ISG15 protein-based dermatomyositis biomarker identification method and system, electronic equipment and storage medium

Through bioinformatics analysis and clinical sample verification, ISG15 protein is identified as a biomarker of dermatomyositis, solving the problem of insufficient specificity and sensitivity of diagnostic methods in the prior art, and achieving more accurate diagnosis and personality of dermatomyositis Chemical medicine.

CN119028447APending Publication Date: 2024-11-26GENERAL HOSPITAL OF SOUTHERN THEATRE COMMAND OF PLA
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202411048498.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-01
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

Existing diagnostic methods for dermatomyositis have insufficient specificity and sensitivity, leading to the risk of misdiagnosis and misdiagnosis. It is urgent to develop a novel blood biomarker with low invasiveness, high sensitivity and specificity to achieve early diagnosis and accurate reflection. Condition.

Method used

By integrating bioinformatics analysis and clinical sample verification, differential expression analysis, functional correlation analysis, machine learning feature selection and weighted gene co-expression network analysis were used to identify ISG15 protein as a potential biomarker of dermatomyositis, improving diagnosis specificity Sex and accuracy.

Benefits of technology

This method can more accurately identify biomarkers related to dermatomyositis, improve the specificity and sensitivity of the diagnosis, ensure the generalization ability and stability of the identified markers, lay a solid foundation for clinical application, and help develop more accurate Diagnostic tools and treatment strategies to achieve personalized medical care.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119028447A_ABST
    Figure CN119028447A_ABST
Patent Text Reader

Abstract

The invention discloses a dermatomyositis biomarker identification method and system based on ISG15 protein, electronic equipment and a storage medium. The identification method comprises the steps of data acquisition and preprocessing; differential expression analysis and function correlation analysis: analyzing differential expression genes; selecting machine learning features; carrying out weighted gene co-expression network analysis; and verifying by using an independent data set. The present application aims to discover and validate specific DM diagnostic biomarkers by integrating bioinformatics analysis and clinical sample validation.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of biomarkers, and in particular to an identification method and system, electronic equipment and storage medium for a dermatomyositis biomarker based on an ISG15 protein. Background Art

[0002] Dermatomyositis (DM) is a rare inflammatory myopathy characterized by muscle weakness and skin lesions. Due to its complex etiology and diverse manifestations, the diagnosis of DM has always been a difficult problem in clinical medicine. Early diagnosis and accurate evaluation are crucial for disease control and treatment effectiveness. However, existing diagnostic methods and markers have certain limitations in specificity and sensitivity, leading to the risk of misdiagnosis and missed diagnosis.

[0003] Traditional methods for the diagnosis of dermatomyositis rely mainly on clinical manifestations, muscle enzyme levels, and muscle biopsy. Although these methods have certain clinical applications, they have many shortcomings, such as strong invasiveness, complex operation, and insufficient sensitivity and specificity. In recent years, with the development of molecular biology and genomics, researchers have discovered many biomarkers associated with dermatomyositis, including myositis-specific autoantibodies (MSAs), anti-MDA5 antibodies, interferon-related markers, etc.; these markers help to understand the pathological mechanism of dermatomyositis and provide new ideas for diagnosis and treatment. The hallmark autoimmune antibodies of dermatomyositis include anti-synthetase antibodies (30–40%), anti-Jo-1 antibodies (seen in 30%, with high specificity), anti-Mi-2 antibodies (<10%), etc., but these antibodies cannot fully cover all patients with dermatomyositis. Many patients have negative results for conventional antibody tests, which often leads to delays in diagnosis and treatment. Therefore, there is an urgent need to develop a new blood biomarker with low invasiveness, high sensitivity and specificity to facilitate early diagnosis and accurately reflect the condition of dermatomyositis.

[0004] ISG15 is a class of ubiquitin-like protein molecules induced by interferon (IFN) with a size of 15KD. The mature ISG15 protein contains two ubiquitin-like domains and has pleiotropic functions, acting as both a soluble molecule and a protein modifier. As a new potential biomarker, ISG15 (interferon-stimulated gene 15) protein has attracted widespread attention. ISG15 is a small molecule protein that plays an important role in antiviral immune response, and its expression levels vary significantly in a variety of diseases, including viral infection, autoimmune diseases, and tumors. Although the role of ISG15 in antiviral and anti-tumor immunity has been widely studied, its specific role in autoimmune diseases, including dermatomyositis, is not fully understood. At present, there are relatively few studies on the specific mechanism of action and biomarker potential of ISG15 in dermatomyositis. Summary of the invention

[0005] In order to overcome the shortcomings of the prior art, the purpose of this application is to provide a method and system for identifying dermatomyositis biomarkers based on ISG15 protein, which aims to discover and verify specific DM diagnostic biomarkers by integrating bioinformatics analysis and clinical sample verification.

[0006] To solve the above problems, the technical solutions adopted in this application are as follows:

[0007] The present application embodiment provides a method for identifying a dermatomyositis biomarker based on ISG15 protein, comprising:

[0008] Data acquisition and preprocessing: Gene expression datasets containing skin lesion samples from DM patients and control groups were obtained, and the datasets were merged for batch effect correction. The corrected expression matrix was used to identify differentially expressed genes.

[0009] Differential expression analysis and functional correlation analysis: analyze differentially expressed genes, identify genes with significant differential expression, and reveal the functions of differentially expressed genes in biological processes and pathways;

[0010] Machine learning feature selection, screening out potential biomarker candidate genes based on different machine learning algorithms;

[0011] Weighted gene co-expression network analysis, through gene co-expression network analysis, calculate the median absolute deviation of each gene, exclude the lower 50% of genes, identify module genes, and determine the key module genes;

[0012] Independent datasets were used for validation, and Venn diagrams were drawn to intersect differentially expressed genes, genes selected by machine learning, and WGCNA key module genes to identify potential biomarkers.

[0013] As a further preferred solution, the data acquisition and preprocessing steps described in the embodiment of the present application include:

[0014] Data acquisition: Gene expression datasets GSE46239, GSE142807, and GSE128314 containing skin lesion samples from DM patients and controls were obtained from the GEO database;

[0015] Data merging, using the R package “inSilicoMerging”, GSE46239 and GSE142807 datasets were merged for the discovery phase; GSE128314 was used for validation;

[0016] Batch effect correction: the merged data set was corrected for batch effect using the “removeBatchEffect” function in the R package “limma”;

[0017] Differentially expressed gene identification,The corrected expression matrix was used to identify differentially expressed,genes.

[0018] As a further preferred solution, in the differentially expressed gene identification process described in the examples of this application, the standards are |LogFC|>1 and adj.p<0.05

[0019] As a further preferred solution, the differential expression analysis and functional correlation analysis described in the examples of the present application include:

[0020] Differential expression analysis, using the R software package “limma” to perform differentially expressed gene analysis and identify genes with significant differential expression;

[0021] Functional correlation analysis, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were performed to reveal the functions of differentially expressed genes in biological processes and pathways.

[0022] As a further preferred scheme, in the feature selection process described in the embodiments of the present application, a machine learning algorithm is used for feature selection, and the least absolute shrinkage and selection operator, support vector machine recursive feature elimination and Bayesian ridge regression algorithm are respectively used for feature selection, and potential biomarker candidate genes are screened out according to different machine learning algorithms.

[0023] As a further preferred solution, the weighted gene co-expression network analysis described in the embodiments of the present application includes:

[0024] Identify modular genes, construct scale-free co-expression networks, calculate topological overlap matrices (TOMs), and identify modular genes;

[0025] Modular gene analysis: perform dissimilarity analysis of characteristic genes on modular genes to identify key modular genes.

[0026] As a further preferred scheme, in the machine learning feature selection step described in the embodiment of the present application, a Venn diagram is drawn for the differentially expressed genes (DEGs), the genes selected by machine learning, and the WGCNA key module genes to obtain the intersection. The results are a Venn diagram drawn using the Venn online tool. The independent data set GSE128314 is used for verification, which includes skin samples from 8 DM patients and 5 healthy controls. By drawing the ROC curve, the performance of DM patients and normal individuals is compared, showing AUC = 0.950.

[0027] The present application embodiment also provides a dermatomyositis biomarker identification system based on ISG15 protein, comprising:

[0028] A data acquisition and preprocessing unit is used to obtain a gene expression dataset containing skin lesion samples from DM patients and a control group, merge the datasets for batch effect correction, and use the corrected expression matrix to identify differentially expressed genes;

[0029] Differential expression analysis and functional correlation analysis units are used to analyze differentially expressed genes, identify genes with significant differential expression, and reveal the functions of differentially expressed genes in biological processes and pathways;

[0030] A machine learning feature selection unit, which is used to screen potential biomarker candidate genes based on different machine learning algorithms;

[0031] Weighted gene co-expression network analysis unit, used to calculate the median absolute deviation of each gene through gene co-expression network analysis, exclude the lower 50% of genes, identify module genes, and determine key module genes;

[0032] An independent dataset was used as a validation unit to draw a Venn diagram to obtain the intersection of differentially expressed genes, genes selected by machine learning, and WGCNA key module genes to identify potential biomarkers.

[0033] An embodiment of the present application also provides an electronic device, comprising a memory storing executable program code and a processor coupled to the memory; the processor calls the executable program code stored in the memory to execute the method for identifying dermatomyositis biomarkers based on ISG15 protein described in an embodiment of the present application.

[0034] The embodiment of the present application also provides a computer-readable storage medium storing a computer program, wherein the computer program enables a computer to execute the method for identifying a dermatomyositis biomarker based on the ISG15 protein as described in the embodiment of the present application.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] 1. The method for identifying dermatomyositis biomarkers based on ISG15 protein described in this application can comprehensively evaluate the expression pattern and function of genes from multiple dimensions by comprehensively applying multiple analysis methods (differential expression analysis, functional correlation analysis, machine learning feature selection, weighted gene co-expression network analysis), thereby more accurately identifying biomarkers associated with dermatomyositis.

[0037] 2. In the method for identifying dermatomyositis biomarkers based on ISG15 protein described in this application, a machine learning algorithm is used for feature selection, and a weighted gene co-expression network analysis is used to determine key module genes, which helps to screen out genes that are highly specifically associated with dermatomyositis as potential biomarkers and improve the specificity of diagnosis.

[0038] 3. Furthermore, the ISG15 protein-based dermatomyositis biomarker identification method described in this application is verified by using an independent data set, which can evaluate the generalization ability and stability of the identified biomarker, ensure its applicability in different samples, and lay a solid foundation for subsequent clinical applications.

[0039] 4. The identification method of dermatomyositis biomarkers based on ISG15 protein described in this application can identify dermatomyositis biomarkers based on ISG15 protein, which will help develop more accurate diagnostic tools and treatment strategies, achieve personalized medicine for dermatomyositis, and improve the quality of life of patients.

[0040] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0042] Figure 1 A volcano plot of differential expression analysis and functional correlation analysis in the embodiment of this application;

[0043] Figure 2 This is a heat map of differential expression analysis and functional correlation analysis in the examples of this application;

[0044] Figure 3 The GO analysis results of differential expression analysis and functional correlation analysis in the examples of this application;

[0045] Figure 4 KEGG analysis results of differential expression analysis and functional correlation analysis of the embodiments of this application;

[0046] Figure 5 This is a flow chart of feature selection for machine learning in the present application, where (A) biomarker screening in the Lasso model. The number of genes corresponding to the lowest point of the curve (n = 11) is most suitable for DM diagnosis, (B) biomarker screening based on SVM-RFE, (C) biomarker analysis of Bayesian Ridge, and genes are ranked according to importance scores.

[0047] Figure 6This is a flow chart of the weighted gene co-expression network analysis of an embodiment of the present application; wherein, (A) soft threshold power (left) and average connectivity (right) of the WGCNA network, (B) gene co-expression modules represented by different colors under the gene tree, (C) clustering heat map of module feature vectors, and (D) heat map of the correlation between module feature genes and DM clinical characteristics.

[0048] Figure 7 It is the intersection Venn diagram of DEGs, WGCNA module genes, LASSO, SVM-RFE and Bayesian Ridge analysis in the embodiment of the present application;

[0049] Figure 8 The ROC curve for evaluating the diagnostic effectiveness of ISG15 in the embodiment of the present application;

[0050] Fig. 9 This is a graph showing the results of validating ISG15 using the GSE128314 dataset.

[0051] Fig.10 Results of immunohistochemical (IHC) analysis of clinical skin biopsy specimens of the DM group are shown in the figure, wherein, (A) immunohistochemical analysis of ISG15 expression in the DM group (X10), (B) immunohistochemical analysis of ISG15 expression in the control group (X10), (C) immunohistochemical analysis of ISG15 expression in surrounding vascular areas and lymphocytes in the DM group (X40), (D) immunohistochemical analysis of ISG15 expression in the control group (X40).

[0052] Fig.11 Schematic diagram of immune cell infiltration and correlation analysis, including (A) the proportion of 22 immune cells in each DM sample, (B) immune cell infiltration between DM and normal controls, and (C) the association between ISG15 and different immune cell infiltrations in DM. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0054] The term "comprising" and other equivalent descriptions involved in the specification and claims of this application are intended to cover non-exclusive inclusions, including both the contents clearly described in the specification and claims, and the steps or units that are not described in the specification and claims but are inherent in the product, method or structure.

[0055] Example 1

[0056] The present application embodiment provides a method for identifying a dermatomyositis biomarker based on ISG15 protein, comprising:

[0057] S1, data acquisition and preprocessing, gene expression datasets containing skin lesion samples from DM patients and control group were obtained, the datasets were merged for batch effect correction, and the corrected expression matrix was used to identify differentially expressed genes.

[0058] In the examples of the present application, the gene expression datasets GSE46239, GSE142807 and GSE128314 containing skin lesion samples of DM patients and control groups were obtained from the GEO (Gene Expression Omnibus, https: / / www.ncbi.nlm.nih.gov / geo / ) database. The specific research content, sample type, experimental conditions and other information contained in GSE46239 need to be directly accessed to the GEO database for query; the GSE142807 dataset is used to study the key genes and potential molecular pathways of dermatomyositis (DM); the GSE128314 dataset can be used to explore gene expression changes, differentially expressed gene analysis, gene set enrichment analysis and potential biomarkers or therapeutic targets in specific disease states.

[0059] In some embodiments of the present application, the R software package "inSilicoMerging" is used to merge the GSE46239 and GSE142807 data sets for the discovery phase. Specific operations include: before merging, it is necessary to download the data of GSE46239 and GSE142807 from GEO (Gene Expression Omnibus), which are usually CEL files or processed expression matrices, load the data after downloading, use the functions in the inSilicoMerging package to merge the data, and save the merged data. In other embodiments, some basic data analysis, such as PCA analysis, can also be performed after the merger to view the effect of the data merger.

[0060] Batch effect refers to the systematic differences between different batches due to differences in experimental conditions, sample processing or measurement platforms. In order to eliminate these differences, the "removeBatchEffect" function in the R software package "limma" is used in the embodiment of the present application to correct the batch effect of the merged data set; before calling the removeBatchEffect function, a variable representing the batch information needs to be determined. This variable is a vector with the same number of columns as mergedData, and each element corresponds to a batch of samples. In this function model, it is assumed that the observed gene expression value is composed of three parts: the true expression value of the gene, the batch effect and the random error. The goal is to estimate and remove the batch effect to obtain the corrected gene expression value.

[0061] The specific mathematical model can be expressed as:

[0062] Y=Xβ+Zγ+∈

[0063] Where Y is the observed gene expression value matrix, X is the design matrix, β is the true expression value of the gene, Z is the batch indicator matrix, γ is the batch effect, and ∈ is the random error.

[0064] The "removeBatchEffect" function estimates and removes γ through the least squares method, thereby correcting the batch effect.

[0065] The corrected expression matrix was used to identify differentially expressed genes with the criteria of |LogFC|>1 and adj.p<0.05.

[0066] S2, differential expression analysis and functional correlation analysis, analyze differentially expressed genes, identify significantly differentially expressed genes, and reveal the functions of differentially expressed genes in biological processes and pathways.

[0067] In the present application example, the R software package "limma" was further used to perform differentially expressed gene analysis to identify genes with significant differential expression; the process is shown in Figure 1-Figure 4 ,in Figure 1 For the volcano map, Figure 2 For heatmap, Figure 3 The GO analysis results, Figure 4The volcano plot is the result of KEGG analysis. Volcano plots are often used to screen genes that are significantly upregulated or downregulated under specific conditions. These genes may be disease markers, drug targets, or key molecules in biological processes. The horizontal axis of the graph represents the magnitude of the gene expression change, and the vertical axis represents the statistical significance. By interpreting the volcano plot, genes with significant differential expression can be quickly identified. In the heat map, each row represents a gene, and each column represents a sample or condition. The depth of the color indicates the level of gene expression, thus forming an intuitive "heat" pattern. Through the heat map, we can observe the expression changes of genes in different samples or conditions, and then reveal their potential biological functions or regulatory mechanisms. Figure 3 The results of GO analysis are shown, where different colors or shapes represent different GO classifications, and the size or color depth of the points indicate the degree of enrichment or significance. Through GO analysis, we can have a deeper understanding of the role of differentially expressed genes in biological processes. Figure 4 The results of KEGG analysis are shown, in which different pathways are represented by different colors or shapes, and the size or color depth of the points reflects the enrichment or significance of differentially expressed genes in the pathways. Through KEGG analysis, we can further understand the functions and interactions of differentially expressed genes in biological pathways.

[0068] S3, machine learning feature selection, screening potential biomarker candidate genes based on different machine learning algorithms;

[0069] In the feature selection process described in the embodiment of the present application, a machine learning algorithm is used for feature selection, and the least absolute shrinkage and selection operator (LASSO), support vector machine recursive feature elimination (SVM-RFE) and Bayesian ridge regression algorithm are used for feature selection, and potential biomarker candidate genes are screened according to different machine learning algorithms. See the machine learning feature selection flow chart for details. Figure 5 , where (A) Biomarker screening in the Lasso model. The number of genes corresponding to the lowest point of the curve (n = 11) is most suitable for DM diagnosis, (B) Biomarker screening based on SVM-RFE, (C) Biomarker analysis of Bayesian Ridge, genes are ranked according to importance scores.

[0070] S4, Weighted gene co-expression network analysis,WGCNA software package was used to perform gene co-expression network analysis, calculate the median absolute deviation (MAD) of each gene, exclude the lower 50% of genes, identify module genes, and determine the key module genes.

[0071] In the present application embodiment, weighted gene co-expression network analysis includes module gene identification, and constructs a scale-free co-expression network. In the gene co-expression network, this structure reflects the complexity of key genes and regulatory networks in biological systems; The key to constructing a scale-free co-expression network is to select a suitable soft threshold (soft threshold) to perform weighted conversion on the gene expression correlation matrix, so as to obtain an adjacency matrix. In the present application embodiment, the soft threshold (softthreshold) is set to set p value <0.05 as a significant threshold, and multiple test correction is performed using the adjusted p value; Calculate the topological overlap matrix (TOM), and identify module genes; In the gene co-expression network, the TOM matrix can reduce noise and false correlations. After obtaining the TOM matrix, the module genes can be further identified by cluster analysis. Commonly used clustering methods include hierarchical clustering, module detection algorithms (such as the module detection algorithm in WGCNA), etc., and are grouped into different modules according to the similarity in the TOM matrix, and the key modules and genes related to specific diseases or traits can be further identified by comparing the correlation of module genes with phenotypic data or other biological characteristics, and key module genes can be determined. The weighted gene co-expression network analysis process can be found in Figure 6 .

[0072] S5, using independent datasets for validation, a Venn diagram was drawn to intersect the differentially expressed genes, genes selected by machine learning, and WGCNA key module genes to identify potential biomarkers.

[0073] In the validation step using an independent data set described in the embodiment of the present application, a Venn diagram was performed on the differentially expressed genes (DEGs), the genes selected by machine learning, and the WGCNA key module genes (see Figure 7 ) were used to draw the intersection, and the result was a Venn diagram drawn using the Venn online tool. The independent dataset GSE128314 was used for validation, which included skin samples from 8 DM patients and 5 healthy controls. The ROC curve was drawn (see Figure 8 ), comparing the performance of DM patients with normal individuals, showed AUC = 0.950. It has high specificity and sensitivity. The validation results can be found in Fig. 9 , the expression of ISG15 in DM patients was verified by GSE128314 dataset.

[0074] Furthermore, in the above embodiment, the independent data set GSE128314 was used to perform reliability analysis to evaluate the consistency, repeatability and stability of the experimental data.

[0075] Example 2

[0076] The present application embodiment also provides a dermatomyositis biomarker identification system based on ISG15 protein, comprising:

[0077] A data acquisition and preprocessing unit is used to obtain a gene expression dataset containing skin lesion samples from DM patients and a control group, merge the datasets for batch effect correction, and use the corrected expression matrix to identify differentially expressed genes;

[0078] Differential expression analysis and functional correlation analysis units are used to analyze differentially expressed genes, identify genes with significant differential expression, and reveal the functions of differentially expressed genes in biological processes and pathways;

[0079] A machine learning feature selection unit, which is used to screen potential biomarker candidate genes based on different machine learning algorithms;

[0080] Weighted gene co-expression network analysis unit, used to calculate the median absolute deviation of each gene through gene co-expression network analysis, exclude the lower 50% of genes, identify module genes, and determine key module genes;

[0081] An independent dataset was used as a validation unit to draw a Venn diagram to obtain the intersection of differentially expressed genes, genes selected by machine learning, and WGCNA key module genes to identify potential biomarkers.

[0082] Furthermore, based on Example 1 of the present application, the embodiment of the present application also provides an electronic device, including a memory storing executable program code and a processor coupled to the memory; the processor calls the executable program code stored in the memory to execute the method for identifying dermatomyositis biomarkers based on ISG15 protein described in Example 1 of the present application.

[0083] Furthermore, based on Example 1 of the present application, the embodiment of the present application also provides a computer-readable storage medium, which stores a computer program, wherein the computer program enables the computer to execute the method for identifying dermatomyositis biomarkers based on ISG15 protein described in Example 1 of the present application.

[0084] Application Example 1

[0085] Human samples and immunohistochemistry (IHC): Skin biopsy samples were obtained from DM patients and healthy controls.

[0086] Samples: Immunohistochemistry (IHC) experiments were performed on skin biopsy specimens from 9 dermatomyositis patients and 12 healthy controls to validate our bioinformatics findings.

[0087] Experimental procedures: First, paraffin-embedded tissue sections were dewaxed with xylene and rehydrated through a series of increasing ethanol solutions. Next, antigen retrieval was performed using 1 mM EDTA, followed by preincubation with Tris-buffered saline containing 5% goat serum to reduce nonspecific binding. Subsequently, the sections were probed with rabbit anti-ISG15 monoclonal antibody (EPR24482-49, diluted 1:500). After applying horseradish peroxidase-labeled secondary antibodies, 3,3'-diaminobenzidine matrix was used for color development and counterstained with hematoxylin to increase contrast. The results of immunohistochemistry (IHC) analysis of clinical skin biopsy specimens in the DM group showed (see Fig.10 ), ISG15 expression was generally positive ( Fig.10 A. Fig.10 C), while negative expression was observed in the control group ( Fig.10 B. Fig.10 D). In addition, immunostaining was mainly observed in the perivascular areas and lymphocytes in the skin tissue ( Fig.10 C).

[0088] Correlation analysis of immune cells showed that ISG15 was associated with a variety of immune cells, suggesting its role in the pathogenesis of dermatomyositis.

[0089] Application Example 2

[0090] Immune cell infiltration assessment and correlation analysis

[0091] The CIBERSORT algorithm was used to evaluate the relative abundance of 22 different immune cell types in the samples. By analyzing the immune cell infiltration pattern, we can further understand the characteristics of the immune response in DM patients. The Spearman correlation coefficient between the level of infiltrating immune cells and the proposed diagnostic biomarker ISG15 was calculated to analyze the relationship between immune cell infiltration and biomarker expression and to find potential mechanistic connections. The analysis results are shown in Fig.11 .

[0092] Fig.11 (A) The proportion of 22 immune cells in each sample of DM, (B) Immune cell infiltration between DM and normal control. (C) The association of ISG15 with different immune cell infiltration in DM.

[0093] The above-mentioned embodiments are only preferred embodiments of the present invention and cannot be used to limit the scope of protection of the present invention. Any non-substantial changes and substitutions made by technicians in this field on the basis of the present invention shall fall within the scope of protection required by the present invention.

Claims

1. A method for identifying dermatomyositis biomarkers based on ISG15 protein, It is characterized in that include Data acquisition and preprocessing: Gene expression datasets containing skin lesion samples from DM patients and control groups were obtained, and the datasets were merged for batch effect correction. The corrected expression matrix was used to identify differentially expressed genes. Differential expression analysis and functional correlation analysis: analyze differentially expressed genes, identify genes with significant differential expression, and reveal the functions of differentially expressed genes in biological processes and pathways; Machine learning feature selection, screening out potential biomarker candidate genes based on different machine learning algorithms; Weighted gene co-expression network analysis, through gene co-expression network analysis, calculate the median absolute deviation of each gene, exclude the lower 50% of genes, identify module genes, and determine the key module genes; Independent datasets were used for validation, and Venn diagrams were drawn to intersect differentially expressed genes, genes selected by machine learning, and WGCNA key module genes to identify potential biomarkers.

2. The method for identifying a dermatomyositis biomarker based on ISG15 protein according to claim 1, It is characterized in that The data acquisition and preprocessing steps include: Data acquisition: Gene expression datasets GSE46239, GSE142807, and GSE128314 containing skin lesion samples from DM patients and controls were obtained from the GEO database; Data merging: The R package "inSilicoMerging" was used to merge the GSE46239 and GSE142807 datasets for the discovery phase; GSE128314 was used for validation; Batch effect correction: use the "removeBatchEffect" function in the R package "limma" to correct the batch effect of the merged data set; Differentially expressed gene identification,The corrected expression matrix was used to identify differentially expressed,genes.

3. The method for identifying dermatomyositis biomarkers based on ISG15 protein according to claim 2, It is characterized in that During the identification of differentially expressed genes, the criteria were |LogFC|>1 and adj.p<0.

05.

4. The method for identifying dermatomyositis biomarkers based on ISG15 protein according to claim 1, It is characterized in that The differential expression analysis and functional correlation analysis include: Differential expression analysis, using the R software package "limma" to perform differentially expressed gene analysis and identify genes with significant differential expression; Functional correlation analysis, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were performed to reveal the functions of differentially expressed genes in biological processes and pathways.

5. The method for identifying dermatomyositis biomarkers based on ISG15 protein according to claim 1, It is characterized in that During the feature selection process, machine learning algorithms were used for feature selection, and the least absolute shrinkage and selection operator, support vector machine recursive feature elimination, and Bayesian ridge regression algorithm were applied for feature selection, respectively, and potential biomarker candidate genes were screened out according to different machine learning algorithms.

6. The method for identifying dermatomyositis biomarkers based on ISG15 protein according to claim 1, It is characterized in that The weighted gene co-expression network analysis includes: Identify modular genes, construct scale-free co-expression networks, calculate topological overlap matrices (TOMs), and identify modular genes; Modular gene analysis: perform dissimilarity analysis of characteristic genes on modular genes to identify key modular genes.

7. The method for identifying dermatomyositis biomarkers based on ISG15 protein according to claim 1, It is characterized in that The differentially expressed genes (DEGs), genes selected by machine learning, and WGCNA key module genes were intersected by Venn diagram drawing. The results were drawn using the Venn diagram drawn by the Venn online tool. The independent dataset GSE128314 was used for validation, which included skin samples from 8 DM patients and 5 healthy controls. The ROC curve was drawn to compare the performance of DM patients with that of normal individuals, showing AUC = 0.

950.

8. A recognition system for dermatomyositis biomarkers based on ISG15 protein, It is characterized in that include: A data acquisition and preprocessing unit is used to obtain a gene expression dataset containing skin lesion samples from DM patients and a control group, merge the datasets for batch effect correction, and use the corrected expression matrix to identify differentially expressed genes; Differential expression analysis and functional correlation analysis units are used to analyze differentially expressed genes, identify genes with significant differential expression, and reveal the functions of differentially expressed genes in biological processes and pathways; A machine learning feature selection unit, which is used to screen potential biomarker candidate genes based on different machine learning algorithms; Weighted gene co-expression network analysis unit, used to calculate the median absolute deviation of each gene through gene co-expression network analysis, exclude the lower 50% of genes, identify module genes, and determine key module genes; An independent dataset was used as a validation unit to draw a Venn diagram to obtain the intersection of differentially expressed genes, genes selected by machine learning, and WGCNA key module genes to identify potential biomarkers.

9. An electronic device, It is characterized in that It comprises a memory storing executable program code and a processor coupled to the memory; the processor calls the executable program code stored in the memory to execute the method for identifying dermatomyositis biomarkers based on ISG15 protein according to any one of claims 1 to 7.

10. A computer-readable storage medium, It is characterized in that The computer program is stored therein, wherein the computer program enables a computer to execute the method for identifying a dermatomyositis biomarker based on the ISG15 protein according to any one of claims 1 to 7.

Citation Information

Cited By

  • Blood protein data quality evaluation and correction method, system, equipment and medium

    CN120236657A