Use of nectin family and nectin-like molecules in low-grade glioma prognostic models

Abstract

This invention belongs to the fields of bioinformatics and biomedical technology, specifically relating to the application of the NECTIN family and NECTIN-like molecules in a prognostic model for low-grade glioma (LGG). By evaluating expression profiles in three independent databases, this invention identified four prognostic-related genes—CADM2, NECTIN1, NECTIN2, and PVR—between nectin and necls, which have significant potential as therapeutic targets and prognostic biomarkers for LGG. Furthermore, we discussed potential pathways involved in the novel four-gene risk score, finding that the risk score model plays a crucial role in the immune process of LGG. This invention provides extensive nectin and necls analysis and a four-gene model for LGG prognostic prediction, offering insights for further research on CADM2, NECTIN1 / 2, and PVR as potential clinical and immune targets for LGG, and possesses a broad clinical application scope.

Description

Technical Field

[0001] This invention belongs to the fields of bioinformatics and biomedical technology, specifically involving the application of the NECTIN family and NECTIN-like molecules in prognostic models of low-grade gliomas. Background Technology

[0002] Nectin and Nectin-like molecules (Necls) are cell adhesion molecules involved in intercellular adhesion and important cellular activities. Multiple studies have shown that Nectin and Necls play important roles in tumorigenesis and tumor immune surveillance. However, the specific roles of Nectin and Necls in low-grade gliomas remain unclear.

[0003] Low-grade gliomas (LGGs) are a common primary malignant tumor of the central nervous system, exhibiting high heterogeneity in their biological behavior. Despite treatment involving maximal surgical resection and postoperative radiotherapy and chemotherapy, most LGG cases eventually progress to treatment-resistant, highly aggressive gliomas with no significant improvement in prognosis. Therefore, identifying novel prognostic factors for LGG is crucial. Several biomarkers, such as isocitrate dehydrogenase (IDH) mutations and co-deletion of chromosome arms 1p and 19q (1p / 19q co-deletion), have been included in the 2016 WHO Classification of Brain Tumors to further subdivide patient types. However, LGG is a disease with a complex and variable genetic background, and these biomarkers cannot adequately differentiate patient prognoses. Therefore, it is urgent and important to utilize multiple advanced molecular platforms to identify novel prognostic biomarkers to improve the prognostic classification of LGG. Summary of the Invention

[0004] To address the aforementioned problems, this invention provides a novel prognostic model for overall survival in patients with low-grade gliomas.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] This invention provides a method for predicting overall survival in the prognosis of low-grade gliomas, characterized in that the method includes the following steps:

[0007] (1) Acquisition Unit: Acquire data on biomarkers related to the prognosis of patients with low-grade gliomas in the test sample;

[0008] (2) Processing unit: imports the data obtained in (1) into the pre-established prognostic model and calculates the risk score;

[0009] (3) Prediction unit: Prognostic analysis of risk scores is performed by univariate and multivariate Cox regression analysis and survival analysis to obtain prognostic parameters of low-grade glioma in the test sample.

[0010] Furthermore, the data on biomarkers related to the prognosis of patients with low-grade gliomas mentioned in step (1) are the expression levels of the NECTIN family and NECTIN-like substances.

[0011] Preferably, the NECTIN family and NECTIN class include CADM2, NECTIN1, NECTIN2 and PVR.

[0012] Further, the risk scoring formula in step (2) is: Risk score = −0.386*CADM2 + −0.659*NECTIN1 + 0.538*NECTIN2 + 0.746*PVR.

[0013] Further, the prognostic parameters in step (3) include prognostic survival risk grouping and / or survival probability values ​​for a specified number of years.

[0014] Preferably, the survival probability value for the specified number of years includes the probability values ​​for a prognostic survival period of six months, one year, and three years.

[0015] This invention also provides a method for constructing a prognostic model for overall survival in low-grade gliomas, characterized in that the method includes the following steps:

[0016] (1) Search for the relationship between the NECTIN gene family and Nectin-like molecules and the clinical features of glioma from the Cancer Genome Atlas TCGA, the Chinese Glioma Genome Atlas CGGA and the Rembrandt Atlas;

[0017] (2) Using the R package survival, patients were divided into high and low groups according to gene expression levels in the three databases mentioned in (1). P < 0.05 was used as the cutoff value for screening prognostic genes. CADM2, NECTIN1, NECTIN2 and PVR were selected as prognostic genes for constructing a prognostic model.

[0018] (3) The four prognostic-related gene models selected in (2) were established using the L1-penalized Coxproportional hazards regression method, and the LGG patients were divided into high-risk and low-risk groups based on the risk scores calculated by the gene models in the three databases mentioned in (1).

[0019] (4) Using univariate risk scores for prognostic analysis can serve as an independent prognostic factor for low-grade gliomas; the area under the curve (AUC) at multiple time points is calculated using the R package ROC to assess the discriminative ability of the prognostic model.

[0020] The present invention also provides a device for predicting the overall survival of low-grade gliomas, characterized in that the device comprises:

[0021] Memory;

[0022] Communication interface;

[0023] The processor is configured to perform the following steps:

[0024] S1 acquires data on biomarkers related to the prognosis of patients with low-grade gliomas in the test sample;

[0025] S2 imports the data obtained in S1 into the pre-established prognostic model to calculate the risk score;

[0026] S3 uses univariate and multivariate Cox regression analysis and survival analysis to perform prognostic analysis on risk scores and obtain prognostic parameters for low-grade gliomas in the test samples.

[0027] Furthermore, the processor is further configured to perform any or all of the method steps described in this invention.

[0028] The beneficial effects of the present invention compared with the prior art.

[0029] (1) This invention establishes a prognostic model with four NECTIN gene families and NECTIN-like molecules, and classifies low-grade glioma patients into high- and low-risk groups. In three different databases, the risk score of LGG patients was significantly associated with overall survival (OS) (P < 0.001). ROC curve analysis showed that AUC > 0.7 at 1-, 3-, and 5-year follow-up. The predictive performance of our gene model was validated in multiple databases.

[0030] (2) Enrichment analysis of the 1000 genes most associated with risk scores revealed that gene models of NECTIN and NECTIN-like molecules are closely related to multiple immune pathways. Using the TIMER website, we found differences in the number of macrophages and myeloid dendritic cells between high- and low-risk LGG patients. Attached Figure Description

[0031] Figure 1 Expression of NECTIN and necls genes in various types of cancer and corresponding matched normal tissues.

[0032] Figure 2 The expression of the NECTIN family and necl genes was analyzed in the TCGA, CGGA and REMBRANDT databases.

[0033] Figure 3The relationship between the NECTIN family and necl gene and the histological grading of LGG samples was analyzed in the TCGA, CGGA, and REMBRANDT databases.

[0034] Figure 4 Differential expression of the NECTIN family and necl genes is correlated with IDH mutation status and 1p19q co-deletion status. A and B represent the correlation between differential expression of the NECTIN family and necl genes and IDH status; C and D represent the correlation between differential expression of the NECTIN family and necl genes and 1p19q co-deletion status.

[0035] Figure 5 Analysis based on the UALCAN website revealed expression differences in the NECTIN family and necls based on different histological LGG subtypes.

[0036] Figure 6 Analyze the changes in the NECTIN family and necl gene in TCGA LGG patients.

[0037] Figure 7 Kaplan-Meier survival curve results for NECTIN family members and necls in the TCGA database.

[0038] Figure 8 Kaplan-Meier survival curves of NECTIN family members and necls in the CGGA database.

[0039] Figure 9 Kaplan-Meier survival curve results for NECTIN family members and necls in the REMBRANDT database.

[0040] Figure 10 Results of univariate Cox regression analysis based on three databases: TCGA, CGGA, and REMBRANDT.

[0041] Figure 11 Construct a feature scoring system for LGG based on TCGA, using the NECTIN family and necl.

[0042] Figure 12 The relationship between the distribution of risk scores and different clinical characteristics in LGG patients.

[0043] Figure 13 Enrichment analysis was performed using the 1000 genes most associated with risk scores to identify pathways related to the NECTIN family and the necl gene. A represents the enrichment data from the TCGA database; B represents the enrichment data from the CGGA database.

[0044] Figure 14 The risk scores of the NECTIN family and necls are correlated with several important immune-related pathways and immune molecules. A represents the four most representative immune-related pathways; B shows all statistically significant pathways.

[0045] Figure 15 A protein-protein interaction network of risk score-related genes was constructed using GeneMANIA.

[0046] Figure 16 The correlation between risk score and immunity, matrix and tumor purity was observed based on the TCGA cohort.

[0047] Figure 17 The correlation between risk score and immunity, matrix and tumor purity was observed based on the CGGA cohort.

[0048] Figure 18 Estimation of immune cell infiltration results from high-risk samples in the TCGA cohort.

[0049] Figure 19 Estimation of immune cell infiltration outcomes from low-risk samples in the TCGA cohort.

[0050] Figure 20 Pearson correlation analysis was used to analyze the correlation between risk scores and important immune molecules.

[0051] Figure 21 Correlation between risk score and IL2, IL8, and IFNγ. Detailed Implementation

[0052] To better understand the above-described objects, features, and advantages of the present invention, the present invention will be further described below with reference to specific embodiments. Many specific details are set forth in the following description to provide a thorough understanding of the present invention; however, the present invention may be practiced in other ways than those described herein, and therefore, the present invention is not limited to the specific embodiments disclosed in the following specification.

[0053] Example 1: Model construction method and effect verification.

[0054] 1. Materials and methods.

[0055] 1.1 Oncomine Database:

[0056] The expression levels of various NECTIN family members and necls in different types of cancer were analyzed using the Oncomine database (AnnArbon, MI, USA) (http: / / www.oncomine.org / ). Oncomine provides 715 datasets, microarray information, and several online functions. The p-values ​​for differences in NECTIN family genes between cancer specimens and normal controls were obtained using t-tests. The fold change threshold was set to 2, and the p-value was set to 0.05.

[0057] 1.2 Expression data of LGG patients:

[0058] The data of LGG patients in this invention were obtained from the Tumor Genome Atlas (TCGA) LGG cohort (downloaded from http: / / xena.ucsc.edu / ), the Chinese Glioma Genome Atlas (CGGA; http: / / www.cgga.org.cn) database, and the REMBRANDT (https: / / caintegrator.nci.nih.gov / rembrandt / ) database. The TCGA LGG cohort contains 463 tumor samples, the CGGA cohort contains 172 tumor samples, and the REMBRANDT cohort contains 107 tumor samples. The basic characteristics of the LGG patients relevant to this invention are shown in Table 1.

[0059] Table 1. Clinical characteristics of LGG patients included in the study.

[0060] .

[0061] 1.3 Study on mRNA expression and correlation of NECTIN and necls in LGG:

[0062] Using the free online data analysis platform Sanger Box (http: / / www.sangerbox.com / tool), we compared the expression of NECTIN and necls genes in LGG tissues at different levels, and analyzed the expression differences of the NECTIN family based on IDH status or 1p19q co-deletion status. We further explored the correlation between NECTIN and necls expression in LGG using the corrplot package, and visualized the results using Sanger Box.

[0063] 1.4 UALCAN Database:

[0064] The UALCAN website provides a comprehensive assessment of cancer transcriptome data (https: / / ualcan.path.uab.edu / ). The expression levels of NECTIN and necl genes in different histological subtypes of LGG were analyzed using the UALCAN database.

[0065] 1.5cBioportal Data Extraction:

[0066] The cBioportal database provides comprehensive data analysis of complex transcriptional and clinical cancer genomic atlases (TCGA) (https: / / www.cbioportal.org / ). Genetic alterations in the NECTIN family and necls (amplifications, deletions, and missense mutations) in TCGLGG patients are evaluated by cBioportal.

[0067] 1.6 Survival analysis of NECTIN family and necls in LGG patients:

[0068] The following two steps were used to assess the prognostic value of NECTIN members and necls in the TCGA LGG, CGGA LGG, and REMBRANDT cohorts: (1) Patients were divided into two groups based on median gene expression using the R package “survival”. The Kaplan-Meier method and log-rank test were used to assess the impact of expression of each nectin member and necls on overall survival (OS) in LGG patients in the three independent cohorts. (2) Univariate Cox proportional hazards regression analysis was used to assess the significant association between NECTIN family and necls with OS in the three databases. Genes with p-values ​​less than 0.05 in steps (1) and (2) in all three databases were considered prognostic genes and were used for further analysis. All analyses were performed using the R package “survival” in R3.6.3 software.

[0069] 1.7 Establishment and Statistical Analysis of Feature Scores:

[0070] The risk formula for the characteristic score was constructed based on a linear combination of the expression levels of prognostic-related genes determined by the above methods, and weighted using the corresponding regression coefficients from univariate Cox proportional hazards regression analysis. The TCGA cohort was used as the training set, and the CGGA and REMBRANDT cohorts were used as the validation set. Patients were divided into high-risk and low-risk groups according to the risk score. Differences in histological grade, IDH status, and 1p19q co-deletion status between high-risk and low-risk LGG patients were analyzed using t-tests. Kaplan-Meier survival curves were used to assess the difference in survival time between the high-risk and low-risk groups, and the log-rank test was used to detect the difference in overall survival (OS) between the two groups. The prognostic significance of the risk score model at 0.5 years, 1 year, and 3 years was evaluated using the R package "survivalROC". All statistical methods were set to a minimum of 0.05, and the results were plotted using the Sanger box online tool.

[0071] 1.8 Enrichment analysis, GSEA and PPI analysis of the functions and interactions between the NECTIN family and necls:

[0072] Pearson correlation analysis was used to obtain genes for functional annotation. The top 1000 genes associated with risk scores and p-values ​​<0.05 were then uploaded to Metascape for Gene Ontology (GO) enrichment analysis annotation and visualization (http: / / metascape.org). Furthermore, gene set enrichment analysis (GSEA, version 4.0.1; http: / / software.broadinstitute.org / gsea / index.jsp) was performed to identify pathways associated with risk score expression in TCGA LGG patients. The gene set annotation file c5.go.bp.v7.4.entrez.gmt (from the Msig database) was used as a reference. Randomization was repeated 1000 times, with p-values ​​<0.05 used as target pathways. Genes associated with risk scores were analyzed for protein-protein interaction networks using the GeneMANIA (HTTPS: / / www.genemania.org) multi-association network integration algorithm.

[0073] 1.9 Immune infiltration analysis:

[0074] Differences in immune cell abundance (including B cells, CD4+ T cells, CD8+ T cells, neutrophils, macrophages, and dendritic cells) between the high- and low-risk score groups of the TCGA and CGGA cohorts were calculated or obtained using the Tumor Immunology Estimation Resource (TIMER; https: / / cistrome.shinyapps.io / timer / ). Based on expression data, the "estimate" software package was used to analyze the immune score, stromal score, and tumor purity of each LGG patient to assess the level of immune cell infiltration and the level of stromal cells in the tumor tissue. Pearson correlation coefficients were calculated to assess the correlation between the risk score and the above three scores or the expression levels of IL2, IL8, and IFNγ.

[0075] 2. Results

[0076] 2.1 Expression patterns of the NECTIN family and necls gene in LGG patients.

[0077] First, this invention determined the chromosomal location of the NECTIN family and the necls gene. Details are shown in Table 2. Next, the ONCOMINE database was examined to compare the mRNA expression of the NECTIN family in cancerous and normal tissues across various cancers. Figure 1 As shown, the expression of NECTIN and necls genes differs across cancer types and corresponding matched normal tissues. CADM3 and CADM4 are upregulated in most cancers, while NECTIN2 is downregulated in most cancers. In brain and central nervous system tumor samples, CADM1, CADM2, NECTIN1, and NECTIN3 are downregulated, while NECTIN2 is overexpressed.

[0078] Table 2. Characteristics of the NECTIN gene family and NECTIN-like molecules

[0079] .

[0080] Subsequently, the relationship between the expression of the NECTIN family and necl genes and the histological grading of LGG samples was analyzed in the TCGA, CGGA, and REMBRANDT databases. In TCGA samples, the expression levels of CRTAM, NECTIN2, NECTIN4, and PVR were upregulated with increasing grade, while the expression levels of CADM2, CADM3, NECTIN1, and NECTIN3 were downregulated. Figure 2 A). In the CGGA samples, the expression of CRTAM, NECTIN2, and PVR was upregulated with increasing level, while the expression of CADM2 and NECTIN1 was downregulated. Figure 2B). In the REMBRANDT sample, the levels of CADM4, NECTIN2, and PVR were upregulated at higher levels, while the levels of CADM2 and NECTIN1 were downregulated. Figure 2 C). Therefore, in all three databases, NECTIN2 and PVR were upregulated, while CADM2 and NECTIN1 were downregulated. Further analysis revealed correlations among members of the NECTIN family. Figure 3 As shown in Figure A, in the TCGA LGG samples, there is a strong positive correlation between CADM2 and NECTIN1 / 3, and a strong negative correlation between NECTIN2 and either CADM2 or NECTIN1 / 3. However, in the CGGA and REMBRANDT databases, the correlation among nectin family members is not significant. Figure 3 B, Figure 3 C).

[0081] 2.2 Relationship between NECTIN family and necls gene alterations and LGG subtype, IDH1 status and 1P19q status.

[0082] Independent analysis of 635 LGG samples from the TCGA and CGGA databases revealed that differential expression of the NECTIN family and necl genes was significantly associated with IDH mutation status, 1p19q co-deletion status, and histological subtype. Results from these two databases showed that CADM1, CRTAM, NECTIN2, and PVR were highly expressed in IDH wild-type samples, while CADM2, NECTIN1, and NECTIN3 were highly expressed in IDH mutant samples. Figure 4 (A, 4B). Furthermore, CADM2 and NECTIN1 were highly expressed in 1p19q co-deleted samples, while CADM4, CRTAM, NECTIN2, and PVR were highly expressed in 1p19q non-co-deleted samples. Figure 4 C, 4D). Analysis based on the UALCAN website showed differential expression of the NECTIN family and necls based on different histological LGG subtypes. Figure 5 Secondly, we analyzed the gene changes in TCGA LGG patients to gain a deeper understanding of the molecular mechanisms underlying differential expression of the NECTIN family and necl genes. For example... Figure 6 As shown, NECTIN1 showed the highest probability of alteration (10%), followed by PVR (8%), while NECTIN4 showed the lowest probability of alteration (2.8%). These results indicate that NECTIN family gene expression is closely related to a range of important clinical parameters.

[0083] 2.3 Identification of four prognostic-related NECTIN family and necls genes in LGG samples

[0084] The prognostic value of NECTIN family members and necls was then analyzed. Kaplan-Meier survival curves for all NECTIN family members and necls in the TCGA, CGGA, and REMBRANDT databases are shown below. Figure 7 , 8 9. Independent analysis of the three databases revealed that patients with high expression of NECTIN2 and PVR, and low expression of CADM2 and NECTIN1 had a poorer prognosis (P<0.05); Figure 7 A, 8A, 9). Univariate Cox regression analysis based on three databases showed that CADM2, NECTIN1, NECTIN2, and PVR were associated with overall survival (OS, p < 0.05). Figure 10 (A, B, C). Therefore, this invention identified the above four genes as prognostic-related genes of the NECTIN family and necl in LGG samples.

[0085] 2.4 Constructing a TCGA-based LGG feature scoring system based on the NECTIN family and necl

[0086] After independently screening four prognostic-related genes from three databases, a risk score formula was constructed based on the expression levels of the four genes; the regression coefficient was: Risk Score = −0.386*CADM2 + −0.659*NECTIN1 + 0.538*NECTIN2 + 0.746*PVR. ​​All LGG patients were divided into high-risk and low-risk groups according to the median risk score. Kaplan-Meier survival analysis showed that, based on the TCGA training set (log-rank < 0.0001) and two validation cohorts (CGGA and REMBRANDT, both with log-rank p < 0.0001), the risk score was significantly higher than that of LGG patients. Figure 11 A, 11E, 11I). Furthermore, the risk scoring model demonstrates better ability to predict 0.5-year, 1-year, and 3-year survival times, with corresponding area under the ROC curve values ​​of 0.733, 0.801, and 0.801 in the TCGA training cohort, respectively. Figure 11 B~11D), in the CGGA verification queue 0.752, 0.793 and 0.765 ( Figure 11 F, 11H), in the REMBRANDT validation cohort, were 0.69, 0.729, and 0.757, respectively, indicating that NECTIN family markers have a strong role in predicting overall survival in LGG patients. Figure 11Next, based on the clinical characteristics of all patients in the TCGA and CGGA cohorts, univariate and multivariate Cox analyses were used to verify whether the risk scoring model could serve as an independent prognostic factor. We found that the risk scoring model was significantly associated with overall survival, with HR=1.22 (95% confidence interval [CI] = 1.02-1.46; p=0.03) in the TCGA training set and HR=1.50 (95% CI = 1.28-1.84; p<0.001) in the CGGA validation set (Table 3).

[0087] Table 3. Results of COX analysis on the TCGA training set and CGGA validation set.

[0088] .

[0089] 2.5 Clinical characteristics and pathways related to risk scores in LGG patients

[0090] First, this invention explores the relationship between the distribution of risk scores and different clinical characteristics in LGG patients. Patients with higher risk scores also have higher malignancy of gliomas; for example, these patients have higher LGG grades, wild-type IDH status, and no 1p19q deletion. Figure 12 These results were also verified in AC). Figure 12 DF).

[0091] Subsequently, to explore biological pathways associated with risk scores, we first attempted to identify the top 1000 genes in TCGA and CGGA most relevant to our risk scores, as shown in Tables 4 and 5. These differentially expressed genes were then enriched using Metascape (http: / / metascape.org). Interestingly, in addition to functions such as localization and bioadhesion mentioned in many studies, immune system pathways such as immune effector pathways, T cell activation, and cytokine signal transduction in the immune system were also associated with the NECTIN family and necl genes. Figure 13 A, B). Meanwhile, gene set enrichment analysis (GSEA) showed that the risk scores of the NECTIN family and necls are associated with several important immune-related pathways and immune molecules, such as innate immune responses, lymphocyte activation, IL2, IL8, and interferon (…). Figure 14 A, B). Furthermore, this invention also utilizes GeneMANIA to construct a protein-protein interaction network of risk score-related genes (A, B). Figure 15This study aimed to explore the potential interactions between them. SRC, an important tyrosine kinase primarily involved in immune regulation, is one of the neighboring genes predicted to interact with four signature genes, further revealing the deep connection between the NECTIN family and the necl signature and immune processes.

[0092] Table 4. 998 genes most associated with risk scores in the TCGA database

[0093] .

[0094] Table 5. 999 genes most associated with risk scores in the CGGA database.

[0095] .

[0096] 2.6 Relationship between Nectin family risk score and immune infiltration.

[0097] Using the ESTIMATE method, we observed the correlation between risk scores and immunity, matrix, and tumor purity based on TCGA and CGGA cohorts, respectively. Risk scores showed a strong positive correlation with immunity and matrix scores, and a negative correlation with tumor purity, indicating that high-risk patients have a more complex immune microenvironment. Figure 16 , Figure 17 Next, the tumor immune estimation resource website (https: / / cistrome.shinyapps.io / timer / ) was used for each sample in the TCGA cohort to estimate immune cell infiltration. Figure 18 As shown, the high-risk group had a higher number of macrophages and myeloid dendritic cells. In contrast, the low-risk group had a higher number of B cells and neutrophils. Figure 19 Pearson correlation analysis showed the correlation between risk scores and key immune molecules. Figure 20 The results showed that the risk score was positively correlated with IL2, IL8, and IFNγ. Figure 21 These data indicate that the NECTIN family and necl marker-related genes play an important role in the immune regulation of LGG.

[0098] In summary, by evaluating expression profiles in three independent databases, this invention identified four prognostic-related genes (CADM2, NECTIN1, NECTIN2, and PVR) between nectin and necls, which have significant potential as therapeutic targets and prognostic biomarkers for LGG. Furthermore, we discussed potential pathways involved in the novel four-gene risk scoring system, revealing that the risk scoring model plays a crucial role in the immune process of LGG.

Claims

1. A method for predicting overall survival in the prognosis of low-grade gliomas, characterized in that, The method includes the following steps: (1) Acquisition Unit: Acquire data on biomarkers related to the prognosis of patients with low-grade glioma in the sample to be tested; the biomarker data are the expression levels of the NECTIN family and NECTIN-like substances; the NECTIN family and NECTIN-like substances include CADM2, NECTIN1, NECTIN2 and PVR; (2) Processing unit: imports the data obtained in (1) into the pre-established prognostic model and calculates the risk score; the risk score = −0.386*CADM2+−0.659*NECTIN1+0.538*NECTIN2+0.746*PVR; (3) Prediction unit: Prognostic analysis of risk scores is performed by univariate and multivariate Cox regression analysis and survival analysis to obtain prognostic parameters of low-grade glioma in the test sample.

2. The method for predicting overall survival of low-grade glioma according to claim 1, characterized in that, The prognostic parameters mentioned in step (3) include prognostic survival risk grouping and / or survival probability values ​​for a specified number of years.

3. The method according to claim 2, characterized in that, The survival probability values ​​for the specified number of years include the probability values ​​for a prognostic survival period of six months, one year, and three years.

4. A method for constructing a prognostic model for overall survival in low-grade gliomas, characterized in that, The method includes the following steps: (1) Search for the relationship between the NECTIN gene family and Nectin-like molecules and the clinical features of glioma from the Cancer Genome Atlas TCGA, the Chinese Glioma Genome Atlas CGGA and the Rembrandt Atlas; (2) Using the R package survival, patients were divided into high and low groups according to gene expression levels in the three databases in (1). P<0.05 was used as the cutoff value for screening prognostic genes. CADM2, NECTIN1, NECTIN2 and PVR were selected as prognostic genes and used to construct a prognostic model. (3) The four prognostic-related gene models selected in (2) were established using the L1-penalized Coxproportional hazards regression method, and the LGG patients were divided into high-risk and low-risk groups based on the risk scores calculated by the gene models in the three databases mentioned in (1). (4) Use univariate risk scores for prognostic analysis as an independent prognostic factor for low-grade gliomas; use R package ROC to calculate the area under the curve (AUC) at multiple time points to assess the discriminative ability of the prognostic model.

5. A device for predicting overall survival in the prognosis of low-grade gliomas, characterized in that, The device includes: a memory; a communication interface; and the processor is configured to perform the method as described in any one of claims 1-3.

Images