Molecular docking analysis method of core target based on RCSB database

Abstract

This invention discloses a molecular docking analysis method for obtaining core targets based on the RCSB database, comprising: screening drug-disease-immunity intersection targets through multiple databases; screening core targets through protein interaction network analysis and multi-topology algorithms; obtaining core target structure files from the RCSB database and downloading active ingredient structure files from the PubChem database; performing molecular docking through the CB-DOCK2 database; screening effective binding pairs using binding energy as an indicator and visualizing the results; and finally outputting the results through functional annotation and pathway enrichment analysis. This method improves the accuracy of core target screening and the reliability of molecular docking through multi-database integration, multi-algorithm collaboration, and multi-dimensional evaluation, forming a complete technical chain and providing efficient technical support for the analysis of the mechanisms of action of traditional Chinese medicine compound prescriptions.

Description

Technical Field

[0001] This invention relates to the field of molecular docking analysis technology, and in particular to a molecular docking analysis method based on the RCSB database for obtaining core target points. Background Technology

[0002] Traditional Chinese medicine (TCM) compound formulas, with their synergistic effects involving multiple components, targets, and pathways, exhibit unique advantages in disease treatment. However, their complex composition and ambiguous mechanisms of action significantly limit their clinical application and modern development. Accurate identification of core targets and molecular docking analysis are key technologies for elucidating the mechanisms of action of TCM compound formulas. The RCSB database, as an authoritative protein structure resource, provides reliable support for target structure acquisition. Currently, the mainstream technical approach for TCM compound formula mechanism research is to screen drug-disease-immunity intersection targets based on multi-database integration, combine this with protein-protein interaction network analysis to screen core targets, and then verify the binding activity between targets and active ingredients through molecular docking. However, existing methods still have room for improvement in terms of core target screening accuracy, molecular docking efficiency, and result reliability. There is an urgent need to construct a systematic and efficient integrated analysis method to provide technical support for elucidating the mechanisms of action of TCM compound formulas.

[0003] Existing technologies have two significant drawbacks: First, in the process of core target screening, most methods rely on only a single topology algorithm or a few parameters to sort targets, failing to fully integrate multi-dimensional topological features. This results in insufficient targeting of the screened core targets, making it difficult to comprehensively reflect the interaction between drugs and diseases and the immune network, thus affecting the accuracy of subsequent molecular docking and mechanism analysis. Second, the molecular docking and result evaluation system is not perfect. The accuracy of identifying target active sites during docking is limited, and the stability assessment often relies on a single binding energy index, failing to fully consider key factors such as the type of interaction bonds and conformational stability between targets and components. This leads to the omission of some potential effective target-component combinations, reducing the comprehensiveness and reliability of the analysis results. Summary of the Invention

[0004] To overcome the shortcomings and deficiencies of existing technologies, this invention provides a molecular docking analysis method for obtaining core target points based on the RCSB database.

[0005] The technical solution adopted in this invention is a molecular docking analysis method for obtaining core targets based on the RCSB database. It is characterized by the following steps: S1, screening potential targets corresponding to the blood-entering active components of two or more medicinal materials in traditional Chinese medicine compound formulas using TCMSP, ETCM, and HERB databases; simultaneously searching for associated targets of target diseases and immune-related diseases using GeneCards, DisGeNET, and DrugBank databases; and using the Venny tool for target mapping to obtain drug-disease-immunity intersection targets; S2, importing the intersection targets into the STRING database and setting confidence thresholds to construct a protein-protein interaction network; exporting the network data and performing module analysis using the MCODE plugin of Cytoscape software; and using the CytoHubba plugin based on degree value, closure, and... Four topology algorithms—eccentricity, MNC, and others—were used to screen core targets. S3: PDB format files corresponding to the screened core targets were obtained from the RCSB database, and two-dimensional structure files of the analyzed active ingredients in the traditional Chinese medicine compound were downloaded from the PubChem database. S4: Molecular docking was performed on the PDB format files of the core targets and the two-dimensional structure files of the active ingredients using the CB-DOCK2 database, with default database parameter settings used for the docking operation. S5: Binding energy was used as the docking effectiveness evaluation index to screen target-component binding pairs whose binding energies met preset thresholds, and binding pattern diagrams were generated using a molecular docking visualization tool. S6: GO functional annotation and KEGG pathway enrichment analysis were performed on the screened core targets using the DAVID database, setting significance criteria and outputting the enrichment results.

[0006] Furthermore, the topology algorithm in S2 uses the following model formula when selecting core target points: ,in, For the comprehensive score of the target, These are the weighting coefficients for each topology parameter. The target point value. The target closure coefficient. For the target eccentricity parameter, This represents the maximum number of neighborhood components of the target.

[0007] Furthermore, the binding energy prediction expression for the molecular docking process in S4 is as follows: ,in, For binding energy, For van der Waals action energy weights, For van der Waals action energy, The electrostatic interaction energy coefficient is... It is the energy of electrostatic interaction. The hydrogen bond interaction energy weights, The hydrogen bond interaction energy, The solvation energy coefficient, This is the solvation entropy value.

[0008] Furthermore, the expression for determining the significance of KEGG pathway enrichment in S6 is as follows: ,in, The P-value represents the significance of the enrichment difference. This represents the total number of pathway-related targets in the database. The total number of core targets, To enrich the number of core targets in the target pathway, The total number of targets included in the target pathway. This is the pathway enrichment correction coefficient.

[0009] Furthermore, the intersection target selection expression in S1 is: ,in, To screen confidence levels for intersection targets, It is a collection of drug active ingredient targets. A set of disease-related targets, A collection of immune-related targets, The intersection weight coefficients of the sets To filter deviation correction values.

[0010] Furthermore, the expression for evaluating the target-component binding stability in S5 is as follows: ,in, To incorporate stability scoring, The coefficient of contribution to binding energy. For binding energy, This represents the number of interaction bonds between the target and the component. For conformational stability weights, To incorporate conformational deviation rate.

[0011] Further, S2 includes the following steps: S21, after standardizing the gene names of the intersection target points, import them into the STRING database, set the confidence parameters and select the species matching option, and submit to obtain the interaction data between target points; S22, convert the exported interaction data into a Cytoscape compatible format, import it into the software, and then use the MCODE plugin to set the node degree threshold and clustering coefficient threshold for module partitioning; S23, start the CytoHubba plugin, and sequentially select four algorithms: degree value, closure, eccentricity, and MNC, to calculate and sort the topological parameter values ​​of each target point; S24, extract the common target points from the sorting results of the four algorithms, remove duplicates, form a core target point set, and output it.

[0012] Furthermore, S3 includes the following steps: S31, log in to the RCSB database, input the standard gene symbol of the core target, screen crystal structure files with consistent species origin and meeting resolution requirements, and download the PDB format file; S32, access the PubChem database, search for the chemical name or CAS number of the active ingredient identified in the traditional Chinese medicine compound, select the compound entry with clear purity identification, and download its two-dimensional structure SDF format file; S33, verify the target structure of the downloaded PDB file, confirm the integrity of the active site, perform structural standardization processing on the SDF file, and remove redundant atoms and chemical bond information.

[0013] Further, step S4 includes the following steps: S41, opening the CB-DOCK2 database online platform, uploading the PDB format file of the core target, specifying the active site prediction region or using the automatic identification mode; S42, uploading the SDF format file of the active ingredient, setting the grid parameters for docking operation and the default parameters for the number of conformations searched, and submitting the docking task; S43, waiting for the operation to complete, downloading the docking result file, including binding energy data, target-component interaction list, and conformation file; S44, converting the format of the result file to ensure that the data is compatible with subsequent visualization tools.

[0014] Further, S5 includes the following steps: S51, extracting binding energy data from the docking results, comparing it with a preset threshold, and screening out target-component combinations that meet the conditions; S52, importing the screened conformational files into the PyMOL visualization tool, adjusting the perspective to display the binding mode of the target and components, and marking hydrogen bonds and hydrophobic interactions to calibrate the interactions; S53, counting the number and type of interaction bonds for each target-component combination to form a quantitative analysis table; S54, integrating the binding energy data and interaction analysis results to generate a comprehensive molecular docking evaluation report.

[0015] Beneficial Effects: This invention proposes a molecular docking analysis method for obtaining core targets based on the RCSB database. First, it integrates multiple databases to screen for drug-disease-immunity intersection targets. Then, it utilizes a multi-topology algorithm combination to screen core targets, fully integrating topological features from different dimensions. This solves the problem of insufficient targeting in traditional single-algorithm screening, improving the comprehensiveness and accuracy of core target screening and ensuring that the screening results truly reflect the interaction relationships among the three. Through standardized target and active ingredient structure acquisition and processing procedures, combined with a multi-factor comprehensive binding energy prediction and binding stability assessment model, it supplements the limitations of single binding energy indicators. Simultaneously, it refines the active site identification and conformational analysis steps, reducing the omission of potential effective target-component combinations and significantly improving the reliability and comprehensiveness of molecular docking results. Furthermore, this method combines molecular docking with functional annotation and pathway enrichment analysis systems to form a complete technical chain. This ensures the systematic nature of core target screening and docking analysis, and reduces technical errors through standardized operating procedures. It provides more accurate and comprehensive technical support for the analysis of the mechanisms of action of traditional Chinese medicine compound prescriptions, effectively compensating for the shortcomings of existing technologies in terms of screening accuracy and evaluation systems. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the overall process of the method of the present invention.

[0017] Figure 2 This is a flowchart of method step S2 of the present invention;

[0018] Figure 3 This is a flowchart of method step S3 of the present invention;

[0019] Figure 4 This is a flowchart of method step S4 of the present invention;

[0020] Figure 5 This is a flowchart of step S5 of the method of the present invention. Detailed Implementation

[0021] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0022] like Figure 1As shown, the molecular docking analysis method for obtaining core targets based on the RCSB database includes the following steps: S1, screening potential targets corresponding to the blood-entering active components of two or more medicinal materials in traditional Chinese medicine compound prescriptions through TCMSP, ETCM, and HERB databases; simultaneously searching for related targets of target diseases and immune-related diseases through GeneCards, DisGeNET, and DrugBank databases; and using the Venny tool to perform target mapping to obtain drug-disease-immunity intersection targets; S2, importing the intersection targets into the STRING database and setting confidence thresholds to construct a protein interaction network; exporting the network data and performing module analysis using the MCODE plugin of Cytoscape software; and using the CytoHubba plugin based on degree value, closure, eccentricity, and MNC four-factor analysis. S3. A topological algorithm is used to screen core targets; S4. PDB format files corresponding to the screened core targets are obtained from the RCSB database, and two-dimensional structure files of the active ingredients in the traditional Chinese medicine compound are downloaded from the PubChem database; S5. Molecular docking is performed on the PDB format files of the core targets and the two-dimensional structure files of the active ingredients using the CB-DOCK2 database, and the docking operation is completed using the default parameters of the database; S6. Using the binding energy value as the docking effect evaluation index, target-component binding pairs with binding energies meeting the preset threshold are screened, and a binding pattern diagram is generated using a molecular docking visualization tool; S7. GO function annotation and KEGG pathway enrichment analysis are performed on the screened core targets using the DAVID database, and the criteria for judging the significance of differences are set and the enrichment results are output.

[0023] Step S1 aims to screen drug-disease-immunity intersection targets, laying the foundation for subsequent core target discovery. Its implementation must strictly adhere to the technical specifications for multi-database collaborative retrieval and target mapping. Specifically, for traditional Chinese medicine (TCM) compound prescriptions, two or more core medicinal materials are selected. The TCMSP database is used to screen for active ingredients with oral bioavailability of at least 30% and drug-likeness of at least 0.18 that enter the bloodstream. Simultaneously, the ETCM database is used to supplement information on target-related active ingredients of traditional Chinese medicine, and the HERB database is used to retrieve known targets corresponding to the ingredients. This combination ensures the comprehensiveness of potential drug targets. On the other hand, for target diseases and immune-related diseases, disease-related targets are obtained by setting a relevance score of at least 20 in the GeneCards database. The DisGeNET database is used to screen for high-confidence targets with a confidence score of at least 0.7. Combined with drug target information related to disease treatment in the DrugBank database, a set of disease- and immune-related targets is constructed. Subsequently, the set of potential drug targets and the set of disease and immune-related targets were imported into the Venny tool. The tool's built-in target mapping algorithm was used to perform intersection calculations. This calculation process is based on gene name standardization matching, which automatically removes redundant targets with inconsistent gene names or synonyms with different names. Finally, a list of intersection targets that exist simultaneously in the three dimensions of drugs, diseases, and immunity is output. This step effectively improves the relevance and reliability of targets through complementary retrieval of multiple databases and strict screening threshold settings, providing a high-quality target dataset for subsequent protein-protein interaction network analysis.

[0024] Step S2 uses protein-protein interaction network analysis and multi-topology algorithms to screen core targets. Its implementation requires combining database parameter settings with professional software analysis functions to achieve accurate screening. Specifically, the gene names of the intersection target points obtained in Step S1 are first standardized to ensure that the gene symbols are consistent with the naming conventions of the STRING database. Then, the standardized targets are imported into the STRING database, with the species set to human and the confidence threshold set to 0.7. Three interaction evidence sources are selected: "Experimental Validation," "Database Annotation," and "Text Mining." After submission, a protein-protein interaction network is generated, and the network data file, including nodes, edges, and interaction scores, is exported. The exported network data is converted to a Cytoscape-compatible CSV format, imported into the software, and the MCODE plugin is launched. Parameters are set as follows: node degree threshold of 2, clustering coefficient threshold of 0.2, node density threshold of 0.15, and maximum depth of 100. The protein-protein interaction network is then divided into modules, and target modules with close functional relationships are screened out. Next, the CytoHubba plugin was launched, and four topology algorithms—degree, closure, eccentricity, and MNC—were run sequentially. The degree algorithm calculates the number of direct interactions between each target and other targets. The closure algorithm assesses the density of the local network where the target is located. The eccentricity algorithm reflects the centrality of the target in the entire network. The MNC algorithm counts the number of components in the target's largest neighborhood. The four algorithms quantify the network importance of the target from different dimensions. The software automatically calculates and outputs the four topology parameters for each target. Then, targets whose rankings for all four parameters are in the top 20% are extracted. After removing duplicates, a core target set is formed. This step, through multi-algorithm collaborative analysis, effectively screens out core targets that play a key regulatory role in the protein-protein interaction network, improving the targeting and efficiency of subsequent molecular docking.

[0025] Step S3 involves obtaining standardized core target structure files and active ingredient structure files to provide qualified input data for molecular docking. The completeness and standardization of the structure files must be strictly controlled during this process. For obtaining core target structure files, log in to the RCSB database website and enter the standard gene symbol of each core target obtained in step S2 into the search bar. The search results should be filtered for entries from human species, with a resolution no higher than 2.5 Å, no mutation sites, and a crystal structure integrity of 95% or higher. Priority should be given to crystal structures including natural ligands to ensure the accuracy of active sites. After identifying the target entries, download the corresponding PDB format file. This file includes key information such as the atomic coordinates, chemical bond connections, amino acid sequences, and spatial conformation of the target. After downloading, the file format should be initially verified using a text editor to ensure there is no missing data or format errors. Regarding the acquisition of active ingredient structure files, based on the blood-entry active ingredients screened in step S1, the PubChem database was accessed. Compounds with a purity of 98% or higher and no isomer confusion were selected and downloaded in SDF format. This format fully preserves the molecular structure, functional group positions, and chemical bond types of the compound. After obtaining both structure files, preliminary structure verification is required. For PDB format files, the focus is on verifying the completeness of atomic coordinates in the active site region and the absence of missing amino acid residues. For SDF format files, the molecular structure is checked for redundant atoms and the correctness of chemical bond connections. This ensures that both structure files meet the format requirements and quality standards for subsequent molecular docking. This rigorous structure screening and verification provides fundamental data support for the accuracy of molecular docking.

[0026] Step S4 is the core implementation step of molecular docking. It utilizes algorithms from a professional database to achieve virtual binding between the core target and the active ingredient. The implementation process must strictly adhere to docking parameter settings and computational specifications. Specifically, the CB-DOCK2 online database platform is opened. This platform integrates automatic target identification, conformation search, and binding energy calculation functions, making it suitable for large-scale molecular docking analysis. First, click the target file upload entry and select the core target PDB format file obtained in step S3 for upload. After upload, the platform automatically parses the file and identifies potential active sites. At this point, you can choose to use the platform's automatic identification mode or manually specify the coordinate range of the active site based on known target structure information. When manually specifying, the coordinate range in the x, y, and z axes must be set based on the amino acid residue positions of the active site in the PDB file to ensure complete coverage of the active pocket region. Next, click the ingredient file upload entry and batch upload the active ingredient SDF format files obtained in step S3. During the upload process, ensure that the file naming is standardized to avoid duplicate names or format errors that could lead to upload failure. The parameter setting stage uses the database's default optimal parameter configuration, with the grid box size set to 60×60×60, the grid spacing to 0.375, and 100 different conformations generated for each component. The docking algorithm employs an improved version based on the Lamarckian genetic algorithm, which can efficiently search for the optimal binding conformation between the target and the component, while simultaneously calculating the binding energy using molecular mechanics and solvation models. After parameter settings are completed, the docking task is submitted, and the platform's backend starts the computation program. Multiple target-component docking combinations are processed using parallel computing technology, with real-time progress feedback during the computation. Once all docking combinations are completed, the platform generates a result file including binding energy data, conformation information, and interaction types for each target-component combination. This step, through standardized parameter settings and efficient algorithm computation, achieves precise virtual docking between the core target and the active ingredient, providing data support for subsequent binding activity screening.

[0027] The main function of step S5 is to screen highly active target-component binding pairs and visualize binding patterns. This improves the usability of the docking results through quantitative evaluation and intuitive visualization. The implementation process requires the integration of data screening and visualization analysis techniques. In the binding energy screening stage, binding energy data for all target-component combinations are first extracted from the docking result file of step S4. Based on recognized screening standards in the field, a threshold of no more than -5 for binding energy is set. This threshold indicates a strong binding ability between the target and the component, possessing potential biological activity. Data processing tools are used to compare the binding energy data with the preset threshold, automatically screening target-component binding pairs that meet the threshold requirements and eliminating low-activity combinations with binding energies higher than the threshold. Simultaneously, the specific binding energy value of each highly active combination is recorded, forming a preliminary screening result list. Following this, a visualization analysis of the binding patterns was performed. The conformational files corresponding to the screened highly active combinations were imported into the PyMOL visualization tool, which can display the binding state of the target and the component in three dimensions. In practice, the view angle was first adjusted to clearly show the binding region between the active site and the component. Using the tool's color-coding function, the amino acid residues of the target, the active site pocket, and the component molecule were each set to different colors. Then, the hydrogen bond marking function was used to automatically identify and mark the hydrogen bonds formed between the target and the component, clarifying the donor, acceptor, and bond length of the hydrogen bonds. At the same time, the locations of other interaction types such as hydrophobic interactions and electrostatic interactions were marked. After visualization, a high-resolution binding pattern diagram was generated. The diagram should clearly show the binding posture of the component molecule in the target active pocket, the location and type of key interactions, and combined with the binding energy data obtained from the screening, to form the quantitative and visual analysis results of the target-component binding pairs. This step, through binding energy threshold screening and three-dimensional visualization, ensured the accurate screening of highly active binding pairs and provided an intuitive structural basis for subsequent mechanism analysis.

[0028] Step S6 reveals the biological functions and pathways of action of core targets through functional annotation and pathway enrichment analysis, providing crucial evidence for elucidating the mechanism of action of traditional Chinese medicine compound prescriptions. The implementation process must strictly adhere to the database analysis workflow and significance judgment criteria. Specifically, the core target set obtained in step S2 is first organized into a standard gene symbol list, ensuring no spelling errors or synonyms. Then, the DAVID database is logged in, the gene list upload mode is selected, and the core target gene symbol list is uploaded to the database, setting the species to human. After submission, the database automatically identifies and matches the corresponding gene information. In GO functional annotation analysis, three dimensions are selected for annotation: biological processes, cellular components, and molecular functions. The threshold is set to an enrichment fold of at least 1.5, and the significance judgment criterion is a p-value less than 0.05. The database automatically calculates the annotation results for each core target in the three dimensions by comparing with the built-in GO annotation database, outputting an analysis report including annotation entries, enrichment fold, p-value, and the number of corresponding targets. The biological process dimension reflects the biological events involved by the target, the cellular component dimension clarifies the subcellular location of the target, and the molecular function dimension reveals the molecular mechanism of action of the target. In the KEGG pathway enrichment analysis, a significance criterion of at least 2.0 fold enrichment and a p-value less than 0.05 was set. Based on the association between core targets and KEGG pathways, the database used statistical algorithms to calculate the enrichment level of each pathway, screening out pathways with significant enrichment of core targets. The output included pathway name, enrichment fold, p-value, number of genes, and pathway number, along with links to the pathway's metabolic map. After the analysis, the GO functional annotation and KEGG pathway enrichment results were integrated, and duplicate or biologically insignificant entries were removed to generate a structured analysis report. This step, through systematic functional and pathway analysis, clarified the biological functions of core targets and the signaling pathways they participate in, providing a scientific basis for revealing the synergistic mechanism of multiple components, multiple targets, and multiple pathways in traditional Chinese medicine compound formulas.

[0029] Preferably, the model formula used in S2 for selecting core target points using the topology algorithm is: ,in, For the comprehensive score of the target, These are the weighting coefficients for each topology parameter. The target point value. The target closure coefficient. For the target eccentricity parameter, This represents the maximum number of neighborhood components of the target.

[0030] Specifically, in step S2, the model for selecting core targets using a topology algorithm achieves accurate scoring of core targets through weighted integration of multi-dimensional topology parameters. The implementation process requires combining protein-protein interaction network characteristics with algorithm parameter calibration to ensure the reliability of the results. In practice, the weight coefficients of each topology parameter need to be adjusted based on the size and complexity of the protein-protein interaction network. Typically, α ranges from 0.3 to 0.4, β from 0.25 to 0.35, γ from 0.15 to 0.25, and δ from 0.1 to 0.2, with the sum of the four weight coefficients being 1. The optimal weight combination for screening targets in traditional Chinese medicine compound formulas is determined by training the network data of a large number of known core targets using the analytic hierarchy process (AHP). The target degree value is directly calculated using Cytoscape software, representing the number of direct interactions between the target and other proteins. The value range varies with the network size, typically between 5 and 50. The target closure coefficient reflects the density of the local network where the target is located. It is calculated by the software to determine the proportion of triangular structures around the node, with a value range of 0-1. The closer the value is to 1, the stronger the local network correlation. The target eccentricity parameter is obtained by calculating the ratio of the maximum to the minimum shortest path from the target to all other nodes in the network. The value range is 1 - network diameter. The smaller the value, the closer the target is to the center of the network. The number of the target's maximum neighborhood components is obtained by counting the number of nodes in the maximum connected subgraph formed by the nodes directly connected to the target. The value range is 3-30. The larger the value, the stronger the local influence of the target. During implementation, the four topological parameters of each intersection target point are first calculated using the CytoHubba plugin, and then substituted into the model for comprehensive scoring. The top 20% of the targets in the comprehensive score are selected as core target candidates. By integrating multi-dimensional topological features, this model avoids the one-sidedness of single-parameter screening, making the core targets more reflective of the key regulatory nodes of the protein-protein interaction network, thereby improving the targeting and effectiveness of subsequent molecular docking. The parameter settings can be fine-tuned according to the target network characteristics of different traditional Chinese medicine compound prescriptions to ensure model adaptability and result stability.

[0031] Preferably, the binding energy prediction expression for the molecular docking process in S4 is: ,in, For binding energy, For van der Waals action energy weights, For van der Waals action energy, The electrostatic interaction energy coefficient is... It is the energy of electrostatic interaction. The hydrogen bond interaction energy weights, The hydrogen bond interaction energy, The solvation energy coefficient, This is the solvation entropy value.

[0032] Specifically, the binding energy prediction expression for molecular docking in step S4 achieves accurate calculation of the binding energy by integrating multiple intermolecular interaction energies, providing a quantitative basis for evaluating the target-component binding activity. The implementation process requires setting parameters based on molecular structure characteristics and docking algorithm features. The weights and coefficients of each interaction energy in the model need to be calibrated based on a large amount of known active target-component combination data. Specifically, the van der Waals interaction energy weight ω is set to 0.4-0.5, the electrostatic interaction energy coefficient ζ to 0.3-0.4, the hydrogen bond interaction energy weight η to 0.1-0.15, and the solvation energy coefficient θ to 0.05-0.1. All parameters are optimized through multiple linear regression analysis to ensure that the correlation coefficient between the model predictions and experimental measurements is not less than 0.85. Van der Waals interaction energies are obtained by calculating the hydrophobic interactions and van der Waals attraction and repulsion between the target and component molecules, typically ranging from -50 to -10. Electrostatic interaction energies are calculated based on differences in molecular surface charge distribution, ranging from -30 to 5. Hydrogen bonding interaction energies are calculated based on the number and length of hydrogen bonds formed between the target and component, ranging from -15 to 0. Solvation entropy is obtained by simulating the dissolution state of molecules in aqueous solution, ranging from 10 to 50. During implementation, the CB-DOCK2 database docking algorithm automatically extracts the molecular structural features of the target and component, calculates the values ​​of various interaction energies, and substitutes them into the model to obtain the binding energy results. This model comprehensively considers the key types of intermolecular interactions, overcoming the limitations of single interaction energy assessment, making the binding energy prediction more closely reflect the actual binding state, and providing a precise quantitative standard for subsequent screening of highly active binding pairs. The parameters can be adaptively adjusted according to the chemical structure type of the component and the characteristics of the target's active site.

[0033] Preferably, the expression for determining the significance of KEGG pathway enrichment in S6 is: ,in, The P-value represents the significance of the enrichment difference. This represents the total number of pathway-related targets in the database. The total number of core targets, To enrich the number of core targets in the target pathway, The total number of targets included in the target pathway. This is the pathway enrichment correction coefficient.

[0034] Specifically, the KEGG pathway enrichment significance judgment expression in step S6 quantifies the degree of pathway enrichment using statistical algorithms, providing a basis for screening key pathways that play a crucial role in core targets. The implementation process must strictly adhere to database statistical rules and parameter setting specifications. Each parameter in the model must be determined based on the built-in pathway information and the number of core targets in the database. The total number of pathway-related targets in the database is calculated based on the latest version of the KEGG database's pathway annotation information, with values ​​varying depending on the pathway type, typically between 50 and 500. The total number of core targets is the actual number of core targets screened in step S2, generally controlled between 30 and 100. The number of core targets enriched into the target pathway is obtained through database matching calculations, with values ​​ranging from 3 to 30. The total number of targets included in the target pathway is the total number of targets annotated for that pathway in the KEGG database, with values ​​ranging from 20 to 300. The pathway enrichment correction coefficient λ needs to be set according to the multiple test correction requirements, with values ​​ranging from 0.8 to 1.0, to reduce the false positive rate. During implementation, after uploading the list of core targets to the DAVID database, the database automatically matches the KEGG pathways corresponding to each target. The model calculates the significance P-value for the enrichment difference of each pathway, setting a P-value less than 0.05 as the significant enrichment criterion. Pathways highly involved by the core targets are screened out. This model uses rigorous statistical calculations to eliminate random enrichment interference, ensuring the reliability of the pathway enrichment results. The correction coefficient can be adjusted according to the number of core targets and the total number of pathways, making the judgment criteria more in line with the experimental design requirements, and providing a scientific basis for elucidating the pathway mechanism of action of traditional Chinese medicine compound prescriptions.

[0035] Preferably, the intersection target selection expression in S1 is: ,in, To screen confidence levels for intersection targets, It is a collection of drug active ingredient targets. A set of disease-related targets, A collection of immune-related targets, The intersection weight coefficients of the sets To filter deviation correction values.

[0036] Specifically, in step S1, the intersection target screening expression enhances the confidence of intersection targets through set operations and parameter correction, ensuring that the screening results simultaneously satisfy the correlation of drug, disease, and immunity dimensions. The implementation process requires setting parameters based on the target characteristics in the database. The set intersection weight coefficient ε ranges from 0.6 to 0.8, determined through training with a large amount of drug-disease-immunity target correlation data, and is used to strengthen the confidence of true intersection targets. The screening bias correction value μ ranges from 0.1 to 0.2, used to correct screening errors caused by differences in target naming and annotation bias between databases. The drug active ingredient target set is the union of screening results from the TCMSP, ETCM, and HERB databases; the disease-related target set is the intersection of search results from the GeneCards and DisGeNET databases; and the immunity-related target set mainly comes from immune-regulating drug targets in the DrugBank database. In practice, the gene names of the three target sets are first standardized. The intersection ratio of the three sets is calculated using the Venny tool. The intersection target screening confidence is obtained by substituting the values ​​into the model. A confidence score of not less than 0.7 is set as the qualified standard. Intersection targets that meet the requirements are screened out. By quantifying the intersection confidence and bias correction, the model reduces the omission or redundancy of targets caused by the limitations of a single database, improves the specificity and reliability of intersection targets, and lays a high-quality data foundation for subsequent core target screening. The parameters can be flexibly adjusted according to the types of Chinese herbal medicines and the type of target disease.

[0037] Preferably, the expression for evaluating the target-component binding stability in S5 is: ,in, To incorporate stability scoring, The coefficient of contribution to binding energy. For binding energy, This represents the number of interaction bonds between the target and the component. For conformational stability weights, To incorporate conformational deviation rate.

[0038] Specifically, the target-component binding stability assessment expression in step S5 comprehensively evaluates the binding reliability of the target and component by integrating multiple dimensions of indicators such as binding energy, number of interaction bonds, and conformational stability. The implementation process requires setting parameters based on docking result data and visualization analysis. The binding energy contribution coefficient κ ranges from 0.5 to 0.6, highlighting the core influence of binding energy on binding stability; the conformational stability weight ν ranges from 0.4 to 0.5, supplementing the impact of conformational deviation on binding stability. The binding energy is the raw data from the docking results in step S4. The number of interaction bonds is statistically analyzed using PyMOL visualization tools, including the total number of key interaction types such as hydrogen bonds, hydrophobic interactions, and electrostatic interactions, with a value range of 3 to 15. The binding conformational deviation rate is calculated by comparing the structural differences between the 100 conformations generated by docking and the optimal conformation, with a value range of 0.01 to 0.1; a smaller value indicates a more stable conformation. During implementation, the binding energy, number of interacting bonds, and conformational deviation rate data of each target-component combination are extracted and substituted into the model to calculate the binding stability score. A score of not less than 0.6 is set as the standard for stable binding. Target-component combinations with both high binding activity and high stability are screened out. This model breaks through the limitations of single binding energy assessment, comprehensively considers binding strength and structural stability, and improves the biological reliability of the screening results. The parameters can be adaptively adjusted according to the size of the component molecules and the structural complexity of the target active site to ensure the rationality and adaptability of the assessment standard.

[0039] Preferred, such as Figure 2 As shown, S2 includes the following steps: S21, after standardizing the gene names of the intersection target points, import them into the STRING database, set the confidence parameters and select the species matching option, and submit to obtain the interaction data between target points; S22, convert the exported interaction data into a Cytoscape compatible format, import it into the software, and then use the MCODE plugin to set the node degree threshold and clustering coefficient threshold to perform module partitioning; S23, start the CytoHubba plugin, and select four algorithms in sequence: degree value, closure, eccentricity, and MNC, to calculate and sort the topological parameter values ​​of each target point; S24, extract the common target points in the sorting results of the four algorithms, remove duplicates, form a core target point set, and output it.

[0040] Specifically, step S2 is implemented in stages, using standardized database operations and software plugins for collaborative analysis to achieve precise screening of core targets. Each stage is closely linked, and the parameters are set scientifically and rigorously. In S21, the gene names of the intersection target points obtained in step S1 are first standardized according to the HGNC naming convention, eliminating targets with non-standard abbreviations or synonymous but different names, ensuring that the gene symbol of each target point is unique and accurate. Then, the STRING database is logged in, the "Multiple proteins" mode is selected, the standardized target list is uploaded, the species is set to human, the confidence threshold is adjusted to 0.7, and three interaction evidence types are selected: experimental verification, database annotation, and text mining. After submission, the database automatically retrieves the interaction relationships between targets and generates network data including node names, interaction scores, and evidence sources. After completion, it is exported as a TSV format file. In S22, a data conversion tool is used to convert the TSV format network data into a Cytoscape-compatible CSV format, ensuring complete mapping of node and edge information. After importing into the software, the MCODE plugin is launched, and parameters are set as follows: node degree threshold of 2, clustering coefficient threshold of 0.2, node density threshold of 0.15, and maximum depth of 100. The plugin automatically identifies functionally interconnected sub-modules in the network and outputs a list of target points and module scores for each module. In S23, the CytoHubba plugin is launched, and four topology algorithms—degree, closure, eccentricity, and MNC—are selected sequentially. The software automatically calculates the specific values ​​of each target point under the four algorithms. The degree algorithm counts the number of nodes directly connected to the target point, the closure algorithm calculates the proportion of triangle structure in the local network where the target point is located, the eccentricity algorithm determines the shortest path length from the target point to the farthest node in the network, and the MNC algorithm counts the number of components in the target point's largest neighborhood. After calculation, the results are sorted in descending order and saved. In S24, targets ranking in the top 20% of the four sorting algorithms are extracted, duplicate targets are removed using Excel, and finally a core target set is formed and exported in TXT format. This step-by-step process ensures the comprehensiveness and reliability of the core targets through standardized data processing and multi-algorithm collaborative screening, providing high-quality target resources for subsequent molecular docking.

[0041] Preferred, such as Figure 3As shown, S3 includes the following steps: S31, log in to the RCSB database, input the standard gene symbol of the core target, screen crystal structure files with consistent species origin and meeting resolution requirements, and download the PDB format file; S32, access the PubChem database, search for the chemical name or CAS number of the active ingredient identified in the traditional Chinese medicine compound, select the compound entry with clear purity identification, and download its two-dimensional structure SDF format file; S33, verify the target structure of the downloaded PDB file, confirm the integrity of the active site, perform structural standardization processing on the SDF file, and remove redundant atoms and chemical bond information.

[0042] Specifically, step S3 involves the precise acquisition and standardization of structural files for core targets and active ingredients. Each step strictly controls structural quality to meet molecular docking requirements. In step S31, the RCSB database website is accessed. The standard gene symbol for each core target obtained in step S2 is entered into the search box. The search results page is then filtered for entries from human species, with a resolution no higher than 2.5 Å, no mutation modifications, and a crystal structure integrity of 95% or higher. Crystal structures including natural ligands are prioritized to clarify the location of active sites. After identifying the target entry, the download button is clicked, and a PDB format file is selected and saved locally. After downloading, the file is opened with a text editor to verify the completeness of key information such as atomic coordinates, amino acid sequences, and chemical bond connections. Files with missing data or incorrect formats are discarded to ensure the target structure files meet docking standards. S32 Access the PubChem database, enter the chemical name or CAS number of the key active ingredient screened in step S1 in the search bar, select compound entries with a purity of 98% or higher, no isomer confusion, and an authoritative source from the search results, click the "Download" button and select SDF format to download the two-dimensional structure file. This format can completely preserve the key information such as the molecular structure, functional group positions, and chemical bond types of the compound. After downloading, open the file with ChemDraw software to check whether there are redundant atoms in the molecular structure and whether the chemical bonds are connected correctly. In step S33, the target file in PDB format undergoes structural verification, focusing on checking the integrity of amino acid residues and the rationality of atomic coordinates in the active site region. PyMOL is used to examine the spatial conformation of the target to ensure a clear active pocket structure. The component file in SDF format undergoes structural standardization. ChemDraw software is used to remove redundant solvent molecules, adjust bond lengths and bond angles to standard states, and remove redundant charge information from the molecular structure. After processing, it is re-saved in SDF format. This step, through rigorous structural screening and standardization, provides qualified input data for molecular docking, ensuring the accuracy and reliability of the docking results.

[0043] Preferred, such as Figure 4As shown, step S4 includes the following steps: S41, opening the CB-DOCK2 database online platform, uploading the PDB format file of the core target, specifying the active site prediction region or using the automatic identification mode; S42, uploading the SDF format file of the active ingredient, setting the grid parameters for docking operation, the default parameters for the number of conformations searched, and submitting the docking task; S43, waiting for the operation to complete, downloading the docking result file, including binding energy data, target-component interaction list, and conformation file; S44, converting the format of the result file to ensure that the data is compatible with subsequent visualization tools.

[0044] Specifically, step S4 is implemented in steps that follow the complete molecular docking process, forming a standardized operation from file upload and parameter setting to result acquisition. Each step works closely together to achieve efficient docking. During S41, open the CB-DOCK2 database online platform, click the "TargetUpload" button, select the core target PDB format file obtained in step S3, and upload it. After the file is uploaded, the platform automatically parses the file and identifies potential active sites. At this point, you can select the "Auto-detect" automatic identification mode, or manually input the x, y, and z axis coordinate ranges of the active sites based on known target structure information. When setting manually, ensure that the coordinate range completely covers the active pocket region to avoid missing key binding sites. After setting, click "Next" to proceed to the component upload stage. In step S42, click the "LigandUpload" button to batch upload the SDF format files of the active ingredients processed in step S3. During the upload process, ensure the files are named correctly to avoid duplicate names or format errors that could cause upload failure. After the upload is complete, the platform automatically displays the ingredient name and structure preview. After verifying that everything is correct, proceed to the parameter settings page. Use the database's default optimal parameter configuration, where the grid box size is set to 60×60×60, the grid spacing is 0.375, the conformation search quantity is 100 different conformations generated for each ingredient, and the docking algorithm uses an improved version based on the Lamarckian genetic algorithm. This algorithm can efficiently search for the optimal binding conformation between the target and the ingredient. After confirming the parameters are correct, click "Submit" to submit the docking task. In step S43, the platform's background starts a parallel computing program to process multiple target-ingredient docking combinations, displaying the task progress bar and remaining time in real time. After all docking combinations are calculated, the platform generates a compressed result package including the binding energy data, interaction type, and conformation file for each target-ingredient combination. Click the download button to save it locally, and after decompression, view the detailed docking information for each combination. In S44, a format conversion tool is used to convert the decompressed conformational files from PDBQT format to PDB format to ensure compatibility with subsequent visualization tools. At the same time, the binding energy data is organized into an Excel spreadsheet, including information such as target name, component name, binding energy value, and interaction bond type. This step-by-step process achieves efficient docking between core targets and active ingredients through standardized platform operation and parameter settings, providing comprehensive data support for subsequent binding activity screening.

[0045] Preferred, such as Figure 5As shown, S5 includes the following steps: S51, extracting binding energy data from the docking results, comparing it with a preset threshold, and screening out target-component combinations that meet the conditions; S52, importing the screened conformational files into the PyMOL visualization tool, adjusting the perspective to display the binding mode of the target and the component, and marking hydrogen bonds and hydrophobic interactions to calibrate the interactions; S53, counting the number and type of interaction bonds for each target-component combination to form a quantitative analysis table; S54, integrating the binding energy data and interaction analysis results to generate a comprehensive molecular docking evaluation report.

[0046] Specifically, step S5 involves a multi-step implementation process, including binding energy screening, visualization analysis, quantitative statistics, and comprehensive evaluation, to fully analyze the target-component binding patterns. Each step progressively enhances the usability of the results. During S51, binding energy data for all target-component combinations are extracted from the docking results of step S4. A screening threshold of no more than -5 binding energies is set based on internationally recognized standards. Excel is used to compare the binding energy data with the threshold, automatically selecting highly active target-component combinations that meet the threshold requirements and eliminating low-activity combinations with binding energies higher than the threshold. Simultaneously, the specific binding energy value of each highly active combination is recorded, forming a preliminary screening result list, including key information such as target name, component name, and binding energy value. In S52, the conformational files corresponding to the screened high-activity combinations were imported into the PyMOL visualization tool. First, the view angle was adjusted to clearly show the binding region between the active site and the component. The tool's color-coding function was used to set the amino acid residues of the target to gray, the active site pocket to blue, and the component molecule to red. Then, the hydrogen bond labeling function was enabled to automatically identify and label the hydrogen bonds formed between the target and the component, clarifying the donor, acceptor, and bond length of the hydrogen bonds. Simultaneously, the locations of other interaction types, such as hydrophobic and electrostatic interactions, were marked. The image resolution was adjusted to 300 dpi and saved as a PNG file. In S53, based on the visualization results, the number and type of interaction bonds for each target-component combination were statistically analyzed. An Excel spreadsheet was used to record the specific number of hydrogen bonds, hydrophobic bonds, and electrostatic bonds, while also labeling the names of key interacting amino acid residues, forming a quantitative analysis table to ensure accurate recording of interaction information. In S54, binding energy data and interaction quantitative analysis results are integrated, and target-component combinations are arranged in descending order of binding energy value. Each combination is comprehensively evaluated based on the number and type of interaction bonds, generating a comprehensive evaluation report that includes binding energy, interaction details, and a link to the binding pattern diagram. This step-by-step process, which combines quantitative screening with visualization, ensures the accurate identification of highly active binding pairs and provides intuitive structural evidence for subsequent mechanism analysis, thereby enhancing the application value of molecular docking results.

[0047] This molecular docking analysis method, based on the RCSB database, integrates multiple databases to screen drug-disease-immune intersection targets. It uses protein-protein interaction network module analysis and multi-topology algorithm combinations to screen core targets, ensuring that they cover key action nodes and possess specificity. Standardized target and component structure files are obtained from authoritative databases, and professional docking tools and visualization analysis are used to accurately present target-component binding patterns. Molecular docking is organically linked with functional annotation and pathway enrichment analysis, forming a complete technical chain from target screening to mechanism analysis. The operation process is standardized and highly reproducible.

[0048] This method addresses the issue of insufficient targeting in traditional single-algorithm target selection. It employs multiple topological algorithms to collaboratively calculate and extract common targets, fully integrating multi-dimensional topological features to improve the comprehensiveness and reliability of core target selection, ensuring that the selection results truly reflect the interaction relationship among the three elements. Furthermore, to address the shortcomings of existing single-factor docking evaluation systems, it constructs a multi-factor comprehensive binding energy prediction and stability evaluation system. By combining key indicators such as interaction bond type and conformational stability, it supplements the limitations of single binding energy evaluation, reduces the omission of potential effective combinations, and enhances the consistency and reliability of the analysis results through standardized structural processing and parameter settings.

[0049] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," "link," and "fix" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0050] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various equivalent changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A molecular docking analysis method for obtaining core target points based on the RCSB database, characterized in that, Includes the following steps: S1. Potential targets corresponding to the blood-entering active components of two or more medicinal materials in traditional Chinese medicine compound prescriptions are screened using TCMSP, ETCM, and HERB databases. Simultaneously, related targets for the target disease and immune-related diseases are retrieved using GeneCards, DisGeNET, and DrugBank databases. Venny tool is used for target mapping to obtain drug-disease-immunity intersection targets. S2. The intersection targets are imported into the STRING database, and a protein-protein interaction network is constructed by setting confidence thresholds. After exporting the network data, module analysis is performed using the MCODE plugin of Cytoscape software. Core targets are screened using the CytoHubba plugin based on four topology algorithms: degree value, closure, eccentricity, and MNC. S3. S4. Obtain the PDB format files corresponding to the core targets screened from the RCSB database, and download the two-dimensional structure files of the active ingredients in the traditional Chinese medicine compound from the PubChem database; S5. Perform molecular docking processing on the PDB format files of the core targets and the two-dimensional structure files of the active ingredients using the CB-DOCK2 database, and complete the docking operation using the default parameters of the database; S6. Use the binding energy value as the docking effect evaluation index to screen out target-component binding pairs whose binding energies meet the preset threshold, and generate a binding pattern diagram using a molecular docking visualization tool; S7. Use the DAVID database to perform GO function annotation and KEGG pathway enrichment analysis on the core targets screened out, set the criteria for judging the significance of differences, and output the enrichment results.

2. The molecular docking analysis method for obtaining core target points based on the RCSB database according to claim 1, characterized in that, The model formula used in S2 for selecting core target points by the topology algorithm is as follows: ,in, For the comprehensive score of the target, These are the weighting coefficients for each topology parameter. The target point value. The target closure coefficient. For the target eccentricity parameter, This represents the maximum number of neighborhood components of the target.

3. The molecular docking analysis method for obtaining core target points based on the RCSB database according to claim 1, characterized in that, The binding energy prediction expression for the molecular docking process in S4 is as follows: ,in, For binding energy, For van der Waals action energy weights, For van der Waals action energy, The electrostatic interaction energy coefficient is... It is the energy of electrostatic interaction. The hydrogen bond interaction energy weights, The hydrogen bond interaction energy, The solvation energy coefficient, This is the solvation entropy value.

4. The molecular docking analysis method for obtaining core target points based on the RCSB database according to claim 1, characterized in that, The expression for determining the significance of KEGG pathway enrichment in S6 is as follows: ,in, The P-value represents the significance of the enrichment difference. This represents the total number of pathway-related targets in the database. The total number of core targets, To enrich the number of core targets in the target pathway, The total number of targets included in the target pathway. This is the pathway enrichment correction coefficient.

5. The molecular docking analysis method for obtaining core target points based on the RCSB database according to claim 1, characterized in that, The intersection target selection expression in S1 is: ,in, To screen confidence levels for intersection targets, It is a collection of drug active ingredient targets. A set of disease-related targets, A collection of immune-related targets, The intersection weight coefficients of the sets To filter deviation correction values.

6. The molecular docking analysis method for obtaining core target points based on the RCSB database according to claim 1, characterized in that, The expression for evaluating the target-component binding stability in S5 is as follows: ,in, To incorporate stability scoring, The coefficient of contribution to binding energy. For binding energy, This represents the number of interaction bonds between the target and the component. For conformational stability weights, To incorporate conformational deviation rate.

7. The molecular docking analysis method for obtaining core target points based on the RCSB database according to claim 1, characterized in that, S2 includes the following steps: S21, after standardizing the gene names of the intersection target points, import them into the STRING database, set the confidence parameters and select the species matching option, and submit to obtain the interaction data between target points; S22, convert the exported interaction data into a Cytoscape compatible format, import it into the software, and then use the MCODE plugin to set the node degree threshold and clustering coefficient threshold for module division; S23, start the CytoHubba plugin, and select four algorithms in sequence: degree value, closure, eccentricity, and MNC, to calculate and sort the topological parameter values ​​of each target point; S24, extract the common target points in the sorting results of the four algorithms, remove duplicates, form a core target point set, and output it.

8. The molecular docking analysis method for obtaining core target points based on the RCSB database according to claim 1, characterized in that, S3 includes the following steps: S31, log in to the RCSB database, input the standard gene symbol of the core target, screen crystal structure files with consistent species origin and meeting resolution requirements, and download the PDB format file; S32, access the PubChem database, search for the chemical name or CAS number of the identified active ingredient in the traditional Chinese medicine compound, select the compound entry with clear purity identification, and download its two-dimensional structure SDF format file; S33, verify the target structure of the downloaded PDB file, confirm the integrity of the active site, perform structural standardization processing on the SDF file, and remove redundant atoms and chemical bond information.

9. The molecular docking analysis method for obtaining core target points based on the RCSB database according to claim 1, characterized in that, S4 includes the following steps: S41, open the CB-DOCK2 database online platform, upload the PDB format file of the core target, specify the active site prediction region or use the automatic identification mode; S42, upload the SDF format file of the active ingredient, set the grid parameters and default parameters for the docking operation and the number of conformations searched, and submit the docking task; S43, wait for the operation to complete, download the docking result file, including binding energy data, target-component interaction list and conformation file; S44, convert the format of the result file to ensure that the data is compatible with subsequent visualization tools.

10. The molecular docking analysis method for obtaining core target points based on the RCSB database according to claim 1, characterized in that, S5 includes the following steps: S51, extracting binding energy data from the docking results, comparing it with a preset threshold, and screening out target-component combinations that meet the conditions; S52, importing the screened conformational files into the PyMOL visualization tool, adjusting the perspective to display the binding mode of the target and component, and marking hydrogen bonds and hydrophobic interactions to identify interactions; S53, counting the number and type of interaction bonds for each target-component combination to form a quantitative analysis table; S54, integrating the binding energy data and interaction analysis results to generate a comprehensive molecular docking evaluation report.

Images