the component-target mapping weights in a linear combination by comprehensively considering multiple dimensions of evidence level, component content, and bioactivity, integrating the credibility differences of data from different sources, flexibly adjusting the influence of each factor on the mapping relationship, laying the foundation for constructing a reliable multi-target regulatory network, and improving the scientificity and accuracy of the weighting system by eliminating outliers through subsequent verification.

[0048] For example, such as Figure 2 As shown, in the traditional Chinese medicine formula Buzhong Yiqi Tang, the content of astragaloside A is 2.1 mg / g. In vitro experiments showed that 30 μg / mL could increase the expression level of IGF-1 in fibroblasts by 35%. Based on the consensus in the field of traditional Chinese medicine on the mechanism of action of components and targets, a judgment matrix was constructed by AHP. The direct influence of biological activity on cell proliferation regulation was set to have a higher weight than that of molecular structure similarity. The pharmacological threshold of component content as a material basis needs to be comprehensively considered. After column normalization and row averaging, the weighting coefficients of molecular structure similarity (weight 0.22), biological activity correlation (0.61), and component content (0.17) were determined. The rationality of the weight allocation was verified by the consistency test (CR=0.031<0.1). The astragaloside A content was normalized to 0.52 and substituted into the linear weighting formula to calculate the mapping weight of astragaloside A→IGF-1R target as 0.69. Meanwhile, the same method was used to obtain the component-target mapping weights of glycyrrhizic acid → NF-κB (weight 0.58) and Codonopsis pilosula polysaccharide → IL-2 (weight 0.63). Chemical components and targets were defined as network nodes, and mapping weights were used as edges. First, a binary network was constructed based on the component-target data of single Chinese medicines such as Astragalus membranaceus and Glycyrrhiza uralensis. Then, by accumulating the edge weights of the same components, the multi-target regulatory network of Buzhong Yiqi Decoction was generated.

[0049] In this embodiment of the application, the Louvain algorithm is used to identify core functional pathway clusters, including: constructing the Louvain algorithm; dividing the multi-target regulatory network into modules according to the Louvain algorithm, and selecting high-confidence modules as candidate core pathway clusters by combining module degree and internal density indicators; and identifying core functional pathway clusters by performing enrichment analysis on the gene ontology and KEGG pathway of the target in each pathway cluster according to the candidate core pathway clusters and combining the network topology index screening method.

[0050] Enrichment analysis involves statistically analyzing the gene function and pathways of the target to screen for biological processes or signaling pathways related to the efficacy of the prescription. The formula is:

[0051] in, The probability of significance for enrichment; This refers to the number of genes belonging to this functional category or pathway among the candidate targets; This refers to the number of genes in the entire genome that belong to a certain functional category or pathway. The total number of genes in the candidate target set; The total number of genes in the entire genome; It is an index variable.

[0052] It is understood that the embodiments of this application screen biological processes or signaling pathways related to the efficacy of prescriptions by performing gene ontology and KEGG pathway statistical analysis on the targets in the candidate core pathway clusters, verifying the biological relevance of the pathway clusters from a functional perspective, identifying core efficacy pathway clusters with biological function enrichment characteristics, providing functional basis for the quantitative analysis of prescription efficacy, improving the accuracy and reliability of core pathway identification, connecting multi-target network analysis with the mapping of TCM efficacy terms, and improving the biological rationality of prescription efficacy analysis.

[0053] It should be noted that the network topology index screening method is based on the structural characteristics of multi-target control networks. It filters core functional pathway clusters by quantifying the importance of nodes' positions within the network and the tightness of module connections. The formula is:

[0054]

[0055]

[0056] in, For nodes Degree centrality; For nodes The number of directly connected edges; The total number of nodes in the network; For nodes Betweenness centrality; For nodes To the node The number of shortest paths; For the nodes nodes To the node The number of shortest paths; For nodes Proximity centrality; For nodes To the node The shortest path distance.

[0057] It is understood that the embodiments of this application, by regulating the network structure characteristics based on multiple targets and by quantifying indicators such as node degree centrality, betweenness centrality, and proximity centrality, evaluate the pivotal role of nodes in the network, information transmission efficiency, and module connection tightness, identify targets and pathways that play a key role in network connectivity from the topological structure level, provide a structural scientific basis for the quantitative analysis of prescription efficacy, and improve the reliability of core pathway cluster screening.

[0058] For example, such as Figure 3 As shown, in the quantitative analysis of the efficacy of the traditional Chinese medicine formula Ma Xing Shi Gan Tang, its chemical composition spectrum was obtained. A multi-target regulatory network was constructed using a database of Chinese herbal medicine targets, including ephedrine, amygdalin, and their potential targets. The Louvain algorithm was used to divide the network into modules, and two candidate modules with high Q-values ​​and high internal density were selected. Gene Ontology (GO) and KEGG pathway enrichment analyses were performed on the targets in the candidate modules. It was found that the targets in one module were significantly enriched in the "bronchial smooth muscle relaxation" biological process (p<0.01) and the "cAMP-PKA signaling pathway." This module contained eight targets, of which four targets, including PKA and cAMP, showed an enrichment fold of more than 2.2 times in the pathway. Meanwhile, the importance of targets within the module was assessed using a network topology index screening method: PKA's node degree (connecting 10 edges) and betweenness centrality (0.75) both ranked in the top 3, indicating its pivotal role in the network; cAMP's proximity centrality (0.70) showed its high communication efficiency with other targets. Combining enrichment analysis and topology indices, this module was identified as a core functional pathway cluster, corresponding to the traditional Chinese medicine effect of "ventilating the lungs and relieving asthma." Subsequent calculations using a weighted aggregation model showed that the activation intensity value of this pathway cluster was 0.83, consistent with the experimental results of Ma Xing Shi Gan Tang promoting bronchial smooth muscle relaxation in an in vitro cell model, thus verifying the rationality of the screening method.

[0059] In this embodiment of the application, the Louvain algorithm formula is:

[0060] in, Modularity; This represents the sum of the weights of all edges in the network. , For any two target points in the network; Target With target Edge weights between them; Target The sum of the weights of all connected edges; Target The sum of the weights of all connected edges; Target The module tag it belongs to; Target The module tag it belongs to; This is an indicator function.

[0061] It is understood that the embodiments of this application, based on the topology of a multi-target regulatory network, efficiently partition the network by maximizing the modularity Q-value, and identify candidate core pathway clusters with high confidence. This allows for the screening of tightly connected and structurally significant modules from complex networks, providing a foundation for subsequent screening of core efficacy pathway clusters using enrichment analysis and network topology indicators. It reduces network complexity, focuses on key pathways, and supports the quantitative analysis of the efficacy of prescriptions.

[0062] For example, in the efficacy study of the traditional Chinese medicine formula Liuwei Dihuang Wan, its chemical composition spectrum was obtained, including components such as paeonol and ursolic acid. A multi-target regulatory network containing 36 components and 65 targets was constructed using a database of traditional Chinese medicine targets. These targets include apoptosis-related proteins such as AKT and Bcl-2. This network was then imported into the Louvain algorithm model. The algorithm iteratively optimizes the network's module structure and calculates the module degree (Q-value) for each possible module partitioning method. During module partitioning, the Louvain algorithm first treats each target as an independent module, then gradually merges adjacent modules, calculating the change in module degree with each merge and retaining merge operations that increase the module degree until no further improvement is possible. After several iterations, the algorithm divides the network into four initial modules. Subsequently, combining the module degree (Q>0.35) and internal density index, two high-confidence candidate modules are selected. One of these modules contains 12 targets, with the sum of its internal edge weights accounting for 18% of the total network edge weights, indicating tight internal connections. Gene ontology enrichment analysis of this module revealed significant enrichment in the biological process of "apoptosis regulation." Network topology screening identified targets such as AKT and Bcl-2 as having high node degree and betweenness centrality, ultimately determining this module as a core functional pathway cluster, corresponding to the traditional Chinese medicine efficacy of Liuwei Dihuang Wan (a traditional Chinese medicine formula) of "nourishing Yin and tonifying the kidneys." This module segmentation process extracted key functional modules from a complex network, laying the foundation for subsequent quantitative efficacy analysis.

[0063] In this embodiment of the application, the formula for the pathway target membership weighted model is:

[0064]

[0065]

[0066] in, For pathway The activation strength value; For pathways Membership degree; Target The activity weight; target Activation direction; path The total number of targets included in the calculation; Target The original activity value; It is the minimum value among all the original values ​​of the target points; The maximum value among all original target values; This represents the upper limit of the summation; For index variables; Target With indicators The correlation between them.

[0067] It is understood that the embodiments of this application quantify the activation intensity value of pathways by integrating the membership degree, activity weight, and activation direction of targets to pathways, providing a quantitative basis for the screening of core functional pathway clusters and serving as the basis for generating the efficacy intensity vector of prescriptions. By normalizing the target activity weights, the importance of different targets is dynamically adjusted to reflect the contribution of highly active or highly correlated targets to the pathway, filter out the influence of low-confidence targets, improve the reliability of prescription efficacy analysis, and provide a scientific quantitative method for calculating the functional intensity of multi-target regulatory networks.

[0068] For example, such as Figure 4 As shown, in the efficacy study of the traditional Chinese medicine formula Longdan Xiegan Tang, a pathway target membership weighted model was used to calculate the activation intensity of the core efficacy pathway cluster "MAPK signaling pathway". By obtaining the content data of components such as gentiopicrin and baicalin in Longdan Xiegan Tang, and combining them with a target database, nine related targets, including ERK, JNK, and p38, were identified. The model first calculated the membership degree of each target to the MAPK pathway. For example, the average correlation between ERK and five inflammatory markers in the pathway reached 0.76, and the membership degree was assigned to 0.76; the average correlation between JNK and three markers was 0.67, and the membership degree was 0.67. Next, the target activity weights were calculated: the in vitro anti-inflammatory activity data of each target (such as the inhibition rate of inflammatory factors) were normalized. After normalization, the activity value of ERK was weighted at 0.83, and that of JNK at 0.74. Simultaneously, the activation direction was determined (+1 indicates activation, -1 indicates inhibition). All targets in the Gentianae Radix et Rhizoma Decoction exhibited inhibition of the MAPK pathway, therefore the activation direction was set to -1. Finally, the model weighted and integrated membership degrees, activity weights, and activation directions to calculate the pathway activation intensity value: the contribution of ERK was 0.76 × 0.83 × (-1), JNK was 0.67 × 0.74 × (-1), and so on. After summing these values ​​and dividing by the total membership degree, the activation intensity value of the MAPK pathway was obtained as -0.75, indicating that the Gentianae Radix et Rhizoma Decoction has a significant inhibitory effect on this pathway.

[0069] In step S103, based on the core efficacy pathway cluster, a mapping relationship between the core efficacy pathway cluster and TCM efficacy terms is established through ontology mapping and semantic mapping. Based on the pathway activation intensity value, a formula efficacy intensity vector is generated through a weighted aggregation model.

[0070] Among them, the core functional pathway cluster refers to a functional set of interconnected signaling pathways in an organism. They work together to regulate key physiological activities or disease-related processes, playing a core driving role in maintaining homeostasis, exerting drug therapeutic effects, or mediating the occurrence and development of diseases.

[0071] It is understood that the embodiments of this application obtain efficacy pathway clusters through multi-target regulatory network module division and biological function enrichment analysis, which can accurately locate the key biological signaling pathways of the prescription, provide molecular mechanism support for efficacy quantification, and improve the accuracy of mechanism analysis of efficacy quantification of traditional Chinese medicine prescriptions.

[0072] In this embodiment, based on the core efficacy pathway clusters, a mapping relationship between the core efficacy pathway clusters and TCM efficacy terms is established through ontology mapping and semantic mapping. Based on the pathway activation intensity values, a formula efficacy intensity vector is generated through a weighted aggregation model. This includes: constructing a basic ontology library; obtaining an initial mapping relationship and weights between the core efficacy pathway clusters and TCM efficacy terms through structured rule matching based on the basic ontology library; calculating semantic similarity using a cosine similarity algorithm based on the initial mapping relationship and weights, and simultaneously optimizing and correcting the mapping weights using a weighted linear combination algorithm; and calculating the intensity value of each efficacy in the formula using a weighted aggregation model based on the corrected mapping weights and the pathway activation intensity values ​​to generate a formula efficacy intensity vector.

[0073] Among them, structured rule matching refers to the structural matching and initial weighting of core efficacy pathway clusters and TCM efficacy terms based on rules such as concept hierarchy, attribute association, and synonym mapping in the basic ontology library, combined with data field correspondence, semantic causal logic, and evidence level priority.

[0074] It should be noted that the cosine similarity algorithm is a method that measures the degree of similarity between two vectors by calculating the cosine of the angle between them. The formula is:

[0075] in, The value is the cosine. The feature vector of the core functional pathway cluster; The feature vector of TCM efficacy terms; The number of features used for matching; For vectors The Middle Feature values ​​in each dimension; For vectors The Middle Feature values ​​in each dimension; For dimensional indexing.

[0076] Linear weighted combination refers to combining the initial weights obtained from structured rule matching with the semantic similarity values ​​calculated by the cosine similarity algorithm, and then using a linear weighting method that incorporates expert experience or domain-specific requirements to correct the accuracy of the mapping relationship. The formula is as follows:

[0077] in, For mapping weights; These are the initial weights; This represents the semantic similarity value. These are preset coefficients.

[0078] It is understood that the embodiments of this application quickly establish the initial mapping and weights of core efficacy pathway clusters and TCM efficacy terms through predefined rules, thereby improving the prior rationality and reliability of the mapping and providing a structured knowledge foundation for weight optimization. By converting pathway clusters and efficacy terms into semantic space vectors and calculating the cosine value of the angle between them to quantify the degree of semantic association, the shortcomings of rule matching at the semantic level are supplemented. By fusing structured rule weights and semantic similarity values ​​through preset coefficients, the mapping weights are dynamically optimized, balancing the contributions of prior knowledge and semantic logic. This makes the mapping relationship more in line with the needs of connecting TCM and Western medicine theories, improves the mapping accuracy of core efficacy pathway clusters and TCM efficacy terms, lays the foundation for generating scientific formula efficacy intensity vectors, and enhances the theoretical logic and clinical applicability of quantitative TCM formula efficacy.

[0079] For example, such as Figure 5As shown, in the efficacy study of the traditional Chinese medicine formula Xuefu Zhuyu Decoction, when determining the mapping weights between the "PI3K-AKT signaling pathway" and the "blood-activating and stasis-removing" effects, initial weights were first obtained through structured rule matching. Based on traditional Chinese medicine theory and literature, components such as aucubin and safflower yellow pigment in Xuefu Zhuyu Decoction can activate the PI3K-AKT pathway, promote the proliferation and migration of vascular endothelial cells, and improve microcirculation disorders. This is directly related to the mechanism of "blood-activating and stasis-removing" effects in clearing blood vessels and dispersing stagnation. Combined with the evidence from the literature, an initial mapping weight of 0.65 was assigned. Next, the cosine similarity algorithm was used to calculate semantic association. Terms related to the "PI3K-AKT signaling pathway" (such as "angiogenesis" and "cell survival regulation") and terms related to "blood-activating and stasis-removing" (such as "removing stasis and clearing the meridians" and "activating blood and relieving pain") were extracted. The abstracts of the literature in PubMed and CNKI were segmented to construct a word vector space, and the semantic similarity between the two was calculated to be 0.78. Finally, the weights were optimized through weighted linear combination. With an initial weighting factor of 0.7 and a semantic similarity factor of 0.3, the corrected mapping weight is calculated to be 0.7 × 0.65 + 0.3 × 0.78 = 0.689. This weight quantifies the contribution of the "PI3K-AKT signaling pathway" to the "blood-activating and stasis-removing" effect, providing key parameters for subsequently generating an efficacy vector based on pathway activation intensity.

[0080] In this embodiment of the application, the weighted aggregation model formula is:

[0081]

[0082]

[0083] in, For the first The original activation intensity of each pathway; This represents the standardized value after processing; It is the minimum value of the activation values ​​of all pathways; The maximum value of all pathway activation values; For the first The intensity of each efficacy dimension; For the first The weight of each path; The number of pathways participating in the calculation of this dimension; This represents the vector of efficacy intensity of the prescription. This represents the total number of efficacy dimensions.

[0084] It is understood that the embodiments of this application generate a formula efficacy intensity vector by standardizing the activation intensity of the pathway and combining it with the modified mapping weight for weighted summation. After being calibrated by feedback from clinical data, the molecular mechanism of the biological pathway is quantified and transformed into the intensity value of TCM efficacy, connecting the core efficacy pathway cluster with TCM efficacy terminology, bridging the differences between molecular mechanisms and TCM theory, and improving the accuracy and scientific nature of the formula efficacy interpretation.

[0085] For example, such as Figure 6 As shown, in the study of the traditional Chinese medicine formula Shenling Baizhu San, activity data of components such as atractylodes macrocephala lactone (0.9 mg / g) and poria cocos polysaccharide (1.5 mg / g) in this formula were obtained. After constructing a multi-target regulatory network, the PI3K-AKT metabolic pathway (activation intensity +0.68) and the intestinal mucosal barrier regulatory pathway (activation intensity +0.73) were identified as core pathways. Through ontology mapping, the mapping weight between the PI3K-AKT pathway and "strengthening the spleen and stomach" was determined to be 0.65, and the weight between the intestinal mucosal barrier pathway and "dampness elimination and diarrhea relief" was determined to be 0.72. The activation intensity of the pathways was standardized (e.g., the original value of the PI3K-AKT pathway +0.68 was standardized to 0.76, and the intestinal mucosal barrier pathway +0.73 was standardized to 0.81), and multiplied by the corresponding weights: the efficacy intensity of "strengthening the spleen and stomach" = 0.76 × 0.65 ≈ 0.494, and the efficacy intensity of "dampness-removing and diarrhea-stopping" = 0.81 × 0.72 ≈ 0.583, initially generating an efficacy vector [0.494, 0.583]. Finally, combined with the intestinal epithelial cell proliferation rate in the in vitro cell model (e.g., a 27% increase in proliferation) and clinical symptom scores (diarrhea and abdominal distension relief rate of 52%), with calibration coefficients of 1.05 and 1.12 respectively, the final efficacy vectors were obtained: "strengthening the spleen and stomach" 0.494 × 1.05 ≈ 0.519, and "dampness-removing and diarrhea-stopping" 0.583 × 1.12 ≈ 0.653, quantifying the efficacy intensity of Shenling Baizhu Powder.

[0086] In step S104, based on the efficacy intensity vector of the prescription, the target regulation efficiency is verified on an in vitro cell model to correct the network weights, and the efficacy vector calibration coefficient is fed back by the change rate of clinical syndrome scores to generate the quantitative efficacy of the prescription.

[0087] Among them, in vitro cell models refer to experimental system models that simulate physiological or pathological states in vivo by culturing cells in a controlled laboratory environment, and are used to study the effects of genes, drugs, etc. on cell function and verify related mechanisms.

[0088] It is understood that the embodiments of this application construct cell models related to the mechanism of action of prescriptions, detect the regulatory efficiency of prescription components on targets, verify the accuracy of component-target mapping relationship and pathway activation intensity value, correct the weights of multi-target regulatory network, eliminate low-confidence associations, optimize the weight system of multi-target network, provide experimental evidence support for the quantitative analysis of prescription efficacy, improve the reliability of network model, and lay a precise foundation for the calibration of efficacy vector by clinical syndrome scores.

[0089] It should be noted that the rate of change in clinical syndrome scores refers to the ratio of the difference in clinical symptom scores before and after treatment to the scores before treatment. This reflects the degree of improvement of the TCM syndrome by the prescription, and the value can serve as a quantitative indicator of actual clinical efficacy, feeding back into the calibration vector of the prescription's efficacy intensity. The calibration formula is:

[0090] in, For the calibrated first The intensity of each efficacy dimension; No. Calibration coefficients in each dimension.

[0091] It is understood that the embodiments of this application, by collecting quantitative data of patients' clinical symptoms to calibrate coefficients, bridge the gap between laboratory models and clinical reality, making the quantitative results of efficacy more consistent with the improvement of patients' symptoms, improving the clinical applicability of prescription efficacy evaluation, providing more valuable quantitative basis for medication, efficacy prediction and prescription optimization, and promoting the deep integration of TCM prescription efficacy research and clinical practice.

[0092] The method for quantifying the efficacy of traditional Chinese medicine (TCM) prescriptions proposed in this application constructs a standardized basic data layer encompassing chemical composition spectra, pharmacodynamic substance content, and bioactivity indicators. Combined with a TCM-target interaction database, it uses similarity algorithms for inference and literature mining to supplement missing associations. A mapping weight system is established using a weighted model of evidence level, component content, and bioactivity, forming a multi-dimensional, multi-target regulatory network that overcomes the limitations of single-dimensional quantification. Secondly, the Louvain algorithm, combined with modularity, gene ontology, and KEGG pathway enrichment analysis, is used to screen core efficacy pathway clusters from both network topology and biological function dimensions. Cosine similarity and weighted linear combination algorithms are used to optimize the mapping weights of TCM efficacy terms, enabling statistical modeling of multi-source heterogeneous data. Finally, the network weights are corrected by verifying target regulation efficiency using in vitro cell models. The efficacy vector calibration coefficient is then fed back by the rate of change in clinical syndrome scores, improving the accuracy of efficacy quantification results and meeting the modern needs of precision medicine and new drug development. This solves the problems of single data dimensions, static models, and insufficient accuracy of quantification results in existing technologies.

[0093] The following will illustrate a specific example of a method for quantifying the efficacy of a traditional Chinese medicine formula, such as... Figure 7 As shown, taking the quantitative analysis of the efficacy of the traditional Chinese medicine formula Yinqiao San as an example, it includes: Acquisition Phase: Yinqiao Powder is composed of traditional Chinese medicines such as honeysuckle, forsythia, and peppermint. Its chemical composition spectrum was obtained, identifying over 50 compounds including chlorogenic acid, forsythoside, and menthol. Simultaneously, the content of pharmacodynamic substances was determined; for example, the concentration of chlorogenic acid in the medicinal materials was 1.2 mg / g, forsythoside was 0.8 mg / g, and menthol was 0.5 mg / g. In the detection of bioactivity indicators, in vitro experiments showed that 15 μg / mL chlorogenic acid reduced IL-6 expression in LPS-induced RAW264.7 macrophages by 38%, and 20 μg / mL forsythoside showed a 45% inhibition rate against H1N1 virus. These data lay the foundation for subsequent analysis.

[0094] Construction phase: Direct matching of chlorogenic acid → COX-2 (Evidence Level A) and forsythoside → TNF-α (Evidence Level B) was performed using the TCMSP database; based on molecular structure and activity data, a similarity algorithm was used. The formula deduced menthol → TRPM8 (molecular fingerprint similarity 0.75), and the mapping accuracy was verified by its anti-inflammatory activity (IC50 = 12 μg / mL). A linear weighted model formula was used. Mapping weights were calculated; for example, the weight for chlorogenic acid → COX-2 was 0.76 [evidence level weight 0.4 × 1 + component content weight 0.3 × (1.2 / 2.0) + bioactivity weight 0.3 × 0.38]. Using chemical components (chlorogenic acid, forsythoside, etc.) and targets (COX-2, TNF-α, etc.) as nodes and mapping weights as edges, a binary network for single-herb traditional Chinese medicine was constructed and then merged into a prescription network (48 nodes, 82 edges). The Louvain algorithm formula was used. The module was divided into submodules, and the module degree Q=0.42 was calculated. Modules with an internal density >0.3, such as the NF-κB inflammatory pathway module (12 targets), were selected and enrichment analysis formulas were used. The GO enrichment for "inflammatory response regulation" (p=0.002) and KEGG enrichment for "NF-κB signaling pathway" (enrichment factor 2.3) were calculated to determine functional correlation. Simultaneously, a network topology index screening method was applied to analyze indicators such as node degree and betweenness centrality, filtering out core nodes such as COX-2 (node ​​degree 15, betweenness centrality 0.82) and IL-6 (closeness centrality 0.75). Combining the enrichment analysis results with topology indices, core functional pathway clusters were identified. Using a pathway target membership weighted model, taking the NF-κB pathway as an example, the membership degree of target COX-2 was μ=0.72 (average correlation with 5 inflammatory markers 0.72), with an activity weight of 0.85 (after normalization), and activation direction a=-1; for IL-6, μ=0.65, w=0.78, a=-1, and the activation direction was determined using the formula... The calculated pathway activation intensity was 0.68.

[0095] Related phase: Based on the basic ontology library, mapping relationships such as "NF-κB pathway inhibition" → "heat-clearing and detoxifying" and "viral replication regulation" → "wind-dispersing and exterior-releasing" are defined through structured rule matching, with initial weights of 0.6 and 0.5 respectively. A cosine similarity algorithm is employed. To optimize semantic association, the cosine value of the word vectors for "NF-κB pathway inhibition" and "clearing heat and detoxifying" is calculated to be 0.72. Then, a weighted linear combination formula is used. The adjusted weight is 0.636. Based on the normalized pathway activation intensity, a weighted aggregation model formula is used. The efficacy dimension strengths were calculated: "clearing heat and detoxifying" was 0.52, and "dispersing wind and releasing the exterior" was 0.53, generating an initial efficacy vector E=[0.52,0.53,...].

[0096] Verification phase: In in vitro validation using RAW264.7 macrophages, Yinqiao Powder (100 μg / mL) inhibited NF-κB nuclear translocation by 42%, consistent with the model-predicted pathway activation intensity of 0.68, and the COX-2→NF-κB side weight was adjusted to 0.82. In H1N1-infected MDCK cell experiments, Yinqiao Powder (50 μg / mL) reduced viral titer by 51%, validating the accuracy of the viral replication pathway activation intensity of 0.75. In clinical syndrome score feedback: after 60 patients with wind-heat common cold took Yinqiao Powder, the fever syndrome score changed by 78%, and the sore throat score changed by 85%. The calibration coefficients for "clearing heat and detoxifying" were calculated to be 1.5 and "relieving wind and releasing the exterior" to be 1.6. The final quantitative efficacy was: the intensity of clearing heat and detoxifying was 0.52×1.5=0.78, and the intensity of relieving wind and releasing the exterior was 0.53×1.6=0.85, which is consistent with the clinical efficacy.

[0097] In summary, this invention acquires the chemical composition spectrum, pharmacodynamic substance content, and bioactivity indicators of Yinqiao Powder, constructs a multi-target regulatory network, uses the Louvain algorithm combined with network topology indicators to screen core efficacy pathway clusters, and then establishes associations with traditional Chinese medicine efficacy terms through ontology mapping and semantic mapping to generate efficacy intensity vectors. These vectors are then dynamically calibrated using in vitro cell models and clinical symptom data to objectively quantify the efficacy of the prescription. This invention breaks through the limitations of traditional experience-based judgment, analyzes the mechanism from the perspective of multi-component and multi-target synergy, bridges the differences between traditional Chinese and Western medicine theories through network topology analysis and semantic mapping, and improves the consistency between results and clinical efficacy through dynamic calibration. It provides a quantitative path for the scientific research of traditional Chinese medicine that integrates modern technology and the holistic view of traditional Chinese medicine, promoting the precision use of traditional Chinese medicine, new drug development, and quality standardization.

[0098] Next, referring to the accompanying drawings, a quantitative system for the efficacy of traditional Chinese medicine prescriptions is described according to an embodiment of this application.

[0099] Figure 8 This is a schematic diagram of a system for quantifying the efficacy of traditional Chinese medicine prescriptions, according to an embodiment of this application.

[0100] like Figure 8 As shown, the quantitative system 10 for the efficacy of traditional Chinese medicine prescriptions includes: an acquisition module 100, a construction module 200, an association module 30, and a verification module 400.

[0101] The module 100 is used to acquire the chemical composition spectrum, active ingredient content, and bioactivity index of the Chinese herbs constituting the prescription. The module 200 is used to construct a multi-target regulatory network for the prescription's effects based on the chemical composition spectrum, active ingredient content, and bioactivity index, combined with a Chinese herbal medicine-target interaction database. It uses a modular algorithm to identify core efficacy pathway clusters and calculates pathway activation intensity values ​​through a pathway target membership weighted model. The module 300 is used to establish a mapping relationship between the core efficacy pathway clusters and Chinese medicine efficacy terms through ontology mapping and semantic mapping. Based on the pathway activation intensity values, it generates a prescription efficacy intensity vector through a weighted aggregation model. The module 400 is used to verify the target regulation efficiency on an in vitro cell model based on the prescription efficacy intensity vector to correct the network weights. It also uses the clinical syndrome integral change rate to feed back the efficacy vector calibration coefficient and generate quantitative prescription efficacy.

[0102] It should be noted that the foregoing explanation of an embodiment of a method for quantifying the efficacy of a traditional Chinese medicine prescription also applies to the system for quantifying the efficacy of a traditional Chinese medicine prescription in this embodiment, and will not be repeated here.

[0103] According to the embodiments of this application, a quantitative system for TCM prescription efficacy is proposed. This system constructs a standardized basic data layer encompassing chemical composition spectra, pharmacodynamic substance content, and bioactivity indicators. Combined with a TCM-target interaction database, it uses similarity algorithms for inference and literature mining to supplement missing associations. A mapping weight system is established using a weighted model of evidence level, component content, and bioactivity, forming a multi-dimensional prescription multi-target regulatory network that overcomes the limitations of single-dimensional quantification. Secondly, the Louvain algorithm, combined with modularity, gene ontology, and KEGG pathway enrichment analysis, is used to screen core efficacy pathway clusters from both network topology and biological function dimensions. Cosine similarity and weighted linear combination algorithms are used to optimize the mapping weights of TCM efficacy terms, enabling statistical modeling of multi-source heterogeneous data. Finally, in vitro cell models are used to verify target regulation efficiency and correct network weights. The clinical syndrome integral change rate is used to feed back into the efficacy vector calibration coefficient, improving the accuracy of efficacy quantification results and meeting the modern needs of clinical precision medicine and new drug development. This solves the problems of single data dimensions, static models, and insufficient accuracy of quantification results in existing technologies.

[0104] Figure 9 A schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device may include: The memory 901, the processor 902, and the computer program stored on the memory 901 and capable of running on the processor 902.

[0105] When the processor 902 executes the program, it implements a method for quantifying the efficacy of traditional Chinese medicine prescriptions provided in the above embodiments.

[0106] Furthermore, electronic devices also include: Communication interface 903 is used for communication between memory 901 and processor 902.

[0107] The memory 901 is used to store computer programs that can run on the processor 902.

[0108] The memory 901 may include high-speed RAM (Random Access Memory) and may also include non-volatile memory, such as at least one disk storage device.

[0109] If the memory 901, processor 902, and communication interface 903 are implemented independently, then the communication interface 903, memory 901, and processor 902 can be interconnected via a bus to complete communication between them. The bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 9 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0110] Optionally, in a specific implementation, if the memory 901, processor 902, and communication interface 903 are integrated on a single chip, then the memory 901, processor 902, and communication interface 903 can communicate with each other through an internal interface.

[0111] The processor 902 may be a CPU (Central Processing Unit), an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of this application.

[0112] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for quantifying the efficacy of traditional Chinese medicine prescriptions.

[0113] In the description of this specification, the references to "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0114] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0115] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0116] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any of the following techniques known in the art, or a combination thereof: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0117] Those skilled in the art will understand that all or part of the steps of the methods described in the above embodiments can be implemented by a program instructing related hardware, and the program can be stored in a computer-readable storage medium. When executed, the program includes one or a combination of the steps of the method embodiments.

[0118] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for quantifying the efficacy of a traditional Chinese medicine formula, characterized in that, include: To obtain the chemical composition spectrum, content of active ingredients, and bioactivity indicators of the Chinese herbs that make up the prescription; Based on the chemical composition spectrum, pharmacodynamic substance content, and bioactivity indicators, a multi-target regulatory network for the effects of traditional Chinese medicine (TCM) formulas is constructed using a TCM-target interaction database. This includes: constructing a standardized basic data layer; directly matching known component-target mapping relationships in the database based on the standardized basic data layer; inferring missing associations based on molecular structure and activity data; supplementing unrecorded information by mining literature; verifying the accuracy of unrecorded component-target mapping relationships using component content and bioactivity data; determining the level of evidence based on the component-target mapping relationship and its source; and then, based on the level of evidence... Combining the component content and bioactivity, a linear weighted model is used to calculate mapping weights. By verifying reliability and removing outliers, a weighting system is established. Based on this weighting system, chemical components and targets are defined as network nodes, and mapping weights are defined as edges, constructing a binary network for single-herb Chinese medicine. Multiple binary networks for single-herb Chinese medicine are merged into a multi-target regulatory network for the prescription by accumulating or averaging the weights of the same components and edges. The Louvain algorithm is used to identify core efficacy pathway clusters, and a pathway activation intensity value is calculated using a pathway target membership weighted model. The formula for the pathway target membership weighted model is: in, For pathway The activation strength value; For pathways Membership degree; Target The activity weight; target Activation direction; path The total number of targets included in the calculation; Target The original activity value; It is the minimum value among all the original values ​​of the target points; The maximum value among all original target values; This represents the upper limit of the summation; For index variables; Target With indicators The correlation between them; Based on the core efficacy pathway clusters, a mapping relationship between the core efficacy pathway clusters and TCM efficacy terms is established through ontology mapping and semantic mapping. Based on the pathway activation intensity values, a weighted aggregation model is used to generate a formula efficacy intensity vector. This includes: constructing a basic ontology library; obtaining the initial mapping relationship and weights between the core efficacy pathway clusters and TCM efficacy terms through structured rule matching based on the basic ontology library; calculating semantic similarity using a cosine similarity algorithm based on the initial mapping relationship and weights, and simultaneously optimizing and correcting the mapping weights using a weighted linear combination algorithm; and calculating the intensity value of each efficacy in the formula using a weighted aggregation model based on the corrected mapping weights and the pathway activation intensity values, thereby generating a formula efficacy intensity vector. The formula for the weighted aggregation model is: in, For the first The original activation intensity of each pathway; This represents the standardized value after processing; It is the minimum value of the activation values ​​of all pathways; The maximum value of all pathway activation values; For the first The intensity of each efficacy dimension; For the first The weight of each path; The number of pathways participating in the calculation of this dimension; This represents the vector of efficacy intensity of the prescription. The total number of efficacy dimensions; Based on the efficacy intensity vector of the formula, the target regulation efficiency is verified in an in vitro cell model to correct the network weights. The efficacy vector calibration coefficient is calculated using the rate of change of clinical syndrome integrals to generate a quantitative estimate of the formula's efficacy. The efficacy vector calibration coefficient is a coefficient used to calibrate the intensity values ​​of each efficacy dimension in the efficacy intensity vector. The quantitative estimate of the formula's efficacy is the quantified efficacy result after calibration with clinical data, composed of the calibrated intensity values ​​of each TCM efficacy dimension. The calibration formula is as follows: in, For the calibrated first The intensity of each efficacy dimension; No. Calibration coefficients in each dimension.

2. The method for quantifying the efficacy of traditional Chinese medicine prescriptions according to claim 1, characterized in that, The Louvain algorithm was used to identify core functional pathway clusters, including: Construct the Louvain algorithm; Based on the Louvain algorithm, the multi-target regulation network is divided into modules, and high-confidence modules are selected as candidate core pathway clusters by combining module degree and internal density indicators. Based on the candidate core pathway clusters, enrichment analysis was performed on the gene ontology and KEGG pathway of the target sites in each pathway cluster, and the core functional pathway clusters were identified by combining the network topology index screening method.

3. The method for quantifying the efficacy of traditional Chinese medicine prescriptions according to claim 1, characterized in that, The Louvain algorithm formula is as follows: in, Modularity; This represents the sum of the weights of all edges in the network. , For any two target points in the network; Target With target Edge weights between them; Target The sum of the weights of all connected edges; Target The sum of the weights of all connected edges; Target The module tag it belongs to; Target The module tag it belongs to; This is an indicator function.

4. A quantitative system for the efficacy of traditional Chinese medicine prescriptions, characterized in that, include: The acquisition module is used to acquire the chemical composition spectrum, pharmacodynamic substance content, and bioactivity indicators of the Chinese herbs that make up the prescription. The construction module is used to construct a multi-target regulatory network of prescription effects based on the chemical composition spectrum, pharmacodynamic substance content, and bioactivity indicators, combined with a traditional Chinese medicine-target interaction database. This includes: constructing a standardized basic data layer; directly matching known component-target mapping relationships in the database based on the standardized basic data layer; inferring missing associations based on molecular structure and activity data; supplementing unrecorded information by mining literature; verifying the accuracy of unrecorded component-target mapping relationships by combining component content and bioactivity data; determining the level of evidence based on the component-target mapping relationship and its source; and further... The evidence level, combined with the component content and bioactivity, is calculated using a linear weighted model to determine the mapping weights. A weighting system is established by verifying credibility and removing outliers. Based on this weighting system, chemical components and targets are defined as network nodes, and mapping weights are defined as edges. A binary network of single-herb Chinese medicine is constructed. Multiple binary networks of single-herb Chinese medicine are merged into a multi-target regulatory network of the prescription by accumulating or averaging the weights of the same components and edges. A modular algorithm is used to identify core efficacy pathway clusters, and the pathway activation intensity value is calculated using a pathway target membership weighted model. The formula for the pathway target membership weighted model is: in, For pathway The activation strength value; For pathways Membership degree; Target The activity weight; target Activation direction; path The total number of targets included in the calculation; Target The original activity value; It is the minimum value among all the original values ​​of the target points; The maximum value among all original target values; This represents the upper limit of the summation; For index variables; Target With indicators The correlation between them; The association module is used to establish a mapping relationship between the core efficacy pathway clusters and TCM efficacy terms through ontology mapping and semantic mapping, and to generate a formula efficacy intensity vector based on the pathway activation intensity value through a weighted aggregation model. This includes: constructing a basic ontology library; obtaining the initial mapping relationship and weights between the core efficacy pathway clusters and TCM efficacy terms through structured rule matching based on the basic ontology library; calculating semantic similarity using a cosine similarity algorithm based on the initial mapping relationship and weights, and simultaneously optimizing and correcting the mapping weights using a weighted linear combination algorithm; and calculating the intensity value of each efficacy in the formula using a weighted aggregation model based on the corrected mapping weights and the pathway activation intensity value, thereby generating a formula efficacy intensity vector. The formula for the weighted aggregation model is: in, For the first The original activation intensity of each pathway; This represents the standardized value after processing; It is the minimum value of the activation values ​​of all pathways; The maximum value of all pathway activation values; For the first The intensity of each efficacy dimension; For the first The weight of each path; The number of pathways participating in the calculation of this dimension; This represents the vector of efficacy intensity of the prescription. The total number of efficacy dimensions; The verification module is used to verify the target regulation efficiency in an in vitro cell model based on the efficacy intensity vector of the prescription to correct the network weights, and to calculate the efficacy vector calibration coefficient through the change rate of clinical syndrome integrals to generate the quantitative efficacy of the prescription. The efficacy vector calibration coefficient is a coefficient used to calibrate the intensity values ​​of each efficacy dimension in the efficacy intensity vector of the prescription. The quantitative efficacy of the prescription is the efficacy quantification result after calibration with clinical data, consisting of the calibrated intensity values ​​of each TCM efficacy dimension. The calibration formula is as follows: in, For the calibrated first The intensity of each efficacy dimension; No. Calibration coefficients in each dimension.

5. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the program to implement a method for quantifying the efficacy of a traditional Chinese medicine prescription as described in any one of claims 1-3.

6. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed, they implement the method for quantifying the efficacy of a traditional Chinese medicine prescription as described in any one of claims 1-3.

Images