An integrated intelligent agent system for new traditional Chinese medicine drug discovery and registration

Abstract

This invention relates to the field of traditional Chinese medicine (TCM) new drug discovery and registration technology, and discloses an integrated intelligent system for TCM new drug discovery and registration, comprising: a data processing unit acquiring multi-source TCM data and constructing a multi-dimensional data association model; a new drug evaluation unit inputting TCM new drug characteristic description information into the multi-dimensional data association model to generate several candidate TCM new drug schemes; performing joint analysis and multi-dimensional evaluation to generate pharmacodynamic evaluation results; a registration application unit extracting evaluation indicators based on a registration application data template, performing data mapping, and generating registration application materials; and a collaborative optimization unit modularly organizing the registration application materials, generating an application material evaluation report; and feeding the report back to the multi-dimensional data association model for dynamic updates. This application improves the success rate and scientific rigor of TCM new drug development and ensures the completeness and compliance of application materials.

Description

Technical Field

[0001] This invention relates to the field of new drug discovery and registration technology for traditional Chinese medicine, and more specifically, to an integrated intelligent system for the discovery and registration of new traditional Chinese medicine. Background Technology

[0002] Traditional Chinese medicine (TCM), as an important component of traditional medicine, contains a wealth of empirical prescriptions and complex chemical component systems. In recent years, with the development of omics technology, computational pharmacology, and artificial intelligence, some progress has been made in the discovery of new TCM drugs based on multi-source data. At the same time, drug registration and supervision have increasingly stringent requirements for safety, efficacy, and data standardization, leading to a growing demand for the modernization, standardization, and industrialization of TCM.

[0003] In existing technologies, the development of new traditional Chinese medicine (TCM) drugs typically faces the following shortcomings: First, the sources of TCM and pharmaceutical data involved are extensive and inconsistent in format, including ancient Chinese medicine texts, clinical medical records, chemical composition databases, in vitro and in vivo efficacy / toxicology test data, high-throughput omics data such as genomics / proteomics, as well as literature and patent information. Data heterogeneity and terminology non-standardization make information integration and knowledge extraction difficult. Second, TCM is a multi-component, multi-target, and multi-pathway system, with complex relationships between medicinal materials, components, targets, and new drugs. Existing methods struggle to efficiently and accurately extract structured knowledge from historical TCM for new drug design and to optimize components and assess safety. Third, the efficacy, toxicity, and clinical applicability assessment of new drug candidates often relies on extensive experiments and human experience. The assessment process is time-consuming, labor-intensive, and highly subjective, affecting R&D efficiency and repeatability. Furthermore, drug registration application documents are numerous and have strict template and format standards. The generation, field matching, and format standardization of registration data heavily rely on manual processing, which can easily lead to data inconsistencies, logical breaks, or non-compliant formats, thus prolonging the application cycle and increasing compliance risks.

[0004] Therefore, it is necessary to design an integrated intelligent agent system for the discovery and registration of new traditional Chinese medicine drugs to solve the problems existing in the current technology. Summary of the Invention

[0005] In view of this, the present invention proposes an integrated intelligent agent system for the discovery and registration of new traditional Chinese medicine drugs, aiming to solve the problems of time-consuming and labor-intensive evaluation process of new drug schemes, strong subjectivity, and impact on R&D efficiency.

[0006] This invention proposes an integrated intelligent agent system for the discovery and registration of new traditional Chinese medicine drugs, comprising:

[0007] The data processing unit is configured to acquire multi-source traditional Chinese medicine data and perform data preprocessing; and to construct a multi-dimensional data association model based on the preprocessed multi-source traditional Chinese medicine data.

[0008] The new drug evaluation unit is configured to input the characteristic description information of new traditional Chinese medicine drugs into the multi-dimensional data association model to generate several candidate new traditional Chinese medicine drug schemes; to perform joint analysis on several candidate new traditional Chinese medicine drug schemes to screen out candidate new traditional Chinese medicine drug schemes that meet the compatibility and safety conditions; and to perform multi-dimensional evaluation on the screened candidate new traditional Chinese medicine drug schemes based on efficacy indicators, toxicity indicators and clinical indicators to generate drug property evaluation results.

[0009] The registration application unit is configured to extract the evaluation indicators corresponding to the registration requirements from the drug efficacy assessment results based on the registration application data template, perform data mapping on the key evaluation indicators through field matching, and perform structured arrangement, format standardization and automatic module filling on the data mapping results to generate registration application materials.

[0010] The collaborative optimization unit is configured to modularly organize the registration application materials, generate an application material evaluation report through data consistency checks and logical integrity verification, and feed the application material evaluation report back to the multi-dimensional data association model for dynamic updates.

[0011] Furthermore, when the data processing unit constructs a multi-dimensional data association model based on the preprocessed multi-source TCM data, it includes:

[0012] The data processing unit acquires medicinal material attribute information based on the multi-source TCM data; determines the active ingredients contained in the medicinal materials, constructs a correspondence between medicinal materials and ingredients; determines the target interaction relationship between the active ingredients and biological targets; extracts the correlation information between diseases and the efficacy of new TCM drugs, and generates new drug correlation information.

[0013] The data processing unit organizes the medicinal material attribute information, medicinal material-component correspondence, target action relationship and new drug association information according to hierarchical relationship, and establishes bidirectional association mapping between adjacent levels to form the multi-dimensional data association model.

[0014] Furthermore, when the new drug evaluation unit inputs the characteristic description information of traditional Chinese medicine new drugs into the multi-dimensional data association model to generate several candidate traditional Chinese medicine new drug schemes, it includes:

[0015] The new drug evaluation unit acquires new drug feature description information and matches new drug association information corresponding to the new drug feature description information in the multi-dimensional data association model;

[0016] The new drug evaluation unit determines a set of historical Chinese medicines corresponding to the new drug feature description information; extracts basic Chinese medicine structures from the set of historical Chinese medicines; optimizes the basic Chinese medicine structures according to the multi-dimensional data association model to generate the candidate Chinese medicine new drug scheme.

[0017] Furthermore, when the new drug evaluation unit performs joint analysis on several candidate traditional Chinese medicine (TCM) new drug schemes and screens TCM new drug candidate schemes that meet the compatibility and safety conditions, it includes:

[0018] The joint analysis includes assessment of the structural rationality of new traditional Chinese medicine drugs, assessment of the composition and concentration of active ingredients, prediction of multi-target regulatory networks, and screening of drug incompatibilities.

[0019] Furthermore, the new drug evaluation unit performs multi-dimensional evaluations of the selected traditional Chinese medicine new drug candidates and generates drug efficacy evaluation results, including:

[0020] The new drug evaluation unit determines the efficacy intensity based on the concentration-effect relationship of the active ingredients in the candidate traditional Chinese medicine drug scheme; determines the toxicity risk based on the probability of adverse events occurring in the candidate traditional Chinese medicine drug scheme; and determines the clinical applicability based on the degree of matching between the candidate traditional Chinese medicine drug scheme and the target indication.

[0021] Furthermore, when the new drug evaluation unit performs multi-dimensional evaluations on the selected traditional Chinese medicine new drug candidates and generates the drug efficacy evaluation results, it also includes:

[0022] The new drug evaluation unit comprehensively scores the efficacy, toxicity risk, and clinical applicability of the drug according to preset weighting coefficients, and generates the drug efficacy evaluation result.

[0023] Furthermore, when the registration application unit generates registration application materials, it includes:

[0024] The registration application unit extracts evaluation indicator data from the drug efficacy assessment results;

[0025] The registration and application unit establishes a mapping relationship between the evaluation indicator data and the registration and application data template, fills the evaluation indicator data into the corresponding positions, and performs structured arrangement and format standardization processing on the filled content to generate the registration and application materials.

[0026] Furthermore, when the collaborative optimization unit generates the application data evaluation report, it includes:

[0027] The collaborative optimization unit divides the registration application materials into pharmaceutical research, pharmacological and toxicological research, and clinical research sections.

[0028] Check the consistency of data across different research sections; verify the logical coherence between different research sections; check whether the registration application materials meet the prescribed requirements and format specifications; and generate an evaluation report for the application materials.

[0029] Furthermore, when the collaborative optimization unit feeds back the application material evaluation report to the multi-dimensional data association model for dynamic updates, it includes:

[0030] The collaborative optimization unit extracts verified and effective medicinal material-component correspondence data, new evidence of target-action relationships, and new drug association information optimization points from the application material evaluation report;

[0031] The medicinal material-component correspondence data is updated to the medicinal material-component correspondence in the multi-dimensional data association model; the new evidence of target action relationship is updated to the target action relationship in the multi-dimensional data association model; the optimization points of new drug association information are updated to the new drug association information in the multi-dimensional data association model.

[0032] Based on the updated medicinal material attribute information, medicinal material-component correspondence, target action relationship and new drug association information, the bidirectional association mapping between adjacent levels in the multi-dimensional data association model is adjusted.

[0033] Furthermore, when acquiring multi-source traditional Chinese medicine data and performing data preprocessing, the following steps are included:

[0034] The data preprocessing includes text structure parsing, terminology standardization, format normalization, filling in missing key information, removing erroneous data, and establishing relationships.

[0035] Compared with existing technologies, the beneficial effects of this invention are as follows: By constructing an integrated intelligent system for the discovery and registration of new traditional Chinese medicine (TCM) drugs, it achieves intelligent management of the entire process, from multi-source TCM data collection, candidate TCM drug generation, pharmacodynamic evaluation to automated generation and optimization of registration application materials. Utilizing multi-dimensional data association models and joint analysis methods, it screens TCM drug candidates with significant efficacy and safe compatibility, improving the success rate and scientific rigor of new drug development. Through comprehensive evaluation of multiple indicators, it can fully grasp the efficacy, toxicity, and clinical characteristics of candidate drugs, reducing R&D risks. The registration application unit automatically completes data mapping, structured arrangement, and format standardization operations, reducing manual processing and filling workload, and improving registration efficiency and accuracy. The collaborative optimization unit ensures the completeness and compliance of application materials through data consistency checks and logical integrity verification, and can use feedback information to dynamically optimize the data association model, realizing intelligent closed-loop management of R&D-application. Attached Figure Description

[0036] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0037] Figure 1 This is a functional block diagram of an integrated intelligent agent system for the discovery and registration of new traditional Chinese medicines, provided in an embodiment of the present invention. Detailed Implementation

[0038] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0039] In some embodiments of this application, see Figure 1 As shown, this application proposes an integrated intelligent agent system for the discovery and registration of new traditional Chinese medicine drugs, comprising:

[0040] The data processing unit is configured to acquire multi-source TCM data and perform data preprocessing; and to construct a multi-dimensional data association model based on the preprocessed multi-source TCM data.

[0041] The new drug evaluation unit is configured to input the characteristic description information of new traditional Chinese medicines into a multi-dimensional data association model to generate several candidate new traditional Chinese medicine schemes; to perform joint analysis on several candidate new traditional Chinese medicine schemes to screen out candidate new traditional Chinese medicine schemes that meet the compatibility and safety conditions; and to perform multi-dimensional evaluation on the screened candidate new traditional Chinese medicine schemes based on efficacy indicators, toxicity indicators, and clinical indicators to generate drug property evaluation results.

[0042] The registration application unit is configured to extract the evaluation indicators corresponding to the registration requirements from the drug efficacy assessment results based on the registration application data template, perform data mapping on the key evaluation indicators through field matching, and perform structured arrangement, format standardization and automatic module filling on the data mapping results to generate registration application materials.

[0043] The collaborative optimization unit is configured to modularly organize the registration application materials, generate an application material evaluation report through data consistency checks and logical integrity verification, and feed the application material evaluation report back to the multi-dimensional data association model for dynamic updates.

[0044] Specifically, the data processing unit is responsible for acquiring multi-source TCM data. This multi-source TCM data refers to TCM information from different sources, types, and formats, including but not limited to: information on Chinese medicinal herbs in ancient texts, modern databases of chemical components of Chinese medicinal herbs, clinical case data, in vitro and in vivo efficacy and toxicology experimental data, omics data such as genomics or proteomics, as well as scientific research literature and patent information. The data processing unit performs data preprocessing on this data, including text structuring parsing (converting unstructured text into computable structured data), terminology standardization (unifying TCM terminology and medicinal material names), format standardization (unifying data formats), missing data imputation (reasonably estimating or supplementing missing information), outlier removal (removing erroneous or unreasonable data), and relationship establishment (constructing preliminary mapping relationships between medicinal materials, components, targets, and new drugs). After preprocessing, the data processing unit constructs a multi-dimensional data association model based on the organized multi-source TCM data. This model hierarchically organizes the information on medicinal material properties, active ingredients, target interactions, and the correlation between diseases and the efficacy of new TCM drugs, and establishes bidirectional association mappings between adjacent levels to support the generation and evaluation of candidate new TCM drugs.

[0045] The new drug evaluation unit is responsible for inputting the characteristic description information of new traditional Chinese medicines (TCMs) into the aforementioned multi-dimensional data association model to generate several candidate TCM new drug schemes. The characteristic description information refers to a structured description reflecting disease manifestations and patient constitution characteristics, including clinical symptoms, tongue and pulse manifestations, and laboratory indicators. Based on model matching of historical TCMs and their corresponding new drugs, the new drug evaluation unit extracts the basic TCM structure and optimizes the basic TCM new drug scheme by combining active ingredients, medicinal material compatibility rules, and multi-target action networks, thereby generating candidate TCM new drug schemes that meet efficacy thresholds and compatibility safety conditions. The joint analysis includes the structural rationality assessment of new TCM drugs (whether they conform to traditional TCM compatibility principles), assessment of active ingredient composition and concentration, prediction of multi-target regulatory networks (predicting the synergistic effect of drugs acting on multiple biological targets), and screening for incompatibilities (avoiding combinations of potentially interacting components). The new drug evaluation unit conducts multi-dimensional assessments of the selected candidate drugs, including: determining the efficacy based on the concentration-effect relationship of the active ingredient (measuring the physiological activity of the candidate drug), determining the toxicity risk based on the probability of adverse events (measuring safety), and determining the clinical suitability based on the degree of matching between the candidate drug and the target indication. The evaluation results can be comprehensively scored using preset weighted coefficients to form a comprehensive drug efficacy assessment result.

[0046] The registration application unit is based on the national or regional drug registration application data template. It automatically extracts the evaluation indicators (such as efficacy, toxicity, and clinical applicability indicators) that correspond to the registration requirements from the drug efficacy assessment results. Through field matching technology, it maps these indicators to the corresponding positions in the template, performs structured arrangement (ensuring clear data hierarchy), standardizes the format (meets the format requirements of regulatory authorities), and automatically fills in the modules, thereby generating complete and standardized registration application materials.

[0047] The collaborative optimization unit modularizes the generated registration application materials, dividing them into pharmaceutical research, pharmacological and toxicological research, and clinical research sections. It then performs data consistency checks and logical integrity verification, generating an evaluation report. This report is further fed back to a multi-dimensional data association model to dynamically update the correspondence between medicinal materials and components, new evidence of target effects, and new drug association information. This achieves closed-loop optimization of the R&D-evaluation-application process, enabling the system to continuously improve the accuracy of candidate drug screening and the standardization of registration materials.

[0048] Understandably, by integrating multi-source TCM data collection, preprocessing, multi-dimensional correlation modeling, candidate TCM new drug generation and joint evaluation, multi-indicator pharmacodynamics assessment, automatic generation of registration application materials, and collaborative optimization into a single system, intelligent management of the entire TCM new drug R&D and registration process has been achieved. It can automatically integrate diverse TCM and pharmaceutical data in various formats, and by constructing a multi-dimensional data correlation model, it realizes a hierarchical mapping relationship between medicinal materials, active ingredients, targets, and new drugs, providing a scientific basis for the generation of candidate TCM new drugs. Using joint analysis and multi-dimensional pharmacodynamics assessment methods, it screens TCM new drug schemes with significant efficacy and safe compatibility, improving the success rate and scientific rigor of R&D. Through field matching, structured arrangement, and format standardization processing in the registration application unit, it can quickly generate standardized application materials that meet registration requirements, reducing the workload of manual processing and the risk of errors. The collaborative optimization unit performs consistency checks and logical verification on the application materials and feeds the evaluation results back to the data correlation model, achieving dynamic optimization and forming a closed-loop intelligent management system for the new drug R&D and registration process.

[0049] In some embodiments of this application, when the data processing unit constructs a multi-dimensional data association model based on the preprocessed multi-source traditional Chinese medicine data, it includes:

[0050] The data processing unit acquires medicinal material attribute information based on multi-source TCM data; determines the active ingredients contained in the medicinal materials, constructs the correspondence between medicinal materials and ingredients; determines the target interaction relationship between active ingredients and biological targets; extracts the correlation information between diseases and the efficacy of new TCM drugs, and generates new drug correlation information.

[0051] The data processing unit organizes medicinal material attribute information, medicinal material-component correspondence, target action relationship and new drug association information according to hierarchical relationship, and establishes bidirectional association mapping between adjacent levels to form a multi-dimensional data association model.

[0052] Specifically, the data processing unit first acquires medicinal material attribute information based on multi-source TCM data. This attribute information comes from pharmacopoeia standards and TCM databases, and is used to characterize the basic attributes of medicinal materials, such as their properties, meridian tropism, efficacy categories, source plants, and parts used. Next, the data processing unit identifies the active ingredients contained in each medicinal material. Active ingredients are chemical substances that can produce definite biological effects in vivo. Their information comes from the analysis results of TCM chemical components, experimental research data, or literature mining results. Based on this, a medicinal material-component correspondence is constructed to describe the mapping relationship between the medicinal material and its contained active ingredients. Finally, based on pharmacological experimental data, molecular docking prediction results, or target validation studies, the data processing unit determines the target-target interaction relationship between active ingredients and biological targets. Biological targets refer to molecular objects related to the occurrence, development, and regulation of diseases, including proteins, enzymes, receptors, or genes. The target-target interaction relationship is used to characterize the regulatory mechanisms and intensity of active ingredients on corresponding biological targets. Simultaneously, the data processing unit extracts correlation information between diseases and the efficacy of new Chinese medicines from TCM clinical research data and historical TCM literature. Diseases are a comprehensive description of the pathogenesis and clinical manifestations of a disease, while the efficacy of new Chinese medicines reflects the therapeutic effect of a specific new Chinese medicine under the corresponding new drug. This information is structured to form new drug correlation information. After completing the above information acquisition and construction, the data processing unit organizes the medicinal material attribute information, medicinal material-component correspondence, target action relationship, and new drug correlation information according to the hierarchical relationship of "medicinal material—component—target—new drug," and establishes bidirectional correlation mappings between adjacent levels. This allows the model to support both forward reasoning analysis starting from medicinal materials and reverse tracing analysis starting from new drugs or targets, thus forming a multi-dimensional data correlation model.

[0053] Understandably, by using a data processing unit to obtain medicinal material attribute information, construct the correspondence between medicinal materials and active ingredients, clarify the target interaction relationship between active ingredients and biological targets, and extract the correlation information between diseases and the efficacy of new Chinese medicine drugs, a hierarchical multi-dimensional data association model is formed. This allows the complex relationship between "medicinal materials - ingredients - targets - new drugs" in Chinese medicine to be expressed systematically and structurally. By establishing bidirectional association mapping between adjacent levels, it can not only support forward reasoning analysis starting from medicinal materials or ingredients, but also realize reverse tracing from new drugs or efficacy results back to specific medicinal materials and ingredients, improving the flexibility and accuracy of model reasoning.

[0054] In some embodiments of this application, when the new drug evaluation unit inputs the characteristic description information of traditional Chinese medicine new drugs into a multi-dimensional data association model to generate several candidate traditional Chinese medicine new drug schemes, it includes:

[0055] The new drug evaluation unit acquires new drug feature description information and matches new drug association information corresponding to the new drug feature description information in a multi-dimensional data association model;

[0056] The new drug evaluation unit identifies a set of historical Chinese medicines corresponding to the new drug feature description information; extracts basic Chinese medicine structures from the historical Chinese medicine set; optimizes the basic Chinese medicine structures based on a multi-dimensional data association model to generate candidate Chinese medicine new drug schemes.

[0057] Specifically, new drug characteristic description information refers to a data set that structurally expresses the pathogenesis and clinical manifestations of a target disease within the framework of Traditional Chinese Medicine (TCM) theory. This information originates from TCM clinical records, guidelines, consensus statements, or expert annotations. New drug characteristic description information includes, but is not limited to, primary symptoms, secondary symptoms, tongue appearance, pulse characteristics, and related laboratory indicators. The new drug evaluation unit first acquires the new drug characteristic description information and inputs it into the aforementioned multi-dimensional data association model. Within this model, it matches the new drug association information with pre-constructed new drug association information. This association information is structured data formed by the correspondence between diseases and historical TCM efficacy, reflecting the degree of association between different new drugs and effective TCM new drugs. Based on the matching results, the new drug evaluation unit determines the historical TCM set corresponding to the new drug characteristic description information. This historical TCM set refers to a set of TCM new drugs that have been proven to have therapeutic effects against the same or similar new drugs in ancient TCM literature, modern clinical research, or guidelines. Subsequently, the new drug evaluation unit extracts basic TCM structures from the historical TCM set. These basic TCM structures characterize the core medicinal material composition and compatibility relationships of the new TCM drugs, reflecting the basic treatment strategies of traditional TCM new drugs. The new drug evaluation unit optimizes the basic structure of traditional Chinese medicine based on the correlation between Chinese medicinal materials, components, targets, and new drugs using a multi-dimensional data association model. This involves adjusting and optimizing the types of medicinal materials, the composition of active ingredients, or the proportion of drugs used, while maintaining the original principles of compatibility and treatment direction of new Chinese medicine drugs. This aims to enhance efficacy, reduce potential risks, or improve targeting, thereby generating candidate new Chinese medicine drug solutions that meet the requirements of new drug development.

[0058] Understandably, by inputting the characteristic description information of new traditional Chinese medicines (TCMs) into a multi-dimensional data association model through a new drug evaluation unit, and matching the corresponding new drug association information in the model, the accurate retrieval of effective TCM knowledge based on real clinical new drugs can be achieved. By identifying a set of historical TCMs corresponding to the characteristic description information of new drugs and extracting basic TCM structures from them, the generation of candidate solutions is based on existing clinical experience and efficacy verification, reducing the blindness of new drug design. Combining the multi-dimensional data association model to optimize the components of basic TCM structures allows for systematic adjustments to efficacy, target effects, and safety while adhering to traditional compatibility principles, thereby generating more targeted and scientific candidate TCM new drug solutions.

[0059] In some embodiments of this application, when the new drug evaluation unit performs joint analysis on several candidate traditional Chinese medicine (TCM) new drug schemes and screens TCM new drug candidate schemes that meet the compatibility and safety conditions, it includes:

[0060] The combined analysis includes assessment of the structural rationality of new traditional Chinese medicine drugs, assessment of the composition and concentration of active ingredients, prediction of multi-target regulatory networks, and screening of drug incompatibilities.

[0061] Specifically, when screening several candidate traditional Chinese medicine (TCM) new drug schemes, the new drug evaluation unit performs conjoint analysis to determine TCM new drug candidate schemes that meet preset efficacy thresholds and compatibility safety conditions. The conjoint analysis first includes an assessment of the structural rationality of the TCM new drug. This assessment determines whether the overall compatibility structure of the candidate TCM new drug conforms to the "principal, assistant, adjuvant, and guide" principles of TCM theory and common formulation rules. The assessment is based on classical TCM theories, historical TCM data, and an expert rule base. Secondly, the conjoint analysis includes an assessment of the composition and concentration of active ingredients. Active ingredients refer to chemical substances that can produce definite biological effects in vivo. The composition and concentration assessment analyzes the types and relative or absolute contents of active ingredients in each candidate TCM new drug. This assessment is based on a TCM chemical composition database, experimental measurement data, or predictive model results, and is used to determine whether the candidate TCM new drug can achieve the expected efficacy level and meet the efficacy threshold requirements. The joint analysis also includes multi-target regulatory network prediction, where the multi-target regulatory network refers to the interaction network formed between the active ingredient and multiple biological targets. These biological targets are molecular objects related to the occurrence and development of diseases. This prediction is achieved through network pharmacology models, molecular docking algorithms, or validated target data, and is used to evaluate the synergistic regulatory capabilities and potential therapeutic effects of candidate traditional Chinese medicine (TCM) drugs at the system level. The joint analysis also includes incompatibility screening. Incompatibility refers to combinations of medicinal materials or components that may cause adverse reactions or reduce efficacy during the combination of TCMs. This screening is based on traditional TCM incompatibility rules, modern toxicology research results, and adverse reaction databases, and is used to identify potential safety hazards in candidate TCM drugs. Through the above multi-dimensional joint analysis, the new drug evaluation unit can comprehensively screen candidate TCM drug regimens from multiple aspects such as efficacy, rationality of action mechanism, and drug safety, thereby obtaining TCM drug candidate regimens that meet the efficacy threshold and compatibility safety conditions.

[0062] Understandably, by conducting a combined analysis on several candidate TCM new drug schemes, including the evaluation of the rationality of TCM new drug structure, the evaluation of the composition and concentration of active ingredients, the prediction of multi-target regulatory networks, and the screening of incompatibilities, the new drug evaluation unit can comprehensively judge the candidate TCM new drugs from multiple dimensions such as the rationality of TCM new drug compatibility, the material basis of pharmacodynamics, the synergistic effect of the mechanism of action, and the safety of medication. This avoids the bias caused by relying on a single evaluation indicator, helps to accurately identify TCM new drug candidate schemes that can simultaneously meet the pharmacodynamic threshold and compatibility safety conditions, and improves the scientificity and reliability of the screening results.

[0063] In some embodiments of this application, the new drug evaluation unit performs multi-dimensional evaluations of the selected traditional Chinese medicine new drug candidates and generates pharmacodynamic evaluation results, including:

[0064] The new drug evaluation unit determines the efficacy intensity based on the concentration-effect relationship of the active ingredients in the candidate traditional Chinese medicine (TCM) new drug regimens; determines the toxicity risk based on the probability of adverse events occurring in the candidate TCM new drug regimens; and determines the clinical applicability based on the degree of matching between the candidate TCM new drug regimens and the target indication.

[0065] Specifically, after completing the initial screening of candidate traditional Chinese medicine (TCM) new drug schemes, the new drug evaluation unit conducts multi-dimensional evaluations of the selected TCM new drug candidates to generate pharmacodynamic evaluation results. Multi-dimensional evaluation refers to a comprehensive analysis of the efficacy, safety, and clinical applicability of candidate schemes from different evaluation perspectives, used to comprehensively characterize the overall pharmacodynamic characteristics of the candidate schemes. The new drug evaluation unit first determines the efficacy intensity based on the concentration-effect relationship of each active ingredient in the TCM new drug candidate scheme. Active ingredients refer to chemical substances that can produce definite biological effects in vivo. The concentration-effect relationship refers to the functional relationship between the strength of the pharmacological effect produced by the active ingredient at different concentration levels. This relationship is derived from pharmacological experimental data, literature reports, or model prediction results, and is used to quantify the therapeutic effect of the candidate scheme. The new drug evaluation unit determines the toxicity risk based on the probability of adverse events occurring in the TCM new drug candidate scheme. Adverse events refer to reactions that may adversely affect the body during medication. Their probability is derived from toxicological experimental data, clinical research reports, adverse reaction databases, or risk prediction models based on component characteristics, used to assess the safety level of the candidate scheme. The new drug evaluation unit determines the clinical applicability based on the degree of matching between the candidate new drug regimen and the target indication. The target indication refers to the disease or pathological state that the candidate drug is intended to treat. The degree of matching is determined by analyzing the consistency between the target, regulatory pathway and new drug characteristics of the candidate regimen and the pathological mechanism and clinical characteristics of the target indication. The data comes from disease mechanism research, clinical guidelines or historical drug use experience.

[0066] Understandably, by comprehensively evaluating candidate traditional Chinese medicine drugs from three dimensions—efficacy intensity, toxicity risk, and clinical applicability—the new drug evaluation unit can comprehensively and objectively characterize the pharmacological characteristics of the candidate drugs. Determining efficacy intensity based on the concentration-effect relationship of active ingredients helps to accurately reflect the potential therapeutic effects of candidate drugs, avoiding judgments of efficacy based solely on experience. Determining toxicity risk by analyzing the probability of adverse events can identify safety hazards in the early stages of research and development, reducing the risk of failure in subsequent trial phases. Assessing clinical applicability by combining the degree of matching between candidate drugs and target indications helps to improve the targeting of drug development and its clinical translational value.

[0067] In some embodiments of this application, when the new drug evaluation unit performs multi-dimensional evaluation on the selected traditional Chinese medicine new drug candidates and generates the drug efficacy evaluation results, it also includes:

[0068] The new drug evaluation unit comprehensively scores the efficacy, toxicity risk, and clinical applicability according to preset weighting coefficients, and generates drug efficacy evaluation results.

[0069] Specifically, after assessing the efficacy, toxicity, and clinical applicability of candidate traditional Chinese medicine (TCM) drugs, the new drug evaluation unit performs a comprehensive score based on preset weighting coefficients to generate the final drug efficacy evaluation result. The preset weighting coefficients are weight parameters set for different evaluation dimensions, reflecting the relative importance of each dimension in the overall drug efficacy judgment. Their values ​​are derived from historical new drug development data statistical analysis, expert experience rules, or regulatory requirements, and can be adjusted according to different disease types or development stages. The comprehensive score refers to the process of integrating multiple evaluation indicators such as efficacy, toxicity, and clinical applicability into one or more comprehensive evaluation indicators through weighted calculation, used to quantify the overall drug efficacy level of the candidate drug. The new drug evaluation unit assigns corresponding weights to the efficacy assessment results, safety-related weights to the toxicity assessment results, and clinical value-related weights to the clinical applicability assessment results, and generates a comprehensive score value through weighted summation or normalization. The comprehensive scoring system can compare and rank different candidate Chinese medicine new drugs while taking into account efficacy, safety and clinical feasibility, thereby forming a structured and quantifiable drug property evaluation result.

[0070] Understandably, by comprehensively scoring drug efficacy, toxicity risk, and clinical applicability according to preset weighting coefficients, the new drug evaluation unit can quantify the efficacy, safety, and clinical applicability of candidate drugs in a unified manner, forming a structured and highly comparable drug efficacy evaluation result. This comprehensive scoring method can avoid judgment errors caused by emphasizing a single indicator while taking into account the relative importance of each evaluation dimension, thus improving the scientificity and reliability of candidate drug selection.

[0071] In some embodiments of this application, when the registration application unit generates registration application materials, it includes:

[0072] The registration application unit extracts evaluation indicator data from the drug efficacy assessment results;

[0073] The registration and application unit establishes a mapping relationship between evaluation indicator data and registration and application data templates, fills the evaluation indicator data into the corresponding positions, and performs structured arrangement and format standardization processing on the filled content to generate registration and application materials.

[0074] Specifically, the registration application unit is used to transform the pharmacodynamic evaluation results generated by the aforementioned new drug evaluation unit into formal documents that can be used for the registration of new traditional Chinese medicines, namely, registration application documents. The registration application unit first extracts various evaluation indicator data from the pharmacodynamic evaluation results. These evaluation indicator data include, but are not limited to, comprehensive scores, scores for each dimension, active ingredient concentration information, toxicity risk levels, and indication matching degrees, to meet the key evaluation information required for the registration review of traditional Chinese medicines. Subsequently, the registration application unit establishes a mapping relationship between the evaluation indicator data and the registration application data template. The registration application data template refers to a standardized information framework pre-designed according to the requirements of drug regulatory agencies, containing field specifications for various modules such as pharmaceutical research, pharmacology and toxicology, clinical research, and comprehensive evaluation, to ensure the structural integrity and format consistency of the application documents. By establishing the mapping relationship, the registration application unit fills the evaluation indicator data into the corresponding template positions, achieving precise matching between the data and the template content. After the data is filled in, the registration application unit performs structured arrangement and format standardization. Structured arrangement refers to organizing various data and text information according to modular logic, ensuring clear logic and hierarchy across different research sections. Format standardization refers to unifying document formats, table formats, units, and measurement standards according to regulatory requirements to ensure the data meets submission standards. Finally, the registration application unit generates complete registration application materials.

[0075] Understandably, by extracting evaluation index data from the drug efficacy assessment results through the registration application unit and establishing a mapping relationship between the evaluation index data and the registration application data template, the data is accurately filled into the corresponding positions. At the same time, structured arrangement and format standardization are carried out, which improves the completeness, standardization and auditability of the registration application materials. It reduces the workload of manual sorting and filling, reduces the risk of human error, and ensures that various efficacy, safety and clinical applicability information is logically clear and hierarchically distinct in the application materials.

[0076] In some embodiments of this application, when the collaborative optimization unit generates the application data evaluation report, it includes:

[0077] The collaborative optimization unit divides the registration application materials into pharmaceutical research, pharmacological and toxicological research, and clinical research sections.

[0078] Check the consistency of data across different research sections; verify the logical coherence between different research sections; check whether the registration application materials meet the prescribed requirements and format specifications; and generate an evaluation report for the application materials.

[0079] Specifically, the collaborative optimization unit first divides the registration application materials into three main modules according to research type: pharmaceutical research, pharmacological and toxicological research, and clinical research. The pharmaceutical research section refers to research involving drug component analysis, formulation structure, and manufacturing process; the pharmacological and toxicological research section refers to research involving the drug's mechanism of action in vivo, toxicity assessment, and safety experiments; and the clinical research section refers to research involving the drug's efficacy, safety, and indication matching in humans or patients with target diseases. Subsequently, the collaborative optimization unit performs consistency checks on the data from each research section, comparing the data for the same indicator across different modules to ensure consistency and identify any conflicts or omissions. It also verifies the logical coherence between different research sections, checking whether the components and formulations in the pharmaceutical research reasonably correspond to the pharmacological and toxicological results, and whether the pharmacological and toxicological results logically match the clinical research observations, ensuring a complete data chain without contradictions. The collaborative optimization unit also checks whether the registration application materials meet the prescribed requirements and format specifications, which are derived from national drug regulatory agencies or international registration guidelines, including document structure, field names, units of measurement, data table formats, and the completeness of required fields. Through the above steps, the collaborative optimization unit can generate a structured, auditable, and logically complete evaluation report of the application materials.

[0080] Understandably, by using collaborative optimization units to divide registration application materials into pharmaceutical research, pharmacological and toxicological research, and clinical research sections, and checking the consistency of data and logical coherence across sections, while also verifying whether the materials meet the prescribed requirements and format specifications, the completeness, accuracy, and standardization of the application materials are improved. This method can promptly identify data conflicts, logical inconsistencies, or format errors, reducing the burden of manual review and the risk of potential omissions.

[0081] In some embodiments of this application, when the collaborative optimization unit feeds back the application data evaluation report to the multi-dimensional data association model for dynamic updating, it includes:

[0082] The collaborative optimization unit extracts validated and effective data on the correspondence between medicinal materials and components, new evidence on target-action relationships, and new drug-related information from the evaluation report of the application materials;

[0083] Update the medicinal material-component correspondence data to the medicinal material-component correspondence in the multi-dimensional data association model; update the new evidence of target-action relationship to the target-action relationship in the multi-dimensional data association model; update the new drug association information optimization points to the new drug association information in the multi-dimensional data association model;

[0084] Based on the updated medicinal material attribute information, medicinal material-component correspondence, target action relationship and new drug association information, the bidirectional association mapping between adjacent levels in the multi-dimensional data association model is adjusted.

[0085] Specifically, the collaborative optimization unit first extracts validated and effective medicinal material-component correspondence data, new evidence of target-action relationships, and new drug association information optimization points from the application evaluation report. Medicinal material-component correspondence data refers to the mapping information between confirmed or optimized medicinal materials and their main active ingredients. New evidence of target-action relationships refers to the information on the interaction between active ingredients and biological targets obtained through verification in registration materials or supplementary experimental data. New drug association information optimization points refer to the improved or supplementary data on the relationship between the efficacy of new drugs and traditional Chinese medicine new drugs identified in the evaluation report. Subsequently, the collaborative optimization unit updates the medicinal material-component correspondence data to the correspondence layer in the multi-dimensional data association model, updates the new evidence of target-action relationships to the target-action relationship layer, and updates the new drug association information optimization points to the new drug association information layer, thereby ensuring the timeliness and accuracy of the data at each layer. Based on the updated medicinal material attribute information, medicinal material-component correspondence, target action relationship and new drug association information, the collaborative optimization unit adjusts the bidirectional association mapping between adjacent levels in the multi-dimensional data association model. The bidirectional association mapping refers to the mapping relationship established between adjacent levels of data in the model that can be queried and updated in both directions, in order to maintain the consistency and integrity of information between levels.

[0086] Understandably, by using collaborative optimization units to feed back the validated medicinal material-component correspondence data, new evidence of target action relationships, and new drug association information optimization points from the application evaluation report to the multi-dimensional data association model, and updating the data and bidirectional association mapping at the corresponding levels, the accuracy and completeness of the model can be continuously optimized. This dynamic update mechanism enables the multi-dimensional data association model to promptly absorb the latest experimental verification and application review results, correct or supplement the original medicinal material component relationships, target actions, and new drug association information, thereby improving the accuracy and reliability of candidate traditional Chinese medicine new drug generation and drug property evaluation.

[0087] In some embodiments of this application, the acquisition of multi-source traditional Chinese medicine data and the subsequent data preprocessing include:

[0088] Data preprocessing includes text structure parsing, terminology standardization, format normalization, filling in missing key information, removing erroneous data, and establishing relationships.

[0089] Specifically, after acquiring multi-source TCM data, the data processing unit first performs data preprocessing to ensure the accuracy and reliability of subsequent multi-dimensional data association model construction. Text structuring involves converting unstructured text data (such as literature and case descriptions) into structured information tables or database records, enabling computer programs to recognize and process information such as medicinal material names, dosages, usages, components, and effects. Terminology standardization unifies different descriptions of the same medicinal material, component, disease, or symptom from different sources into standardized terminology, such as using the National Pharmacopoeia, International Classification of Diseases, or the TCM standardized terminology system. Format standardization unifies the data representation, units, field order, and data type, ensuring direct comparison and integration of data from different sources. Key information missing information filling involves filling missing key fields in data records, such as active ingredient content, dosage, or target information, through literature inference, database supplementation, or algorithmic prediction. Error data removal includes identifying and deleting duplicate, conflicting, or obviously erroneous data records, such as misspelled medicinal material names or outlier dosage values. Association establishment involves establishing associations such as medicinal material-component, component-target, and disease-new drug relationships in the preprocessed data based on literature, experimental, or database evidence. Through the above data preprocessing steps, multi-source TCM data are cleaned, standardized, and structured to form a high-quality dataset that can be used for accurate modeling, generation of candidate TCM new drugs, and drug property evaluation.

[0090] Understandably, by performing text structuring parsing, terminology standardization, format standardization, key information missing filling, error data removal, and relationship establishment on multi-source TCM data, data from different sources, formats, and expressions can be unified into a structured, standardized, and high-quality dataset. This data preprocessing method eliminates information redundancy, data conflicts, and missing data, and improves data consistency, integrity, and usability.

[0091] The following examples illustrate this in detail:

[0092] S1: The data processing unit first acquires multi-source TCM data from multiple sources, including new drug records of "Astragalus membranaceus, Codonopsis pilosula, and Angelica sinensis" from ancient TCM texts databases, active ingredient and target information of these herbs from modern pharmacology databases, and relevant new drug descriptions from clinical medical records. Subsequently, the data processing unit performs data preprocessing: it performs text structured parsing of ancient texts, converting the dosage, usage, and indications of each herb into a calculable tabular form; it standardizes the terminology for different expressions such as "Astragalus membranaceus" in different literature; it standardizes the dosage unit to grams and the format of time and frequency fields; it fills in missing active ingredient content, removes erroneous records, and establishes the correlation between herbs and ingredients, and between ingredients and targets.

[0093] S2: The data processing unit constructs a multi-dimensional data association model based on the preprocessed data. It organizes medicinal material attribute information, medicinal material-component correspondences, target action relationships, and new drug association information hierarchically, and establishes bidirectional association mappings between adjacent levels. For example, information on the immune-regulating targets corresponding to Astragalus polysaccharide, the main active ingredient of Astragalus, is associated with descriptions of new drugs used clinically for Qi deficiency syndrome. This enables the model to quickly match the relationships between medicinal materials, components, and new drugs during the generation of new traditional Chinese medicine drugs.

[0094] S3: In the new drug evaluation stage, the new drug evaluation unit receives the new drug characteristic description information of "qi deficiency and fatigue, pale complexion" from the patient group, inputs it into the multi-dimensional data association model, matches historical Chinese medicine records, determines the set of historical Chinese medicines related to the new drug, such as "Huangqi Jianzhong Decoction", extracts its basic Chinese medicine structure, and optimizes the active ingredients in the model to generate several candidate Chinese medicine new drug solutions, such as increasing the content of astragalus polysaccharide or adjusting the proportion of codonopsis, in order to improve the efficacy and compatibility safety.

[0095] S4: The new drug evaluation unit conducts joint analysis on several candidate traditional Chinese medicine (TCM) new drug schemes to evaluate the rationality of the TCM new drug structure, whether the composition and concentration of active ingredients are within the safe threshold range, and predict the comprehensive effects of the TCM new drug on immune, cardiac and liver targets through multi-target regulatory networks. At the same time, it conducts incompatibility screening to select safe and effective TCM new drug candidate schemes.

[0096] S5: In the multi-dimensional evaluation phase, the new drug evaluation unit calculates the efficacy intensity based on the concentration-effect relationship of the active ingredients in the candidate traditional Chinese medicine (TCM) new drugs, assesses the toxicity risk by combining the probability of adverse events, and evaluates the clinical applicability based on the degree of matching between the TCM new drug and the target indication. Subsequently, the efficacy, toxicity, and clinical applicability are comprehensively scored using preset weights to generate the final drug efficacy evaluation result.

[0097] S6: The registration application unit extracts evaluation indicators, such as efficacy score, toxicity risk value and clinical applicability index, based on the drug efficacy assessment results, maps them to the corresponding fields in the registration application template, completes data filling, and performs structured arrangement and format standardization of the data to form complete registration application materials.

[0098] S7: After generating the application materials, the collaborative optimization unit divides the materials into three parts: pharmaceutical research, pharmacological and toxicological research, and clinical research. It checks the consistency of data across each part and the logical coherence across parts, verifies compliance with format specifications, and generates an application materials evaluation report. Subsequently, the collaborative optimization unit feeds back the validated medicinal material-component correspondences, new evidence of target effects, and new drug association information optimization points from the evaluation report to the multi-dimensional data association model, updating information at each level and adjusting the bidirectional association mapping to achieve dynamic model updates.

[0099] In summary, by constructing an integrated intelligent system for the discovery and registration of new traditional Chinese medicine (TCM) drugs, intelligent management of the entire process—from multi-source TCM data collection, candidate TCM drug generation, pharmacodynamic evaluation, to the automated generation and optimization of registration application materials—was achieved. Utilizing multi-dimensional data association models and joint analysis methods, candidate TCM drugs with significant efficacy and safe compatibility were screened, improving the success rate and scientific rigor of new drug development. Through comprehensive evaluation of multiple indicators, the efficacy, toxicity, and clinical characteristics of candidate drugs were fully understood, reducing R&D risks. The registration application unit automatically completed data mapping, structured arrangement, and format standardization operations, reducing manual processing and filling workload, and improving registration efficiency and accuracy. The collaborative optimization unit ensured the completeness and compliance of application materials through data consistency checks and logical integrity verification, and could use feedback information to dynamically optimize the data association model, achieving intelligent closed-loop management of R&D-application.

[0100] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. An integrated intelligent agent system for the discovery and registration of new traditional Chinese medicine drugs, characterized in that, include: The data processing unit is configured to acquire multi-source traditional Chinese medicine data and perform data preprocessing; A multi-dimensional data association model is constructed based on the preprocessed multi-source traditional Chinese medicine data. The new drug evaluation unit is configured to input the characteristic description information of new traditional Chinese medicine drugs into the multi-dimensional data association model to generate several candidate new traditional Chinese medicine drug schemes; and to perform joint analysis on several candidate new traditional Chinese medicine drug schemes to screen out candidate new traditional Chinese medicine drug schemes that meet the compatibility and safety conditions. Based on efficacy indicators, toxicity indicators, and clinical indicators, the selected candidate Chinese medicine new drugs are evaluated in multiple dimensions to generate drug performance evaluation results. The registration application unit is configured to extract the evaluation indicators corresponding to the registration requirements from the drug efficacy assessment results based on the registration application data template, perform data mapping on the key evaluation indicators through field matching, and perform structured arrangement, format standardization and automatic module filling on the data mapping results to generate registration application materials. The collaborative optimization unit is configured to modularly organize the registration application materials, generate an application material evaluation report through data consistency checks and logical integrity verification, and feed the application material evaluation report back to the multi-dimensional data association model for dynamic updates.

2. The integrated intelligent agent system for the discovery and registration of new traditional Chinese medicine drugs according to claim 1, characterized in that, When the data processing unit constructs a multi-dimensional data association model based on the preprocessed multi-source traditional Chinese medicine data, it includes: The data processing unit acquires medicinal material attribute information based on the multi-source TCM data; determines the active ingredients contained in the medicinal materials, constructs a correspondence between medicinal materials and ingredients; determines the target interaction relationship between the active ingredients and biological targets; extracts the correlation information between diseases and the efficacy of new TCM drugs, and generates new drug correlation information. The data processing unit organizes the medicinal material attribute information, medicinal material-component correspondence, target action relationship and new drug association information according to hierarchical relationship, and establishes bidirectional association mapping between adjacent levels to form the multi-dimensional data association model.

3. The integrated intelligent agent system for the discovery and registration of new traditional Chinese medicine drugs according to claim 2, characterized in that, When the new drug evaluation unit inputs the characteristic description information of traditional Chinese medicine new drugs into the multi-dimensional data association model to generate several candidate traditional Chinese medicine new drug schemes, it includes: The new drug evaluation unit acquires new drug feature description information and matches new drug association information corresponding to the new drug feature description information in the multi-dimensional data association model; The new drug evaluation unit determines a set of historical Chinese medicines corresponding to the new drug feature description information; extracts basic Chinese medicine structures from the set of historical Chinese medicines; optimizes the basic Chinese medicine structures according to the multi-dimensional data association model to generate the candidate Chinese medicine new drug scheme.

4. The integrated intelligent agent system for the discovery and registration of new traditional Chinese medicine drugs according to claim 3, characterized in that, When the new drug evaluation unit performs joint analysis on several candidate traditional Chinese medicine (TCM) new drug schemes and screens for TCM new drug candidate schemes that meet the compatibility and safety conditions, it includes: The joint analysis includes assessment of the structural rationality of new traditional Chinese medicine drugs, assessment of the composition and concentration of active ingredients, prediction of multi-target regulatory networks, and screening of drug incompatibilities.

5. The integrated intelligent agent system for the discovery and registration of new traditional Chinese medicine drugs according to claim 4, characterized in that, The new drug evaluation unit performs multi-dimensional evaluations of the selected traditional Chinese medicine new drug candidates and generates drug efficacy evaluation results, including: The new drug evaluation unit determines the efficacy intensity based on the concentration-effect relationship of the active ingredients in the candidate traditional Chinese medicine drug scheme; determines the toxicity risk based on the probability of adverse events occurring in the candidate traditional Chinese medicine drug scheme; and determines the clinical applicability based on the degree of matching between the candidate traditional Chinese medicine drug scheme and the target indication.

6. The integrated intelligent agent system for the discovery and registration of new traditional Chinese medicine drugs according to claim 5, characterized in that, The new drug evaluation unit performs multi-dimensional evaluations of the selected traditional Chinese medicine new drug candidates and, when generating the drug efficacy evaluation results, also includes: The new drug evaluation unit comprehensively scores the efficacy, toxicity risk, and clinical applicability of the drug according to preset weighting coefficients, and generates the drug efficacy evaluation result.

7. The integrated intelligent agent system for the discovery and registration of new traditional Chinese medicine drugs according to claim 1, characterized in that, When the registration application unit generates registration application materials, it includes: The registration application unit extracts evaluation indicator data from the drug efficacy assessment results; The registration and application unit establishes a mapping relationship between the evaluation indicator data and the registration and application data template, fills the evaluation indicator data into the corresponding positions, and performs structured arrangement and format standardization processing on the filled content to generate the registration and application materials.

8. The integrated intelligent agent system for the discovery and registration of new traditional Chinese medicine drugs according to claim 1, characterized in that, When the collaborative optimization unit generates the evaluation report for the application materials, it includes: The collaborative optimization unit divides the registration application materials into pharmaceutical research, pharmacological and toxicological research, and clinical research sections. Check the consistency of data across different research sections; verify the logical coherence between different research sections; check whether the registration application materials meet the prescribed requirements and format specifications; and generate an evaluation report for the application materials.

9. The integrated intelligent agent system for the discovery and registration of new traditional Chinese medicine drugs according to claim 8, characterized in that, When the collaborative optimization unit feeds back the application material evaluation report to the multi-dimensional data association model for dynamic updates, it includes: The collaborative optimization unit extracts verified and effective medicinal material-component correspondence data, new evidence of target-action relationships, and new drug association information optimization points from the application material evaluation report; The medicinal material-component correspondence data is updated to the medicinal material-component correspondence in the multi-dimensional data association model; the new evidence of target action relationship is updated to the target action relationship in the multi-dimensional data association model; the optimization points of new drug association information are updated to the new drug association information in the multi-dimensional data association model. Based on the updated medicinal material attribute information, medicinal material-component correspondence, target action relationship and new drug association information, the bidirectional association mapping between adjacent levels in the multi-dimensional data association model is adjusted.

10. The integrated intelligent agent system for the discovery and registration of new traditional Chinese medicine drugs according to claim 1, characterized in that, When acquiring multi-source traditional Chinese medicine data and performing data preprocessing, the following steps are included: The data preprocessing includes text structure parsing, terminology standardization, format normalization, filling in missing key information, removing erroneous data, and establishing relationships.

Images