A method and system for screening compounds that modulate blood-brain barrier permeability

Abstract

The application discloses a kind of compound screening method and system for regulating blood-brain barrier permeability, it is related to medical screening technical field, the specific steps of the method are as follows: S100, collect blood-brain barrier related cell multi-omics data to form original set;S200, data is preprocessed and standardized;S300, molecular network is constructed by cross-omics analysis, and key target is screened;S400, target information is integrated to establish structured target pool;S500, effective compound for regulating blood-brain barrier permeability is screened out by molecular docking and primary cell verification, the application constructs gene-protein-metabolite network by multi-omics integration and cross-omics analysis, accurately identifies blood-brain barrier key target, provides high credibility target pool;Combined with molecular docking and primary cell mixed culture verification, realize efficient screening and functional safety evaluation of candidate compound, significantly reduce research and development cost, provide efficient technical path for drug research and development.

Description

Technical Field

[0001] This invention relates to the field of pharmaceutical screening technology, specifically to a method and system for screening compounds that regulate the permeability of the blood-brain barrier. Background Technology

[0002] The blood-brain barrier (BBB), a crucial protective structure of the central nervous system, maintains brain microenvironment homeostasis through strict regulation of substance exchange. However, its tight structure also significantly hinders the penetration of most therapeutic drugs into brain tissue, becoming a core bottleneck in the development of drugs for neurological diseases. In recent years, with the rapid development of omics technologies, genomics, transcriptomics, proteomics, and metabolomics data have provided a wealth of information for elucidating the dynamic regulatory mechanisms of the BBB. How to systematically integrate multi-omics data, reveal the interaction network between genes, proteins, and metabolites, and identify key regulatory targets has become a key direction for overcoming the technological bottleneck in drug development related to the BBB.

[0003] Traditional research on the regulation of blood-brain barrier permeability mainly relies on single-mathematical analysis or empirical screening, which has significant limitations. Single-mathematical methods are difficult to fully reflect the complexity of biological systems and easily overlook cross-level regulatory relationships, resulting in low accuracy of target prediction. Empirical screening, lacking guidance from molecular mechanisms, requires extensive experimental verification, which is time-consuming, costly, and has a low success rate. In addition, traditional cell models mostly use single cell types and cannot simulate the multi-cell interactions of the blood-brain barrier in vivo, leading to a disconnect between compound verification results and actual in vivo effects. These problems seriously restrict the development efficiency of drugs that regulate the blood-brain barrier. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method and system for screening compounds that regulate the permeability of the blood-brain barrier. By systematically collecting genomic, transcriptomic, proteomic, and metabolomic data, constructing a gene-protein-metabolite interaction network, accurately identifying key regulatory targets, and establishing a structured target pool, combined with molecular docking technology and a primary cell co-culture verification system, efficient screening and functional evaluation of candidate compounds can be achieved.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: On one hand, a method for screening compounds that regulate blood-brain barrier permeability, the specific steps of which are as follows:

[0006] S100, Multi-omics Data Acquisition: Collect genomic, transcriptomic, proteomic, and metabolomics data of blood-brain barrier-related cells in normal and specific disease states to form a multi-dimensional raw dataset;

[0007] S200, Data Standardization Processing: Preprocess the collected multi-omics data, remove outliers and technical errors through data cleaning, and standardize the data to generate a standardized dataset;

[0008] S300, Molecular Regulatory Network Analysis: Based on standardized data, cross-omics association analysis is performed to construct a gene-protein-metabolite interaction network, and the node core degree is calculated by the network core degree algorithm to screen out key targets that regulate blood-brain barrier permeability.

[0009] S400, Precision Target Pool Construction: The key targets obtained from the analysis are classified and integrated according to their functions to establish a structured target pool containing information on target molecular properties, interaction relationships and regulatory pathways.

[0010] S500, multi-stage compound screening: Based on the target pool, candidate compounds with good binding ability to key targets are screened through molecular docking; then, the compounds are validated using primary isolated blood-brain barrier-related cell populations to screen out effective compounds.

[0011] Furthermore, in S100, during the multi-omics data acquisition, the blood-brain barrier-related cells include: brain microvascular endothelial cells, astrocytes, pericytes, and microglia; specific disease states include: stroke, Alzheimer's disease, and glioma; and data are collected in the early, middle, and late stages of the disease.

[0012] Furthermore, in S100, during multi-omics data acquisition, the acquisition of genomic data includes: quality control of the raw reads obtained from sequencing, filtering low-quality reads, removing adapter sequences and repetitive sequences, comparing the cleaned reads with the reference genome, and filtering for false positive variants; the acquisition of transcriptomics data includes: capturing the expression profiles of genes and transcripts using RNA sequencing and single-cell RNA sequencing technologies; the acquisition of proteomics data includes: detecting protein expression levels and post-translational modification information using liquid chromatography-tandem mass spectrometry and protein chip technology; and the acquisition of metabolomics data includes: analyzing the concentration changes of small molecule metabolites using gas chromatography-mass spectrometry and ultra-high performance liquid chromatography-mass spectrometry.

[0013] Furthermore, in S100, during the multi-omics data acquisition, the preprocessing of multi-omics data includes: filtering low-quality sequencing reads, removing adapter sequences and repetitive sequences from genomics data, aligning with a reference genome and filtering for false positives; removing adapter contamination and low-quality sequences from transcriptomics data, performing transcript quantification and batch effect correction, and eliminating low-expression genes; identifying and quantifying peptides from proteomics data, filtering proteins with excessively high missing values ​​and filling in the missing values; and identifying, matching, and quantifying peaks from metabolomics data, removing metabolite peaks with low signal-to-noise ratios and processing missing values.

[0014] Furthermore, in S300, the specific steps for constructing gene-protein-metabolite interactions in molecular regulatory network analysis are as follows: extracting gene variant sites and their frequencies from genomics data; extracting gene and RNA expression levels from transcriptomics data; extracting protein expression levels and modification information from proteomics data; and extracting metabolite concentration values ​​from metabolomics data; analyzing linkage relationships between gene variant sites at the genomics level; analyzing co-expression relationships of gene and RNA expression levels at the transcriptomics level; constructing a protein interaction network based on known protein interaction relationships at the proteomics level; and analyzing upstream and downstream transformation relationships between metabolites at the metabolomics level; calculating the correlation between gene variants and gene expression levels, the correlation between gene expression levels and protein expression levels, the correlation between protein expression levels and metabolite concentrations, and the indirect correlation between gene expression levels and metabolite concentrations using a multi-omics association strength algorithm; and using genes, proteins, and metabolites as network nodes, and significant association relationships obtained from intra-omics and cross-omics analyses as connecting edges between nodes, assigning corresponding association strength values ​​to the edges to form a gene-protein-metabolite interaction network.

[0015] Furthermore, in S300, the calculation formula for the multi-omics association strength algorithm in molecular regulatory network analysis is as follows: ,in: Molecular variables in different omics datasets and The strength of multi-omics associations between them No. In each sample The corresponding molecular variable values, It is the first In each sample The corresponding molecular variable values, These are weighting coefficients set based on prior biological knowledge, with values ​​ranging from 0.1 to 1.0. yes The average value of the corresponding molecular variables, yes The average value of the corresponding molecular variables, It refers to the number of samples.

[0016] Furthermore, in the S300 molecular regulatory network analysis, the calculation formula for the network core degree algorithm is as follows: ,in, It is a node The core of, It is the maximum degree value among all nodes. For nodes The degree value, It is the maximum value of the centrality of the number of intermediate nodes. For nodes betweenness centrality, This represents the multiple of difference between the diseased and normal states at this node. , , These are the weighting coefficients, and .

[0017] Furthermore, in S500, the specific steps for screening candidate compounds with good binding ability to key targets through molecular docking based on the target pool in the multi-stage compound screening are as follows: Key protein targets with well-defined three-dimensional structures are selected from the target pool; their crystal structures are obtained from a protein database and preprocessed, including removing water of crystallization, adding Gasteiger charge, and adding hydrogen; the two-dimensional structures of the compounds are extracted from the candidate compound library and converted into three-dimensional structures, and energy minimization is performed; molecular docking software is used to perform docking calculations between the compounds and the active pockets of the target proteins to obtain the binding energy and structural characteristic parameters of the hydrogen bond interaction sites; quantitative scoring is performed based on the target-compound matching algorithm, the calculation formula of which is: ,in, It is a quantitative score of target-compound matching. The structure matching score between the compound and the target is calculated. The score is based on the consistency between the compound's function and the direction of target regulation. To determine the synergistic score of compound-target pairs in the regulatory pathway, score-based screening is performed. Compounds that exhibit specific interactions with active sites are considered candidate compounds with good binding ability to key targets.

[0018] Furthermore, in S500, the specific steps for compound validation using primary isolated blood-brain barrier-related cell populations during multi-stage compound screening are as follows: Brain microvascular endothelial cells, astrocytes, pericytes, and microglia are isolated and purified from normal and specific disease model organisms, and a mixed cell culture system is constructed according to in vivo physiological proportions; candidate compounds are added to this system, and barrier integrity is assessed by detecting changes in transcellular resistance; a fluorescently labeled small molecule tracer is used to detect substance permeation efficiency, reflecting the permeability regulation effect; simultaneously, cell viability is measured using a cell viability assay kit, and compound toxicity is assessed by combining lactate dehydrogenase release; compounds that can regulate transcellular resistance and tracer permeation efficiency, have a cell viability ≥80%, and lactate dehydrogenase release ≤1.5 times that of the normal control group are screened as effective compounds.

[0019] On the other hand, a compound screening system for regulating blood-brain barrier permeability includes: a data acquisition module, a data processing module, a network analysis module, a target pool construction module, and a compound screening module;

[0020] The data acquisition module collects genomic, transcriptomic, proteomic, and metabolomics data of blood-brain barrier-related cells under normal and specific disease states, forming a multi-dimensional raw dataset.

[0021] The data processing module preprocesses the collected multi-omics data, removes outliers and technical errors through data cleaning, and standardizes the data to generate a standardized dataset.

[0022] The network parsing module performs cross-omics association analysis based on standardized data, constructs a gene-protein-metabolite interaction network, calculates node core degree through a network core degree algorithm, and screens out key targets that regulate blood-brain barrier permeability.

[0023] The target pool construction module integrates the key targets obtained from the analysis according to their functions, and establishes a structured target pool containing information on target molecular properties, interaction relationships and regulatory pathways.

[0024] The compound screening module: Based on the target pool, candidate compounds with good binding ability to key targets are screened through molecular docking; then, the compounds are verified using the primary isolated blood-brain barrier-related cell population to screen out effective compounds.

[0025] Compared with existing technologies, this method and system for screening compounds that regulate blood-brain barrier permeability has the following advantages:

[0026] I. This invention overcomes the limitations of traditional single-omics research by integrating multi-omics data and cross-omics association analysis. It collects genomic, transcriptomic, proteomic, and metabolomic data, covering the dynamic changes of blood-brain barrier-related cells under normal and disease states, and constructs a gene-protein-metabolite interaction network. Using a network core degree algorithm, it can accurately identify key targets that regulate blood-brain barrier permeability, avoiding false positive results caused by noise in single-omics data. This multi-dimensional and cross-level analysis strategy significantly improves the accuracy and biological rationality of target discovery, provides a highly reliable structured target pool for subsequent drug screening, and improves the efficiency and success rate of compound development.

[0027] Second, this invention utilizes molecular docking technology to rapidly screen candidate molecules that specifically bind to key targets from a massive compound library, significantly narrowing the scope of experimental verification. Furthermore, it employs a hybrid culture system constructed from primary isolated blood-brain barrier-related cells to simulate the in vivo physiological environment. The efficacy of compounds is comprehensively evaluated using multiple indicators such as transcellular electrical resistance, substance permeability, and cytotoxicity. This phased screening strategy combining dry and wet experiments ensures both the structural matching of candidate compounds and verifies their functional activity and safety, significantly reducing the high trial-and-error costs in traditional drug development and providing an efficient technical path for the development of innovative drugs that regulate blood-brain barrier permeability.

[0028] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0029] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0030] Figure 1 A flowchart illustrating the steps of a method for screening compounds that regulate blood-brain barrier permeability;

[0031] Figure 2 A flowchart of a compound screening system module for regulating blood-brain barrier permeability;

[0032] Figure 3 A graph showing the output relationships at each stage of screening for a compound that regulates blood-brain barrier permeability. Detailed Implementation

[0033] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structure, features and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0034] Example 1:

[0035] Based on the method of this invention, compounds that regulate blood-brain barrier permeability in stroke patients are screened. Specific steps are as follows: Figure 1 As shown.

[0036] Multi-omics data acquisition: Genomic, transcriptomic, proteomic, and metabolomic data were collected from brain tissue of normal organisms and model organisms in the early, middle, and late stages of Alzheimer's disease. The data were obtained by enzymatic digestion followed by density gradient centrifugation and immunomagnetic bead sorting of brain microvascular endothelial cells, astrocytes, pericytes, and microglia, forming a multi-dimensional raw dataset. Genomic data, obtained through quality control of the raw sequencing data, filtering low-quality fragments, and removing adapters and repetitive sequences, reflects gene-level variation information. Transcriptomic data, using RNA sequencing and single-cell RNA sequencing technologies, captures gene and transcript expression profiles, revealing gene expression activity. Proteomic data, using liquid chromatography-tandem mass spectrometry and protein chip technology, detects protein expression levels and post-translational modifications, showcasing protein functional states. Metabolomic data, analyzed using gas chromatography-mass spectrometry and ultra-high performance liquid chromatography-mass spectrometry, analyzes small molecule metabolite concentration changes, reflecting dynamic changes at the metabolic level. These data collectively provide comprehensive molecular-level information for subsequent analysis of regulatory networks, such as… Figure 3 As shown.

[0037] Data standardization: The collected multi-omics data undergoes preprocessing, including data cleaning to remove outliers and technical errors. For example, genomics data is further filtered for low-quality sequencing fragments, residual adapters and repetitive sequences are removed, and the data is compared with a reference genome with false positive filtering to improve the accuracy of gene variation information. Transcriptomics data is cleaned to remove adapter contamination and low-quality sequences, and transcript quantification and batch effect correction are performed to eliminate low-expression genes, ensuring the reliability and comparability of gene expression data. Proteomics data undergoes peptide identification and quantification, filtering proteins with excessively high missing values ​​and filling in missing values ​​to make protein expression and modification data more stable. Metabolomics data undergoes peak identification, matching, and quantification, removing low signal-to-noise ratio metabolite peaks and processing missing values ​​to enhance the effectiveness of metabolite concentration data. These processes generate standardized datasets, laying a unified data foundation for cross-omics association analysis.

[0038] Molecular Regulatory Network Analysis: Based on standardized data, cross-omics association analysis was conducted to construct a gene-protein-metabolite interaction network. First, gene variant sites and their frequencies were extracted from genomic data; gene and RNA expression levels were extracted from transcriptomics data; protein expression levels and modification information were extracted from proteomics data; and metabolite concentrations were extracted from metabolomics data. Next, linkage relationships between gene variant sites were analyzed at the genomic level; co-expression relationships of genes and RNA were analyzed at the transcriptomics level; a protein-protein interaction network was constructed based on known protein-protein interactions at the proteomics level; and upstream and downstream transformation relationships between metabolites were analyzed at the metabolomics level. Finally, the association strength between molecules from different omics was calculated using a multi-omics association strength algorithm. The formula for the multi-omics association strength algorithm is as follows: ,in: Molecular variables in different omics datasets and The strength of multi-omics associations between them No. In each sample The corresponding molecular variable values, It is the first In each sample The corresponding molecular variable values, These are weighting coefficients set based on prior biological knowledge, with values ​​ranging from 0.1 to 1.0. yes The average value of the corresponding molecular variables, yes The average value of the corresponding molecular variables, The sample size includes the indirect correlations between gene variation and gene expression level, gene expression level and protein expression level, protein expression level and metabolite concentration, and gene expression level and metabolite concentration. Finally, genes, proteins, and metabolites are used as network nodes, and significant associations obtained from intra- and cross-omics analyses are used as connecting edges between nodes, assigning corresponding association strength values ​​to the edges to form a gene-protein-metabolite interaction network. Subsequently, the node core degree is calculated using a network core degree algorithm. The formula for the network core degree algorithm is: ,in, It is a node The core of, It is the maximum degree value among all nodes. For nodes The degree value, It is the maximum value of the centrality of the number of intermediate nodes. For nodes betweenness centrality, This represents the multiple of difference between the diseased and normal states at this node. , , These are the weighting coefficients, and The study identified key targets that occupy crucial positions in the network and play an important role in regulating blood-brain barrier permeability. These targets can accurately reflect the core molecules in the regulatory process.

[0039] Construction of a precise target pool: The key targets obtained from the analysis are classified and integrated according to their functions, such as participating in different functional categories such as barrier structure maintenance, signal transduction, and metabolic regulation. A structured target pool is established that includes target molecular attributes (such as molecular type and structural features), interaction relationships (such as binding or regulatory relationships with other molecules), and regulatory pathway information (such as signaling pathways or metabolic pathways involved). This target pool can systematically sort out various types of information about key targets, providing clear and precise references for subsequent compound screening, and improving the targeting of screening.

[0040] Multi-stage compound screening: Compound screening is based on a target pool. First, key protein targets with well-defined three-dimensional structures are selected. Their crystal structures are obtained from a protein database and preprocessed, including removing water of crystallization, adding charge, and adding hydrogen. Two-dimensional structures of compounds are extracted from the candidate compound library and converted to three-dimensional structures, followed by energy minimization. Molecular docking software is used to perform docking calculations between the compounds and the active pockets of the target proteins, obtaining structural characteristic parameters such as binding energy and hydrogen bond interaction sites. Quantitative scoring is performed using a target-compound matching algorithm. The calculation formula for the target-compound matching algorithm is as follows: ,in, It is a quantitative score of target-compound matching. The structure matching score between the compound and the target is calculated. The score is based on the consistency between the compound's function and the direction of target regulation. The score represents the synergistic effect of the compound-target pair in the regulatory pathway. This score comprehensively considers the structural matching, functional and regulatory direction consistency, and synergistic effect of the compound and target in the regulatory pathway, and selects the best candidates based on these factors. Compounds exhibiting specific interactions with active sites were selected as candidate compounds with good binding affinity to key targets, demonstrating their potential for effective binding. Next, primary isolated blood-brain barrier-related cell populations were used for compound validation. Brain microvascular endothelial cells, astrocytes, pericytes, and microglia were isolated and purified from normal and stroke model organisms, and a mixed cell culture system was constructed according to in vivo physiological proportions to simulate the cellular microenvironment of the blood-brain barrier. Candidate compounds were added to this system, and barrier integrity was assessed by detecting changes in transcellular electrical resistance. Fluorescently labeled small-molecule tracers were used to detect substance permeation efficiency to reflect permeability regulation. Simultaneously, cell viability was measured using a cell viability assay kit, and compound toxicity was assessed by combining lactate dehydrogenase release levels. Finally, compounds that could regulate transcellular electrical resistance and tracer permeation efficiency, with cell viability ≥80% and lactate dehydrogenase release levels ≤1.5 times that of the normal control group, were selected as effective compounds. These compounds effectively regulated blood-brain barrier permeability and exhibited good safety.

[0041] In summary, this embodiment comprehensively acquires gene, transcription, protein, and metabolic information of blood-brain barrier-related cells in stroke state through multi-omics data collection, providing a foundation for elucidating the regulatory mechanism. Data errors are eliminated through standardization processing to ensure the reliability of the analysis. Key targets are accurately screened by using cross-omics association analysis and network core degree algorithms. The constructed target pool clearly defines the target objects. Multi-stage screening combined with molecular docking and cell experiments not only ensures the binding potential of compounds to targets but also verifies their effectiveness and safety in simulated in vivo environments. Ultimately, compounds that can regulate blood-brain barrier permeability in stroke state are efficiently screened.

[0042] Example 2:

[0043] Based on the present invention, compounds that regulate blood-brain barrier permeability in Alzheimer's disease are systematically screened. Specific steps are as follows: Figure 2 As shown.

[0044] The data acquisition module works as follows: From the brain tissue of normal organisms and model organisms in the early, middle, and late stages of Alzheimer's disease, genomic, transcriptomic, proteomic, and metabolomic data of brain microvascular endothelial cells, astrocytes, pericytes, and microglia are collected using enzymatic digestion, combined with density gradient centrifugation and immunomagnetic bead sorting. This forms a multi-dimensional raw dataset. Genomic data covers gene variant sites and their frequencies, revealing the potential regulatory basis at the gene level; transcriptomic data includes gene and transcript expression profiles, reflecting gene activation or repression states; proteomic data involves protein expression levels and post-translational modifications, demonstrating protein functional activity; and metabolomic data includes concentration changes of small molecule metabolites, showcasing dynamic metabolic responses. These data collectively provide comprehensive molecular information support for subsequent analysis of the blood-brain barrier permeability regulation mechanisms, ensuring sufficient and diverse raw materials for subsequent analysis.

[0045] The data processing module receives multi-omics data from the data acquisition module, preprocesses it, removes outliers and technical errors through data cleaning, and implements targeted processing for different omics data: For genomics data, it filters low-quality sequencing fragments, removes adapters and repetitive sequences, aligns with a reference genome, and filters for false positives to improve the accuracy of gene variation information; for transcriptomics data, it removes adapter contamination and low-quality sequences, performs transcript quantification and batch effect correction, and eliminates low-expression genes to ensure the stability of gene expression data and the comparability between different groups; for proteomics data, it identifies and quantifies peptides, filters proteins with excessively high missing values ​​and fills in missing values ​​to make protein-related data more reliable; for metabolomics data, it identifies, matches, and quantifies peaks, removes low signal-to-noise ratio metabolite peaks, and processes missing values ​​to enhance the effectiveness of metabolite concentration data. Finally, it generates a standardized dataset, establishing a unified and standardized data platform for cross-omics association analysis and reducing the interference of data differences on subsequent analyses.

[0046] The network parsing module performs cross-omics association analysis based on the standardized data output from the data processing module. First, it extracts key information from each omics dataset, such as gene variant sites in genomics, gene expression levels in transcriptomics, protein expression and modification information in proteomics, and metabolite concentrations in metabolomics. Then, it analyzes the relationships between molecules within each omics, including linkage relationships at the genomic level, co-expression relationships at the transcriptomic level, protein-protein interactions at the proteomic level, and upstream and downstream metabolite transformation relationships at the metabolomics level. Finally, it calculates the association strength between molecules from different omics using a multi-omics association strength algorithm. The formula for the multi-omics association strength algorithm is as follows: The interaction degree between genes, proteins, and metabolites is clarified. Then, genes, proteins, and metabolites are treated as network nodes, and significant correlations are used as connecting edges to form a gene-protein-metabolite interaction network. Next, the core degree of each node is calculated using a network core degree algorithm. The formula for calculating the network core degree algorithm is as follows: The goal is to identify targets that occupy a core position in the network and play a key role in regulating the permeability of the blood-brain barrier. These targets can reflect the core molecular mechanisms in the regulatory process, providing a precise direction for subsequent screening.

[0047] The target pool construction module works by receiving key targets screened by the network analysis module, classifying and integrating them according to their functions, such as participating in barrier structure formation, signal transduction, and substance transport. It establishes a structured target pool that includes target molecular attributes (such as molecular structure and type), interaction relationships (such as binding or regulatory modes with other molecules), and regulatory pathway information (such as signaling pathways or metabolic pathways involved). This target pool system integrates various types of information on key targets, clearly presenting the characteristics and correlations of targets, providing clear and specific references for the targets of compound screening, and improving the accuracy and efficiency of screening.

[0048] The compound screening module works as follows: First, molecular docking screening is performed based on the target pool to select key protein targets with well-defined three-dimensional structures. Their crystal structures are obtained from a protein database and pre-processed to remove water of crystallization, add charge, and add hydrogen. The two-dimensional structures of compounds are extracted from the candidate compound library and converted into three-dimensional structures, followed by energy minimization. Molecular docking software is used to dock the compounds with the active pockets of the target proteins to obtain parameters such as binding energy and hydrogen bond interaction sites. Finally, a target-compound matching algorithm is used for quantitative scoring. The calculation formula for the target-compound matching algorithm is as follows: Scoring criteria were determined by comprehensively considering structural matching, functional consistency, and pathway synergy. Compounds exhibiting specific interactions with active sites are considered candidate compounds with good binding affinity to key targets, demonstrating their potential for effective binding. A mixed culture system was constructed using primary isolated blood-brain barrier-related cells, incorporating brain microvascular endothelial cells, astrocytes, pericytes, and microglia in physiological proportions to simulate the in vivo barrier microenvironment. Candidate compounds were added to the system, and changes in transcellular electrical resistance were measured to assess barrier integrity. Fluorescently labeled small-molecule tracers were used to detect permeability regulation efficiency. Cell viability and lactate dehydrogenase release were simultaneously assessed for toxicity. Finally, compounds that could regulate transcellular electrical resistance and tracer permeability efficiency, with cell viability ≥80% and lactate dehydrogenase release ≤1.5 times that of the normal control group, were selected as effective compounds. These compounds can effectively and safely regulate blood-brain barrier permeability in Alzheimer's disease.

[0049] In summary, this system integrates multi-omics data related to Alzheimer's disease through a data acquisition module. After standardization by the processing module, a network analysis module constructs a regulatory network and screens key targets. A target pool module sorts out target information. The screening module first screens candidate compounds through molecular docking and then uses a cell system to verify their regulatory effects and toxicity. All modules work together to form a complete process from data integration to compound verification, efficiently screening effective compounds that can safely regulate the permeability of the blood-brain barrier in Alzheimer's disease.

[0050] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for screening compounds that regulate blood-brain barrier permeability, characterized in that, The specific steps of this method are as follows: S100, Multi-omics data collection: Collect genomic, transcriptomic, proteomic, and metabolomics data of blood-brain barrier-related cells in normal and specific disease states. The specific disease states include stroke, Alzheimer's disease, and glioma. Data collection is carried out in the early, middle, and late stages of the disease to form a multi-dimensional raw dataset. S200, Data Standardization Processing: Preprocess the collected multi-omics data, remove outliers and technical errors through data cleaning, and standardize the data to generate a standardized dataset; S300, Molecular Regulatory Network Analysis: Based on standardized data, cross-omics association analysis is performed to construct a gene-protein-metabolite interaction network, and the node core degree is calculated by the network core degree algorithm to screen out key targets that regulate blood-brain barrier permeability. S400, Precision Target Pool Construction: The key targets obtained from the analysis are classified and integrated according to their functions to establish a structured target pool containing information on target molecular properties, interaction relationships and regulatory pathways. S500, multi-stage compound screening: Based on the target pool, candidate compounds with good binding ability to key targets are screened through molecular docking; then, the compounds are validated using primary isolated blood-brain barrier-related cell populations to screen out effective compounds.

2. The method for screening compounds that regulate blood-brain barrier permeability according to claim 1, characterized in that, In the S100 multi-omics data acquisition, blood-brain barrier-related cells include: brain microvascular endothelial cells, astrocytes, pericytes, and microglia.

3. The method for screening compounds that regulate blood-brain barrier permeability according to claim 1, characterized in that, In S100, the acquisition of multi-omics data includes: acquiring genomic data by performing quality control on the raw reads obtained from sequencing, filtering low-quality reads, removing adapter sequences and repetitive sequences, comparing the cleaned reads with the reference genome, and filtering for false positive variants; acquiring transcriptomics data by capturing the expression profiles of genes and transcripts using RNA sequencing and single-cell RNA sequencing technologies; acquiring proteomics data by detecting protein expression levels and post-translational modification information using liquid chromatography-tandem mass spectrometry and protein chip technology; and acquiring metabolomics data by analyzing the concentration changes of small molecule metabolites using gas chromatography-mass spectrometry and ultra-high performance liquid chromatography-mass spectrometry.

4. The method for screening compounds that regulate blood-brain barrier permeability according to claim 1, characterized in that, In S100, during the multi-omics data acquisition, the preprocessing of multi-omics data includes: filtering low-quality sequencing reads, removing adapter and repetitive sequences from genomics data, aligning with a reference genome and filtering for false positives; removing adapter contamination and low-quality sequences from transcriptomics data, performing transcript quantification and batch effect correction, and removing low-expression genes; identifying and quantifying peptides from proteomics data, filtering proteins with excessively high missing values ​​and filling in the missing values; and identifying, matching, and quantifying peaks from metabolomics data, removing metabolite peaks with low signal-to-noise ratios and processing missing values.

5. The method for screening compounds that regulate blood-brain barrier permeability according to claim 1, characterized in that, In S300, the specific steps for constructing gene-protein-metabolite interactions in molecular regulatory network analysis are as follows: Gene variant sites and their frequencies are extracted from genomic data; gene and RNA expression levels are extracted from transcriptomics data; protein expression levels and modification information are extracted from proteomics data; and metabolite concentration values ​​are extracted from metabolomics data. At the genomic level, linkage relationships between gene variant sites are analyzed; at the transcriptomics level, co-expression relationships of gene and RNA expression levels are analyzed; at the proteomics level, a protein-protein interaction network is constructed based on known protein-protein interaction relationships; and at the metabolomics level, upstream and downstream transformation relationships between metabolites are analyzed. The correlation between gene variants and gene expression levels, the correlation between gene expression levels and protein expression levels, the correlation between protein expression levels and metabolite concentrations, and the indirect correlation between gene expression levels and metabolite concentrations are calculated using a multi-omics correlation strength algorithm. Genes, proteins, and metabolites are used as network nodes, and significant correlations obtained from intra-omics and cross-omics analyses are used as connecting edges between nodes, assigning corresponding correlation strength values ​​to the edges to form a gene-protein-metabolite interaction network.

6. The method for screening compounds that regulate blood-brain barrier permeability according to claim 5, characterized in that, In the S300 molecular regulatory network analysis, the calculation formula for the multi-omics association strength algorithm is as follows: ， in: Molecular variables in different omics datasets and The strength of multi-omics associations between them No. In each sample The corresponding molecular variable values, It is the first In each sample The corresponding molecular variable values, These are weighting coefficients set based on prior biological knowledge, with values ​​ranging from 0.1 to 1.

0. yes The average value of the corresponding molecular variables, yes The average value of the corresponding molecular variables, It refers to the number of samples.

7. The method for screening compounds that regulate blood-brain barrier permeability according to claim 1, characterized in that, In the S300 molecular regulatory network analysis, the calculation formula for the network core degree algorithm is as follows: ， in, It is a node The core of, It is the maximum degree value among all nodes. For nodes The degree value, It is the maximum value of the centrality of the number of intermediate nodes. For nodes betweenness centrality, This represents the multiple of difference between the diseased and normal states at this node. , , These are the weighting coefficients, and .

8. The method for screening compounds that regulate blood-brain barrier permeability according to claim 1, characterized in that, In the S500 compound multi-stage screening, the specific steps for screening candidate compounds with good binding ability to key targets through molecular docking based on the target pool are as follows: select key protein targets with clear three-dimensional structures from the target pool, obtain their crystal structures through protein databases and perform preprocessing, including removing water of crystallization, adding Gasteiger charge and adding hydrogen. Two-dimensional structures of compounds were extracted from the candidate compound library and converted into three-dimensional structures, followed by energy minimization. Molecular docking software was used to perform docking calculations between the compounds and the active pockets of the target protein, obtaining structural characteristic parameters such as binding energy and hydrogen bond interaction sites. Quantitative scoring was performed based on a target-compound matching algorithm, the calculation formula of which is as follows: ， in, It is a quantitative score of target-compound matching. The structure matching score between the compound and the target is calculated. The score is based on the consistency between the compound's function and the direction of target regulation. To determine the synergistic score of compound-target pairs in the regulatory pathway, score-based screening is performed. Compounds that exhibit specific interactions with active sites are considered candidate compounds with good binding ability to key targets.

9. The method for screening compounds that regulate blood-brain barrier permeability according to claim 1, characterized in that, In the S500 multi-stage compound screening, the specific steps for compound validation using primary isolated blood-brain barrier-related cell populations are as follows: Brain microvascular endothelial cells, astrocytes, pericytes, and microglia are isolated and purified from normal and specific disease model organisms, and a mixed cell culture system is constructed according to in vivo physiological proportions. Candidate compounds are added to this system, and barrier integrity is assessed by detecting changes in transcellular resistance. Fluorescently labeled small molecule tracers are used to detect substance permeation efficiency, reflecting the permeability regulation effect. Simultaneously, cell viability is measured using a cell viability assay kit, and compound toxicity is assessed by combining lactate dehydrogenase release levels. Compounds that can regulate transcellular resistance and tracer permeation efficiency, have a cell viability ≥80%, and lactate dehydrogenase release levels ≤1.5 times that of the normal control group are screened as effective compounds.

10. A compound screening system for regulating blood-brain barrier permeability, the system being applicable to the compound screening method for regulating blood-brain barrier permeability as described in any one of claims 1-9, characterized in that, The system includes: a data acquisition module, a data processing module, a network parsing module, a target pool construction module, and a compound screening module; The data acquisition module collects genomic, transcriptomic, proteomic, and metabolomics data of blood-brain barrier-related cells under normal and specific disease states, forming a multi-dimensional raw dataset. The data processing module preprocesses the collected multi-omics data, removes outliers and technical errors through data cleaning, and standardizes the data to generate a standardized dataset. The network parsing module performs cross-omics association analysis based on standardized data, constructs a gene-protein-metabolite interaction network, calculates node core degree through a network core degree algorithm, and screens out key targets that regulate blood-brain barrier permeability. The target pool construction module integrates the key targets obtained from the analysis according to their functions, and establishes a structured target pool containing information on target molecular properties, interaction relationships and regulatory pathways. The compound screening module: Based on the target pool, candidate compounds with good binding ability to key targets are screened through molecular docking; then, the compounds are verified using the primary isolated blood-brain barrier-related cell population to screen out effective compounds.

Images