Sample typing method, system, equipment and medium for medullary blastoma
By using weighted gene co-expression networks and differential expression analysis, combined with FHDC1 gene status, the problem of misdiagnosis and missed diagnosis of medulloblastoma G3 and G4 subtypes was solved, enabling accurate subtyping and personalized treatment of medulloblastoma samples.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-27
- Publication Date
- 2026-03-24
AI Technical Summary
Current technologies have led to misdiagnosis and missed diagnosis in the G3 and G4 subtype classification of medulloblastoma, resulting in insufficient precision in treatment decisions. They mainly rely on traditional pathological morphological observation and single immunohistochemical markers, which are difficult to cover the complex molecular differences between subtypes.
By acquiring gene expression data and clinical subtype information of medulloblastoma patients, a co-expression module was constructed using a weighted gene co-expression network. Combined with correlation scores and differential expression analysis, target module genes related to the G3/G4 subtype, especially the FHDC1 gene status, were screened to achieve precise subtyping.
It significantly improves the accuracy of medulloblastoma sample subtyping, reduces subjectivity and misjudgment rate, and provides precise basis for personalized treatment and prognostic assessment.
Smart Images
Figure CN121725876A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of sample typing, and more particularly to a method, system, device, and medium for typing samples of medulloblastoma. Background Technology
[0002] Medulloblastoma is one of the most common malignant brain tumors in children, exhibiting high heterogeneity. The G3 and G4 subtypes show significant differences in clinical progression, treatment response, and prognostic outcomes. The G3 subtype is generally more malignant and prone to metastasis, while the G4 subtype has more complex clinical behavior and varying responsiveness to traditional treatments. Accurate differentiation between these two subtypes is a prerequisite for personalized treatment—different subtypes require different chemotherapy regimens, radiotherapy intensities, and follow-up strategies. Ambiguous subtyping can lead to overtreatment or undertreatment, significantly impacting patients' quality of life and long-term survival. Furthermore, accurate subtyping provides crucial information for prognostic assessment, helping clinicians predict the risk of disease progression and develop optimal treatment plans.
[0003] Current techniques for classifying G3 and G4 subtypes of medulloblastoma primarily rely on traditional pathological morphological observation, immunohistochemical detection of a few biomarkers (such as MYC amplification status), and limited gene expression profiling. Traditional pathological classification depends on microscopic cell morphology, which is highly subjective and has poor reproducibility, especially when tumor tissue heterogeneity exists, leading to misdiagnosis. Immunohistochemistry, based on single or a few biomarkers, struggles to cover the complex molecular differences between subtypes, resulting in unclear classification of some borderline samples. While some molecular classification methods incorporate gene expression data, they often rely on small sample cohorts or non-specific gene sets, failing to focus on key genes closely related to the core characteristics of the subtype, resulting in low classification accuracy and insufficient generalizability. These shortcomings directly lead to misdiagnosis or missed diagnosis of G3 and G4 subtypes in clinical practice, affecting the accuracy of treatment decisions. Summary of the Invention
[0004] This invention provides a method, system, device, and medium for medulloblastoma sample typing, which can improve the accuracy of medulloblastoma sample typing.
[0005] An embodiment of the present invention provides a sample typing method for medulloblastoma, including: To obtain gene expression data and clinical subtype information of medulloblastoma patients; Based on the gene expression data, several co-expression modules are constructed through a weighted gene co-expression network, and the characteristic genes of each co-expression module are calculated. Based on the characteristic genes and the relevant scores, target co-expression modules related to the sample to be classified are screened out, and target module genes are extracted from the target co-expression modules. The relevant scores are the correlation scores between the characteristic genes and the clinical subtype information. The sample to be classified includes G3 subtype and G4 subtype. Differentially expressed gene analysis is performed on the samples to be classified to screen out differentially expressed genes that meet the first preset conditions, and subtype candidate genes are determined based on the target module genes and the differentially expressed genes. Based on the status of the FHDC1 gene among the subtype candidate genes, the samples to be classified for medulloblastoma are classified.
[0006] This invention, by acquiring gene expression data and clinical subtype information, can simultaneously incorporate molecular-level gene expression data and clinical phenotypic information, overcoming the limitations of traditional pathology relying solely on morphological observation and providing multi-dimensional reference for subsequent analysis. The use of weighted gene co-expression network analysis to construct modules allows functionally related genes to be clustered into modules, rather than analyzing individual genes in isolation, which better aligns with the holistic principles of biological regulation. By screening target co-expression modules through correlation scoring, it can accurately capture gene sets highly correlated with the core differences in the G3 / G4 subtype, eliminating interference from irrelevant genes and significantly improving the specificity of candidate genes. Differential expression analysis can focus on genes with significant expression changes between different subtypes. Combined with the target module genes screened in previous steps, the intersection of these genes determines candidate genes for the subtype, further narrowing the scope to core genes that are both related to the subtype and show significant differences between subtypes, overcoming the limitations of single immunohistochemical markers. Based on FHDC1 gene status typing, it can significantly reduce the subjectivity and misjudgment rate of traditional typing, achieving accurate differentiation. Compared with existing technologies, this application can improve the accuracy of medulloblastoma sample typing.
[0007] Furthermore, the relevant score includes a correlation coefficient and a significance P-value. The calculation of the characteristic genes of each co-expression module, and the selection of target co-expression modules related to the sample to be genotyped based on the characteristic genes and the relevant score, specifically involves: Principal component analysis is performed on each of the co-expression modules to determine the first principal component of the gene expression data in each co-expression module as the characteristic gene corresponding to each co-expression module; Calculate the correlation coefficient and corresponding significance P-value between each of the characteristic genes and the clinical subtype information. When the correlation coefficient and the significance P-value meet the second preset condition, determine the target co-expression module associated with the sample to be subtyped.
[0008] This approach of using weighted gene co-expression network analysis to construct modules allows functionally related genes to be clustered into modules, rather than analyzing individual genes in isolation, which is more in line with the holistic principles of biological regulation.
[0009] Further, the differentially expressed gene analysis of the sample to be genotyped to screen out differentially expressed genes that meet the first preset condition specifically includes: Based on the clinical subtype information, the samples to be classified are divided into at least two groups of samples to be classified. Differential expression analysis tools were used to perform gene expression differential analysis on the different groups of the samples to be classified, so as to calculate the fold change and significance P value of each target module gene among the different groups of the samples to be classified; Multiple corrections are applied to the significance P-value to obtain the false discovery rate, and genes whose expression fold change and false discovery rate meet the first preset condition are selected as differentially expressed genes.
[0010] By using differential expression analysis, we can focus on genes that show significant changes in expression among different subtypes. Combined with the target module genes screened in the previous steps, we can determine candidate genes for subtypes through intersection. This can further narrow down the scope to core genes that are both related to the subtype and show significant differences among subtypes, overcoming the limitation of single immunohistochemical markers.
[0011] Furthermore, the subtyping of medulloblastoma samples based on the status of the FHDC1 gene among the candidate subtypes specifically involves: Obtain the expression level of the FHDC1 gene among the candidate genes of the subtype; The expression level is compared with a preset expression threshold, wherein the preset expression threshold is determined based on the fold change in the expression of the FHDC1 gene among different sample groups to be genotyped, as analyzed by the differentially expressed gene analysis. When the expression level is higher than or equal to the preset expression threshold, the sample to be classified is determined to be the G4 subtype; when the expression level is lower than the preset expression threshold, the sample to be classified is determined to be the G3 subtype.
[0012] This FHDC1 gene status-based typing method can significantly reduce the subjectivity and misjudgment rate of traditional typing, achieving accurate differentiation.
[0013] Furthermore, before constructing several co-expression modules through a weighted gene co-expression network based on the gene expression data, the method further includes: The original gene expression data and clinical information were subjected to sample quality control, and individuals with mismatched sample numbers or missing clinical information were removed to obtain the first gene expression matrix. The first gene expression matrix is converted into a numerical format to obtain the second gene expression matrix, thus obtaining preprocessed gene expression data and clinical subtype information.
[0014] By eliminating invalid samples and standardizing data formats, the integrity and standardization of gene expression data are ensured, providing a high-quality data foundation for the subsequent construction of co-expression modules, reducing the interference of poor-quality data on module division and gene screening, and improving the reliability of genotyping analysis from the source.
[0015] Furthermore, after typing the medulloblastoma sample to be classified, the process also includes: Based on the typing results, the corresponding FHDC1 gene expression level is obtained, and the prognostic cutoff value of the FHDC1 gene expression level is determined. Based on the cutoff value, patients are divided into high expression group and low expression group. Survival curves were plotted for the high-expression group and the low-expression group, and the prognostic risk of the patients was assessed based on the survival curves.
[0016] This "subtype-prognosis" linkage verification reduces subjective misjudgment, provides key evidence for prognostic assessment, helps clinicians predict the risk of disease progression, and thus develop better treatment plans.
[0017] Another embodiment of the present invention provides a sample typing system for medulloblastoma, comprising: an acquisition module, a screening module, and a typing module; The acquisition module is used to acquire gene expression data and clinical subtype information of medulloblastoma patients; The screening module is used to construct several co-expression modules based on the gene expression data through a weighted gene co-expression network, calculate the characteristic genes of each co-expression module, screen out target co-expression modules related to the sample to be classified based on the characteristic genes and the correlation score, and extract the target module gene from the target co-expression module. The correlation score is the correlation score between the characteristic gene and the clinical subtype information. The sample to be classified includes G3 subtype and G4 subtype. The typing module is used to perform differentially expressed gene analysis on the sample to be typed, to screen out differentially expressed genes that meet the first preset conditions, and to determine subtype candidate genes based on the target module gene and the differentially expressed genes, and to type the medulloblastoma sample based on the status of the FHDC1 gene in the subtype candidate genes.
[0018] This invention, by acquiring gene expression data and clinical subtype information, can simultaneously incorporate molecular-level gene expression data and clinical phenotypic information, overcoming the limitations of traditional pathology relying solely on morphological observation and providing multi-dimensional reference for subsequent analysis. The use of weighted gene co-expression network analysis to construct modules allows functionally related genes to be clustered into modules, rather than analyzing individual genes in isolation, which better aligns with the holistic principles of biological regulation. By screening target co-expression modules through correlation scoring, it can accurately capture gene sets highly correlated with the core differences in the G3 / G4 subtype, eliminating interference from irrelevant genes and significantly improving the specificity of candidate genes. Differential expression analysis can focus on genes with significant expression changes between different subtypes. Combined with the target module genes screened in previous steps, the intersection of these genes determines candidate genes for the subtype, further narrowing the scope to core genes that are both related to the subtype and show significant differences between subtypes, overcoming the limitations of single immunohistochemical markers. Based on FHDC1 gene status typing, it can significantly reduce the subjectivity and misjudgment rate of traditional typing, achieving accurate differentiation. Compared with existing technologies, this application can improve the accuracy of medulloblastoma sample typing.
[0019] Another embodiment of the present invention provides a terminal device, including: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, it implements the steps of the medulloblastoma sample typing method as described in the present invention.
[0020] Another embodiment of the present invention also provides a computer-readable storage medium item, including: a stored computer program, which, when the computer program is executed, controls the device where the computer-readable storage medium is located to perform the steps of the medulloblastoma sample typing method as described in the present invention. Attached Figure Description
[0021] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.
[0022] Figure 1 This is a schematic flowchart of an embodiment of the medulloblastoma sample typing method provided in this application; Figure 2 This is a flowchart illustrating one embodiment of steps S201 to S204 provided in this application; Figure 3 This is a module-subtype correlation heatmap obtained through weighted gene co-expression network analysis provided in this application; Figure 4This is a volcano diagram illustrating one embodiment of the differentially expressed genes between the G3 and G4 subtypes provided in this application; Figure 5 This is a volcano diagram illustrating another embodiment of the differentially expressed genes between the G3 and G4 subtypes provided in this application; Figure 6 This is a schematic diagram of the structure of one embodiment of the medulloblastoma sample typing system provided in this application. Detailed Implementation
[0023] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0024] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.
[0025] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.
[0026] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.
[0027] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.
[0028] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).
[0029] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; 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; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.
[0030] Medulloblastoma is the most common malignant brain tumor in children, exhibiting high heterogeneity. The G3 and G4 subtypes show significant differences in clinical progression, treatment response, and prognosis. Accurate subtyping of the G3 and G4 subtypes is a core prerequisite for personalized treatment and a key basis for prognostic assessment, helping physicians predict risks and optimize treatment plans. Current techniques for subtyping the G3 and G4 subtypes of medulloblastoma mainly rely on traditional pathological morphological observation, immunohistochemical detection of a few biomarkers (such as MYC amplification status), and limited gene expression profiling analysis. However, these methods often lead to misdiagnosis and missed diagnosis of the G3 and G4 subtypes in clinical practice, affecting the accuracy of treatment decisions.
[0031] See Figure 1 To improve the accuracy of medulloblastoma sample typing, an embodiment of the present invention provides a medulloblastoma sample typing method, including steps S101 to S103. Step S101: Obtain gene expression data and clinical subtype information of medulloblastoma patients; In some embodiments, obtaining gene expression data from medulloblastoma patients specifically involves downloading a dataset of medulloblastoma patients from a public database to obtain gene expression data from the dataset.
[0032] It should be noted that the public database can be the Gene Expression Omnibus database. Download the GSE85217 dataset from the database. This dataset contains standardized microarray gene expression data of several cases (e.g., 763 cases) of medulloblastoma patients based on the Affymetrix platform (Human Genome U133 Plus 2.0 Array). It has already undergone background correction and standardization in the public database to ensure the comparability of expression levels between different samples.
[0033] It should be noted that 332 clinical samples from neurosurgical medulloblastoma patients were also included. Fresh tissue samples, cryopreserved at -80°C after surgical resection, underwent whole-genome sequencing (WGS) using the Illumina platform. Sequencing data underwent mutation detection and functional annotation using the GATK standard workflow, and were integrated into a standard Mutation Annotation Format (MAF) file to supplement and validate gene expression data. It should also be noted that for the 322 samples we had on hand, their clinical subtype information was confirmed through a double-blind review by pathology experts, covering different medulloblastoma subtypes. Informed consent was obtained from all patients or their legal guardians for all samples, and the information was approved by the hospital's ethics committee to ensure the accuracy and compliance of the clinical subtype information.
[0034] In some embodiments, clinical subtype information of medulloblastoma patients is obtained, specifically as follows: For the GSE85217 dataset samples, clinical subtype information is extracted from the original follow-up table (clini_763_Michael.xlsx) provided by it, including the subtype information (WNT, SHH, Group3, Group4), gender, age, risk group, survival status (life / death) and survival time (unit: months) of each patient, and the expression sample number is precisely matched with the clinical sample number through a unified format to exclude individuals with missing subtype information.
[0035] Step S102: Based on the gene expression data, several co-expression modules are constructed through a weighted gene co-expression network, and the characteristic genes of each co-expression module are calculated. Based on the characteristic genes and the relevant scores, target co-expression modules related to the sample to be classified are screened out, and target module genes are extracted from the target co-expression modules. The relevant scores are the correlation scores between the characteristic genes and the clinical subtype information. The sample to be classified includes G3 subtype and G4 subtype. In some embodiments, before constructing several co-expression modules based on the gene expression data through a weighted gene co-expression network, the method further includes: performing sample quality control on the acquired raw gene expression data and clinical information, and removing individuals whose sample numbers cannot be matched or whose clinical information is missing, to obtain a first gene expression matrix; converting the first gene expression matrix into a numerical format to obtain a second gene expression matrix, thereby obtaining preprocessed gene expression data and clinical subtype information. Specifically, sample quality control is first performed to remove individuals whose sample numbers do not match or whose clinical information (including follow-up information) is missing from the expression matrix, ensuring a strict one-to-one correspondence between expression data and clinical data to avoid pairing errors, thus obtaining the first gene expression matrix. Next, the first gene expression matrix is transposed and converted to a numerical format to obtain the second gene expression matrix. Subsequently, since the original data has been standardized by the GEO platform, this standardized second gene expression matrix is directly used to ensure the comparability of expression values between platforms in subsequent analyses. Simultaneously, in the subsequent weighted gene co-expression network analysis and identification process, by setting some screening parameters, background genes with extremely low expression levels or poor co-expression can be automatically screened out, thereby completing low-expression gene removal and noise reduction, ultimately obtaining preprocessed gene expression data and clinical subtype information, providing a reliable data foundation for co-expression network construction.
[0036] By eliminating invalid samples and standardizing data formats, the integrity and standardization of gene expression data are ensured, providing a high-quality data foundation for the subsequent construction of co-expression modules, reducing the interference of poor-quality data on module division and gene screening, and improving the reliability of genotyping analysis from the source.
[0037] Please refer to Figure 2 In some embodiments, the construction of several co-expression modules based on the gene expression data through a weighted gene co-expression network includes steps S201 to S204: Step S201: Perform weighted gene co-expression network analysis on the gene expression data, and determine the optimal soft threshold by evaluating the scale-free network fitting index and average connectivity. In some embodiments, the pickSoftThreshold() function of the WGCNA package in R is used to evaluate the scale-free topology fit index (R²) and mean connectivity under a series of soft thresholds. Then, by combining the fit threshold (R² ≥ 0.9) and the mean connectivity (between 20 and 200), the optimal soft threshold (e.g., 6) is determined for subsequent network construction.
[0038] It should be noted that, Figure 3 This application provides a heatmap of module-subtype correlations obtained through weighted co-expression network analysis (WGCNA), showing the correlation analysis results between module characteristic genes and each clinical subtype (WNT, SHH, Group3, Group4). Each row in the graph represents a gene co-expression module (named by different colors), and each column represents a medulloblastoma subtype (WNT, SHH, Group3, Group4). The number in each cell represents the correlation coefficient (R) between the module's characteristic gene and the corresponding clinical subtype, with the p-value in parentheses indicating significance. The intensity of the color indicates the strength of the association; for example, red indicates a positive correlation, and blue indicates a negative correlation. As shown in the figure, module MEyellow (yellow module) is significantly positively correlated with subtype G4 (high correlation coefficient, low p-value) and is therefore identified as the target module associated with subtype G4.
[0039] Step S202: Construct an adjacency matrix between genes based on the optimal soft threshold, and convert the adjacency matrix into a topological overlap matrix; In some embodiments, the blockwiseModules() function is used to construct a co-expression network, with parameters including: a soft threshold power = 6, a minimum number of genes per module of 30, and an unsigned (unweighted) network type. Then, the Pearson correlation coefficients between genes are calculated and an adjacency matrix is constructed. Subsequently, the adjacency matrix is converted into a topological overlap matrix (TOM) to enhance network robustness and reduce noise interference. The conversion process follows the standardized calculation rules of TOM in the weighted gene co-expression network analysis algorithm.
[0040] Step S203: Calculate the dissimilarity coefficients between genes based on the topological overlap matrix, and perform hierarchical clustering analysis on the dissimilarity coefficients to obtain several clustering results; In some embodiments, the dissimilarity coefficient between genes is obtained by calculating "1-TOM" based on the topological overlap matrix (TOM); then, hierarchical clustering analysis is performed on the dissimilarity coefficient, and several clustering results are generated using the hierarchical clustering tree algorithm, wherein the clustering parameters are consistent with the default settings of the blockwiseModules() function.
[0041] Step S204: The dynamic tree cutting algorithm is used to divide the clustering results into modules to obtain several partitioning results, and the partitioning results are merged to obtain several co-expression modules.
[0042] In some embodiments, a dynamic tree cut algorithm is used to divide all clustering results into modules to obtain several initial modules; and several initial modules are merged by merging similar modules (based on the correlation of module feature genes) to obtain several final co-expression modules, and each co-expression module is assigned a unique color label to distinguish them.
[0043] It should be noted that when dividing the modules, the minimum number of genes in a module can be set to 30 (that is, a single co-expression module contains at least 30 genes).
[0044] In some embodiments, the correlation score includes a correlation coefficient and a significance P-value. The step of calculating the characteristic genes of each co-expression module and, based on the characteristic genes and the correlation score, selecting target co-expression modules related to the sample to be classified specifically involves: performing principal component analysis on each co-expression module to determine the first principal component of the gene expression data in each co-expression module as the characteristic gene corresponding to each co-expression module; calculating the correlation coefficient and corresponding significance P-value between each characteristic gene and the clinical subtype information; and determining the target co-expression module related to the sample to be classified when the correlation coefficient and the significance P-value meet a second preset condition. Specifically, firstly, principal component analysis is performed on all gene expression data contained in each co-expression module. The first principal component that can explain the gene expression variation of the module is identified as the characteristic gene of the module (i.e., module characteristic gene ME), which serves as a representative indicator of the overall expression pattern of the co-expression module. Subsequently, Pearson correlation analysis is used to calculate the correlation coefficient (R) and corresponding significance P-value between the characteristic gene of each module and clinical subtype information (including subtype classifications such as medulloblastoma G3 and G4). It is then determined whether the characteristic gene and the correlation score meet the second preset condition to screen target co-expression modules. The second preset condition is that the absolute value of the correlation coefficient |R|>0.5 and the significance P-value<0.05. When the correlation coefficient and significance P-value between the characteristic gene of a co-expression module and the clinical subtype information meet the above conditions, the module is identified as a target co-expression module related to the sample to be classified (G3 and G4 subtypes).
[0045] This approach of using weighted gene co-expression network analysis to construct modules allows functionally related genes to be clustered into modules, rather than analyzing individual genes in isolation, which is more in line with the holistic principles of biological regulation.
[0046] In some embodiments, target module genes are extracted from the target co-expression module. Specifically, after obtaining the target co-expression module, all genes contained in the target co-expression module need to be directly extracted as target module genes. For example, for the target co-expression module related to the G3 subtype, 673 upregulated genes and 281 downregulated genes can be extracted; for the target co-expression module related to the G4 subtype, 2714 upregulated genes and 2304 downregulated genes can be extracted. The above division of upregulated / downregulated genes is determined based on their expression trends in the corresponding subtypes.
[0047] Step S103: Perform differentially expressed gene analysis on the sample to be genotyped to screen out differentially expressed genes that meet the first preset conditions, determine subtype candidate genes based on the target module gene and the differentially expressed genes, and genotype the sample to be genotyped for medulloblastoma based on the status of the FHDC1 gene among the subtype candidate genes.
[0048] In some embodiments, the differentially expressed gene analysis of the samples to be genotyped to screen out differentially expressed genes that meet a first preset condition specifically involves: dividing the samples to be genotyped into at least two groups based on the clinical subtype information; using a differential expression analysis tool to perform gene expression differential analysis on the different groups of samples to be genotyped to calculate the fold change and significance P-value of each target module gene between the groups of samples to be genotyped; performing multiple corrections on the significance P-value to obtain the false detection rate, and screening out genes whose fold change and false detection rate meet the first preset condition as differentially expressed genes. Specifically, firstly, based on the clinical subtype information (such as the Subgroup field), samples of non-target subtypes such as WNT and SHH are excluded, and only samples of medulloblastoma G3 and G4 subtypes are retained to form two groups of samples to be classified (G3 group and G4 group); then, the DESeq2 package in R language is used as a differential expression analysis tool, and the integer gene expression matrix and the above-grouped clinical subtype annotation information are input to construct a DESeqDataSet object, and the DESeq() function is executed to fit the model; then, the following are calculated for each target module gene through the above model: (1) expression fold change (corresponding to log2FoldChange, i.e., the logarithmic difference of gene expression between the two groups by 2); (2) significance P value (original difference significance); then, the Benjamini-Hochberg method is used to perform multiple correction on the significance P value to obtain the false discovery rate (i.e., padj, FDR), and genes that meet the first preset condition of expression fold change and false discovery rate are selected as differentially expressed genes.
[0049] It should be noted that, Figure 4 and Figure 5 The volcano plots provided in this application represent differentially expressed genes between the G3 and G4 subtypes. These two volcano plots are used to screen differentially expressed genes (DEGs) and construct a candidate gene set. Each point in the plot represents a gene. The horizontal axis (Log2 Fold Change) represents the fold change in gene expression between G4 and G3 (positive values indicate high expression in G4, and negative values indicate high expression in G3). The vertical axis (-Log10 Adj.p.Val) represents the negative logarithm of the significance p-value after multiple correction; a larger value indicates greater significance. The two dashed lines in the plot represent the preset significance threshold (FDR < 0.01) and expression change threshold (|Log2FC| > 0.8), respectively. Genes that meet these two conditions are concentrated in the upper left and upper right quadrants of the plot; that is, the scatter points located in the upper left and upper right quadrants are the screened differentially expressed genes (DEGs).
[0050] It should be noted that the model is based on the negative binomial distribution and is used to evaluate the expression difference of each target module gene between the G3 group and the G4 group. The output indicators include: (1) log2FoldChange: the log-2 fold difference in gene expression between the two groups; (2) p value: the original significance of the difference; (3) padj (FDR): the false discovery rate after Benjamini-Hochberg correction.
[0051] It should be noted that the first preset condition is "false detection rate (FDR) < 0.01 and |log2FoldChange| > 0.8", which requires that the difference is statistically significant and the fold change in expression is not less than 1.74 (2^0.8 ≈ 1.74). FDR is mainly used to control the false positive rate in multiple hypothesis testing, while |log2FoldChange| is used to measure the magnitude of expression change. Together, they ensure that the screened genes have biological significance and statistical reliability.
[0052] By using differential expression analysis, we can focus on genes that show significant changes in expression among different subtypes. Combined with the target module genes screened in the previous steps, we can determine candidate genes for subtypes through intersection. This can further narrow down the scope to core genes that are both related to the subtype and show significant differences among subtypes, overcoming the limitation of single immunohistochemical markers.
[0053] In some embodiments, subtype candidate genes are determined based on the target module genes and the differentially expressed genes. Specifically, firstly, the differentially expressed genes are exported in CSV format, and the target module genes extracted from the target co-expression module are subjected to intersection analysis with the differentially expressed genes that meet the first preset conditions and are screened through differential expression analysis. Genes common to both are obtained by matching gene IDs, and the common genes are determined as subtype candidate genes. This dual screening of co-expression features and differential expression features improves the specificity and reliability of candidate genes. For example, the intersection results are as follows: 22 intersecting genes were identified for the G3 subtype; 46 intersecting genes were identified for the G4 subtype.
[0054] The candidate genes screened by these co-expression and differential analysis results form the basis for the subsequent functional enrichment and prognostic biomarker verification of this invention, providing important support for in-depth analysis of the molecular mechanisms and precise subtyping of medulloblastoma subtypes.
[0055] In some embodiments, the step of typing medulloblastoma samples based on the status of the FHDC1 gene among the subtype candidate genes specifically involves: obtaining the expression level of the FHDC1 gene among the subtype candidate genes; comparing the expression level with a preset expression threshold, wherein the preset expression threshold is determined based on the fold change in FHDC1 gene expression among different subtype sample groups by differential expression gene analysis; when the expression level is higher than or equal to the preset expression threshold, the subtype sample is determined to be the G4 subtype, and when the expression level is lower than the preset expression threshold, the subtype sample is determined to be the G3 subtype. Specifically, to further verify whether the subtype-specific candidate genes screened in the aforementioned co-expression network analysis and differential expression analysis have potential clinical relevance, gene mutation analysis of the candidate genes is required. First, the MAF file is imported using the read.maf() function in the maftools package, and the subtype to which the sample belongs is identified in conjunction with clinical information. Subsequently, the expression level and somatic mutation information of the FHDC1 gene in the candidate genes of the subtype are obtained, and its mutation frequency in Group 3 (G3) and Group 4 (G4) subtypes (i.e., the proportion of mutated samples in the total number of samples in this subtype) is calculated to construct a contingency table. To compare the mutation differences between the two subtypes, a 2×2 contingency table (mutated / unmutated × G3 / G4) is constructed, and Fisher's exact test is used to calculate the odds ratio (OddsRatio) of each candidate gene in G3 and G4. The expression level is compared with a preset expression threshold (OR value and P value). Simultaneously, it is determined whether a somatic mutation exists in the FHDC1 gene. When the expression level is higher than or equal to the preset expression threshold and a somatic mutation exists in the FHDC1 gene, the sample to be genotyped is determined to be G4 subtype. When the expression level is lower than the preset expression threshold and no somatic mutation is detected in the FHDC1 gene, the sample to be genotyped is determined to be G3 subtype. When there is an inconsistency between the expression level and the mutation status (e.g., the expression level meets the threshold but there is no mutation, or the expression level does not meet the threshold but there is a mutation), the enrichment trend (OR value and P value) of FHDC1 in the G4 subtype in mutation analysis is used for auxiliary determination to improve the accuracy of genotyping.
[0056] It should be noted that the somatic mutation information is derived from the whole genome sequencing data of the sample to be genotyped, and the mutation annotation file is obtained after standardization processing.
[0057] It should be noted that the preset expression threshold is determined based on the fold change in the expression of the FHDC1 gene between the G3 and G4 subtypes in differential expression gene analysis.
[0058] It should be noted that the analysis results show that most candidate genes that are highly co-expressed and significantly differentially expressed in G3, such as CACNA1A and RP1, exhibit higher mutation frequencies in the G3 subtype; while genes such as ANKRD30B, ZNF804A, and FHDC1 show significant enrichment in the G4 subtype. This result, to some extent, verifies the correlation between dysregulation and somatic mutations, suggesting that these genes may play an important role in the occurrence and maintenance of subtypes.
[0059] This FHDC1 gene status-based typing method can significantly reduce the subjectivity and misjudgment rate of traditional typing, achieving accurate differentiation.
[0060] In some embodiments, after typing the medulloblastoma samples to be genotyped, the method further includes: obtaining the corresponding FHDC1 gene expression level based on the typing result, determining the prognostic cutoff value of the FHDC1 gene expression level, and dividing patients into a high expression group and a low expression group based on the cutoff value; plotting survival curves for the high expression group and the low expression group respectively, and assessing the prognostic risk of patients based on the survival curves. Specifically, FHDC1 gene expression data were extracted from the samples to be subtyped (identified as G3 or G4 subtypes), and corresponding patient clinical follow-up information (including survival status and survival time) was collected. Then, the `maxstat.test()` function in the `maxstat` package was used to calculate the optimal prognostic cutpoint for FHDC1 gene expression based on the maximum log-rank statistic. Based on this cutpoint, patients were divided into a high-expression group and a low-expression group. Next, the Kaplan-Meier method was used to plot the overall survival curves for both groups, and the significance P-value of the survival difference between the two groups was calculated using the Log-Rank test. When the survival curve of the low-expression group was significantly lower than that of the high-expression group (P<0.05), the prognostic risk of the low-expression group was considered increased (shorter overall survival); conversely, the prognostic risk was considered lower.
[0061] It should be noted that among all candidate genes, the expression level of the FHDC1 gene was significantly associated with patient prognosis, with patients expressing low levels of FHDC1 showing a significantly shorter overall survival (P<0.05). Furthermore, this gene also exhibited a higher mutation frequency in the Group 4 subtype in the aforementioned mutation analysis, suggesting that it may participate in the subtype differentiation and disease progression of medulloblastoma through abnormal expression and gene mutation. This finding provides strong support for FHDC1 as a potential prognostic biomarker and lays the foundation for subsequent mechanistic studies and personalized intervention strategies.
[0062] This "subtype-prognosis" linkage verification reduces subjective misjudgment, provides key evidence for prognostic assessment, helps clinicians predict the risk of disease progression, and thus develop better treatment plans.
[0063] This invention, by acquiring gene expression data and clinical subtype information, can simultaneously incorporate molecular-level gene expression data and clinical phenotypic information, overcoming the limitations of traditional pathology relying solely on morphological observation and providing multi-dimensional reference for subsequent analysis. The use of weighted gene co-expression network analysis to construct modules allows functionally related genes to be clustered into modules, rather than analyzing individual genes in isolation, which better aligns with the holistic principles of biological regulation. By screening target co-expression modules through correlation scoring, it can accurately capture gene sets highly correlated with the core differences in the G3 / G4 subtype, eliminating interference from irrelevant genes and significantly improving the specificity of candidate genes. Differential expression analysis can focus on genes with significant expression changes between different subtypes. Combined with the target module genes screened in previous steps, the intersection of these genes determines candidate genes for the subtype, further narrowing the scope to core genes that are both related to the subtype and show significant differences between subtypes, overcoming the limitations of single immunohistochemical markers. Based on FHDC1 gene status typing, it can significantly reduce the subjectivity and misjudgment rate of traditional typing, achieving accurate differentiation. Compared with existing technologies, this application can improve the accuracy of medulloblastoma sample typing.
[0064] like Figure 5 As shown, based on the above method embodiments, corresponding apparatus embodiments are provided; One embodiment of the present invention provides a sample typing system for medulloblastoma, comprising: an acquisition module 100, a screening module 200, and a typing module 300; The acquisition module 100 is used to acquire gene expression data and clinical subtype information of medulloblastoma patients. The screening module 200 is used to construct several co-expression modules based on the gene expression data through a weighted gene co-expression network, calculate the characteristic genes of each co-expression module, screen out target co-expression modules related to the sample to be classified based on the characteristic genes and the correlation score, and extract target module genes from the target co-expression modules. The correlation score is the correlation score between the characteristic genes and the clinical subtype information. The sample to be classified includes G3 subtype and G4 subtype. The typing module 300 is used to perform differentially expressed gene analysis on the sample to be typed, to screen out differentially expressed genes that meet the first preset conditions, and to determine subtype candidate genes based on the target module gene and the differentially expressed genes, and to type the medulloblastoma sample based on the status of the FHDC1 gene in the subtype candidate genes.
[0065] It is understood that the above-described device embodiments correspond to the method embodiments of the present invention, and can implement the medulloblastoma sample typing method provided by any of the above-described method embodiments of the present invention.
[0066] It should be noted that the device embodiments described above are merely illustrative, and some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can specifically be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.
[0067] Based on the above embodiments of the medulloblastoma sample typing method, another embodiment of the present invention provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the medulloblastoma sample typing method of any embodiment of the present invention.
[0068] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the terminal device.
[0069] The terminal device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.
[0070] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting all parts of the terminal device via various interfaces and lines.
[0071] Based on the above-described method embodiments, another embodiment of the present invention provides a computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the sample typing method for medulloblastoma described in any of the above-described method embodiments of the present invention.
[0072] The modules / units integrated in the device / terminal equipment, if implemented as software functional units and sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.
[0073] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.
Claims
1. A method for typing samples of medulloblastoma, characterized in that, include: To obtain gene expression data and clinical subtype information of medulloblastoma patients; Based on the gene expression data, several co-expression modules are constructed through a weighted gene co-expression network, and the characteristic genes of each co-expression module are calculated. Based on the characteristic genes and related scores, target co-expression modules related to the sample to be classified are screened, and target module genes are extracted from the target co-expression modules. The related score is the correlation score between the characteristic genes and the clinical subtype information. The sample to be classified includes G3 subtype and G4 subtype. Differentially expressed gene analysis is performed on the samples to be classified to screen out differentially expressed genes that meet the first preset conditions, and subtype candidate genes are determined based on the target module genes and the differentially expressed genes. Based on the status of the FHDC1 gene among the subtype candidate genes, the samples to be classified for medulloblastoma are classified.
2. The method for typing medulloblastoma samples according to claim 1, characterized in that, Based on the gene expression data, several co-expression modules are constructed through a weighted gene co-expression network, specifically as follows: Weighted gene co-expression network analysis was performed on the gene expression data, and the optimal soft threshold was determined by evaluating the scale-free network fit index and average connectivity. Based on the optimal soft threshold, an adjacency matrix between genes is constructed, and the adjacency matrix is converted into a topological overlap matrix; The dissimilarity coefficients between genes are calculated based on the topological overlap matrix, and hierarchical clustering analysis is performed on the dissimilarity coefficients to obtain several clustering results. The dynamic tree cutting algorithm is used to divide the clustering results into modules, resulting in several partitioning results. The partitioning results are then merged to obtain several co-expression modules.
3. The method for typing medulloblastoma samples according to claim 1, characterized in that, The relevant score includes a correlation coefficient and a significance P-value. The calculation of the characteristic genes of each co-expression module, and the selection of target co-expression modules related to the sample to be genotyped based on the characteristic genes and the relevant score, specifically involves: Principal component analysis is performed on each of the co-expression modules to determine the first principal component of the gene expression data in each co-expression module as the characteristic gene corresponding to each co-expression module; Calculate the correlation coefficient and corresponding significance P-value between each of the characteristic genes and the clinical subtype information. When the correlation coefficient and the significance P-value meet the second preset condition, determine the target co-expression module associated with the sample to be subtyped.
4. The method for typing medulloblastoma samples according to claim 1, characterized in that, The step of performing differentially expressed gene analysis on the sample to be genotyped to screen out differentially expressed genes that meet the first preset condition specifically includes: Based on the clinical subtype information, the samples to be classified are divided into at least two groups of samples to be classified. Differential expression analysis tools were used to perform gene expression differential analysis on the different groups of the samples to be classified, so as to calculate the fold change and significance P value of each target module gene among the different groups of the samples to be classified; Multiple corrections are applied to the significance P-value to obtain the false discovery rate, and genes whose expression fold change and false discovery rate meet the first preset condition are selected as differentially expressed genes.
5. The method for typing medulloblastoma samples according to claim 1, characterized in that, The classification of medulloblastoma samples based on the status of the FHDC1 gene among the candidate subtypes is specifically as follows: Obtain the expression level of the FHDC1 gene among the candidate genes of the subtype; The expression level is compared with a preset expression threshold, wherein the preset expression threshold is determined based on the fold change in the expression of the FHDC1 gene among different sample groups to be genotyped, as analyzed by the differentially expressed gene analysis. When the expression level is higher than or equal to the preset expression threshold, the sample to be classified is determined to be the G4 subtype; when the expression level is lower than the preset expression threshold, the sample to be classified is determined to be the G3 subtype.
6. The method for typing medulloblastoma samples according to any one of claims 1-5, characterized in that, Before constructing several co-expression modules using a weighted gene co-expression network based on the gene expression data, the method further includes: The original gene expression data and clinical information were subjected to sample quality control, and individuals with mismatched sample numbers or missing clinical information were removed to obtain the first gene expression matrix. The first gene expression matrix is converted into a numerical format to obtain the second gene expression matrix, thus obtaining preprocessed gene expression data and clinical subtype information.
7. The method for typing medulloblastoma samples according to any one of claims 1-5, characterized in that, After typing the medulloblastoma samples to be classified, the process also includes: Based on the typing results, the corresponding FHDC1 gene expression level is obtained, and the prognostic cutoff value of the FHDC1 gene expression level is determined. Based on the cutoff value, patients are divided into high expression group and low expression group. Survival curves were plotted for the high-expression group and the low-expression group, and the prognostic risk of the patients was assessed based on the survival curves.
8. A sample typing system for medulloblastoma, characterized in that, include: Acquisition module, filtering module, and classification module; The acquisition module is used to acquire gene expression data and clinical subtype information of medulloblastoma patients; The screening module is used to construct several co-expression modules based on the gene expression data through a weighted gene co-expression network, calculate the characteristic genes of each co-expression module, screen out target co-expression modules related to the sample to be classified based on the characteristic genes and the correlation score, and extract the target module gene from the target co-expression module. The correlation score is the correlation score between the characteristic gene and the clinical subtype information. The sample to be classified includes G3 subtype and G4 subtype. The typing module is used to perform differentially expressed gene analysis on the sample to be typed, to screen out differentially expressed genes that meet the first preset conditions, and to determine subtype candidate genes based on the target module gene and the differentially expressed genes, and to type the medulloblastoma sample based on the status of the FHDC1 gene in the subtype candidate genes.
9. A terminal device, characterized in that, include: One or more processors; A memory, coupled to the processor, for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the steps of the medulloblastoma sample typing method as described in any one of claims 1-7.
10. A computer-readable storage medium, characterized in that, include: A stored computer program, wherein, when the computer program is executed, it controls the device containing the computer-readable storage medium to perform the steps of the medulloblastoma sample typing method as described in any one of claims 1-7.