A gene set, system and method for predicting the immune microenvironment, prognosis and treatment sensitivity of glioblastoma

Abstract

A gene set, system and method for predicting the immune microenvironment, prognosis and treatment sensitivity of glioblastoma, belonging to the field of biomedicine. For the first time, a gene set and a standard risk scoring formula related to T cell activation were established in glioblastoma. A system for predicting the immune microenvironment and prognosis of glioblastoma was established based on the standard risk scoring formula, which can well evaluate the prognosis and immune microenvironment of GBM patients. The establishment of this system can predict the survival time of patients, and it has been well verified in other databases and has clinical transformation value. Further, based on the T cell activation-related risk score and the T cell initiation activation score, a calculation formula for the immunotherapy score IS of glioblastoma was established. Based on the immunotherapy score IS, a system for predicting the treatment sensitivity of glioblastoma was established, which can well predict the effect of immunotherapy on GBM patients, and the results have been stably verified in multiple databases.

Description

Technical Field

[0001] The present invention belongs to the field of biomedicine, and specifically relates to a gene set, system and method for predicting the immune microenvironment, prognosis and treatment sensitivity of glioblastoma. Background Art

[0002] Glioma is the most common malignant tumor among primary CNS tumors. It can be divided into low-grade glioma (LGG, WHO grade II-III) and high-grade glioma (GBM, WHO grade IV) according to its pathological classification. Among them, glioblastoma has the highest degree of malignancy and the worst prognosis. It has the characteristics of rapid progression, strong tumor proliferation and invasion, and easy recurrence. Even if the patient undergoes surgery and is assisted by standard radiotherapy and chemotherapy, the median survival period is only about 14 months. Due to the occurrence of tumors, chemokines secreted by tumor tissues, angiogenesis in tumor tissues, and the destruction of the blood-brain barrier by tumors, a large number of immune cells appear in gliomas. Most of these immune cells that infiltrate into tumor tissues have lost their original tumor-killing function, and even highly express inhibitory immune checkpoints to inhibit the killing of tumor cells by immune cells.

[0003] There is obvious heterogeneity in the immune microenvironment between patients with different grades of gliomas and between patients with different gliomas of the same grade, and the study of the tumor microenvironment is the key to personalized tumor treatment. Although the main immune cells in the brain are macrophages and microglia, more and more studies have shown that T cells are closely related to the development of gliomas. As an important immune cell, T cells are mainly involved in adaptive immune responses. Different subtypes of T cells have different effects in gliomas. For example, T-reg cells and Th2 cells are considered to be cells that inhibit anti-tumor immunity, while CD8+T cells and CD4+T cells are cytotoxic anti-tumor immune cells. In fact, the more complex the immune microenvironment in gliomas, the worse the prognosis. Even tumor-killing T cells may have their immune function suppressed and not have tumor-killing function.

[0004] In recent years, the research on immune checkpoint blockade such as anti-PD1, PD-L1, CTLA4 as therapeutic targets has been increasing, and significant effects have been achieved in some tumors such as melanoma, breast cancer, and bladder cancer. However, due to the unclear effect of immunotherapy in GBM and its toxic effects on the brain, the research on GBM immunotherapy is limited. Most studies can only be limited to the basic level, so it is impossible to obtain large-scale immunotherapy data for GBM patients. T cell activation is closely related to the tumor microenvironment, immunotherapy, and prognosis of glioma. Many studies are underway on how to activate T cell tumor-killing function and explore the causes of T cell immunosuppression in tumors. However, the overall activation of T cells in GBM and prognosis, immune microenvironment, and immunotherapy sensitivity are still unclear. Summary of the invention

[0005] In view of the defects of the prior art, the present invention mainly provides a gene set, system and method for predicting the immune microenvironment, prognosis and treatment sensitivity of glioblastoma. By calculating the T cell activation risk score as an independent risk factor for adverse prognosis, the prognosis of GBM patients can be well evaluated. The immunotherapy score obtained by combining the risk score with the T cell activation score can evaluate the sensitivity of immunotherapy for GBM patients and provide an auxiliary judgment tool for immunotherapy for GBM patients.

[0006] In order to achieve the above-mentioned invention object, the technical solution adopted by the present invention is as follows:

[0007] A gene set for predicting the immune microenvironment, prognosis and treatment sensitivity of glioblastoma, the gene set comprising 5 genes, CD276, IL-15, SLC11A1, TNFSF4 and TREML2.

[0008] A system for predicting the immune microenvironment and prognosis of glioblastoma, the system comprising an input module, an analysis module and an output module, wherein the analysis module is connected to the input module and the output module via data communication;

[0009] The input module is used to input transcriptome expression data of glioblastoma samples, and the transcriptome expression data is mRNA-seq data of glioblastoma samples.

[0010] The analysis module calculates the standardized risk score RS* based on the gene expression data of the glioblastoma sample using Formula I and predicts the immune microenvironment and prognosis of the glioblastoma sample according to the value of the standardized risk score:

[0011] RS * =(RS-8.7401) / 4.4757 (I)

[0012] Among them, RS is the risk score associated with T cell activation, which is calculated using formula II:

[0013] RS = (0.5258×E CD276 )+(0.1477×E IL-15 )+(0.2178×E SLC11A1 )+(0.2346×E TNFSF4 )+(0.2192×E TREML2 ); (II)

[0014] Among them, E CD276 、E IL-15 、E SLC11A1 、E TNFSF4 and E TREML2 They are the expression levels of genes CD276, IL-15, SLC11A1, TNFSF4, and TREML2 in glioblastoma samples;

[0015] The output module is used to output the immune microenvironment and prognosis of glioblastoma samples.

[0016] A system for predicting the treatment sensitivity of glioblastoma, the system comprising an input module, an analysis module and an output module, wherein the analysis module is connected to the input module and the output module by data communication;

[0017] The input module is used to input gene expression data of glioblastoma samples;

[0018] The analysis module calculates the immunotherapy score IS based on the gene expression data of the glioblastoma sample using Formula III and predicts the treatment sensitivity of the glioblastoma sample according to the immunotherapy score:

[0019] IS=Z-score(x)=(x-μ) / δ(III)

[0020] Where x = PS * -RS * , μ is 0.0925, δ is 0.1571;

[0021] PS * The normalized activation score was calculated using Formula IV

[0022] PS * =(PS-0.0502) / 0.4357 (IV)

[0023] Among them, PS is the T cell activation score. GO:0002291 was used as the gene set, and the glioblastoma sample sequencing data was calculated using the ssGSEA algorithm in R 3.6.3. The GO:0002291 gene set includes eight genes, including LILRB1, HLA-DMB, ICAM1, ITGAL, LGALS3, LGALS9, APBB1IP and HAVCR2.

[0024] RS * is the standardized risk score RS* of glioblastoma samples, calculated using formula I.

[0025] The output module is used to output the treatment sensitivity of the glioblastoma sample.

[0026] A gene set prediction method for glioblastoma immune microenvironment, prognosis and treatment sensitivity, using the following steps:

[0027] A) Obtain glioblastoma samples, and perform RNA sequencing on the glioblastoma samples to obtain transcriptome expression data of glioblastoma patients, that is, perform mRNA sequencing on the glioblastoma samples to obtain mRNA expression profiles, and express the obtained mRNA expression profiles in FPKM format.

[0028] B) Calculation of the standardized risk score RS* for glioblastoma samples:

[0029] RS * =(RS-8.7401) / 4.4757 (I)

[0030] RS is the risk score associated with T cell activation, which is calculated using the following formula:

[0031] RS = (0.5258×E CD276 )+(0.1477×E IL-15 )+(0.2178×E SLC11A1 )+(0.2346×E TNFSF4 )+(0.2192×E TREML2 ) (II);

[0032] Among them, E CD276 、E IL-15 、E SLC11A1 、E TNFSF4 and E TREML2 The expression levels of genes CD276, IL-15, SLC11A1, TNFSF4, and TREML2 in glioblastoma samples, respectively;

[0033] C) Calculation of the immunotherapy score IS for glioblastoma samples:

[0034] IS=Z-score(x)=(x-μ) / δ(III)

[0035] Where x = PS * -RS * , μ is 0.0925, δ is 0.1571;

[0036] PS * The normalized activation score was calculated using Formula IV

[0037] PS * =(PS-0.0502) / 0.4357 (IV)

[0038] Among them, PS is the T cell priming activation score, which was calculated using the ssGSEA algorithm in R 3.6.3 using GO:0002291 as the gene set for the glioblastoma sample sequencing data;

[0039] RS * The standardized risk score RS* was calculated using formula I;

[0040] D) The immune microenvironment, prognosis and treatment sensitivity of glioblastoma samples were predicted based on the calculation results of the standardized risk score RS* and the immunotherapy score IS.

[0041] The gene set is used to predict the immune microenvironment, prognosis and treatment sensitivity of glioma patients.

[0042] The input module, analysis module and output module adopt computer operable devices, including but not limited to computer memory, computer processor, computer output display terminal, keyboard for inputting data, interface for importing and extracting data and any hardware and software parts for realizing the functions.

[0043] Compared with the prior art, the present invention has the following advantages and beneficial effects: for the first time, a gene set and a standard risk scoring formula related to T cell activation are established in glioblastoma. A system for predicting the immune microenvironment and prognosis of glioblastoma is established according to the standard risk scoring formula, which can well evaluate the prognosis and immune microenvironment of GBM patients. The establishment of this system can predict the survival time of patients, and has been well verified in other databases, and has clinical transformation value. Further, according to the risk score related to T cell activation and the T cell activation score, a calculation formula for the immunotherapy score IS of glioblastoma is established, and according to the immunotherapy score IS, a system for predicting the sensitivity of glioblastoma treatment is established. The system can well predict the effect of immunotherapy of GBM patients, and the results are stably verified in multiple databases. It has predictive value for the immunotherapy sensitivity of GBM patients and provides a theoretical basis for GBM immunotherapy. The two prediction systems established by risk scoring and immunotherapy scoring use only 5 and 13 genes respectively to evaluate the prognosis of GBM patients and predict the responsiveness to immunotherapy, which greatly reduces the number of genes compared with hundreds of genes, and has extremely high application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figures 1A-1E This is a graph showing the relationship between T cell activation risk score and sample prognosis in the TCGA database. A is the survival curve of the high and low risk score groups, B is the univariate regression analysis graph, C is the multivariate regression analysis graph, D is the 0.5-2 year survival rate nomogram of the sample, and E is a comparison graph of the predicted survival probability and the actual survival rate.

[0045] Figure 2A-2C It is a relationship diagram between immunotherapy score and immunotherapy sensitivity of samples in TCGA database. Among them, A is the correlation curve of activation score, risk score and immunotherapy sensitivity, B is the bar chart of activation score, risk score and immunotherapy sensitivity, and C is the comparison chart of immunotherapy responsiveness of patients in high and low immunotherapy score groups.

[0046] Figures 3A-3H: This is a graph showing the relationship between the T cell activation risk score and the prognosis of GBM samples in the validation group database. A is a univariate regression analysis graph of samples in the CGGA325 database, B is a multivariate regression analysis graph of samples in the CGGA325 database, C is a univariate regression analysis graph of samples in the CGGA693 database, D is a multivariate regression analysis graph of samples in the CGGA693 database, E is a univariate regression analysis graph of samples in the GSE16011 database, F is a multivariate regression analysis graph of samples in the GSE16011 database, G is a comparison graph of the predicted survival probability and the actual survival rate in the CGGA325 database, and H is a comparison graph of the predicted survival probability and the actual survival rate in the CGGA693 database.

[0047] Figures 4A-4F It is the relationship between the T cell activation-related risk score and the immune microenvironment of GBM samples in the TCGA database. Among them, A is the relationship between the T cell activation-related risk score and tumor purity, B is the relationship between the T cell activation-related risk score and the matrix score, C is the relationship between the T cell activation-related risk score and the immune score, D is the relationship between the T cell activation-related risk score and the number of immune cells, E is the relationship between the T cell activation-related risk score and the anti-tumor immune score and the tumor-promoting immunosuppression score, and F is the relationship between the T cell activation-related risk score and the immune cell content.

[0048] Figures 5A-5F It is a relationship diagram between immunotherapy score and immunotherapy sensitivity of samples in the validation group database. Among them, A is a correlation curve diagram of activation score, risk score and immunotherapy sensitivity in the CGGA325 database, B is a correlation curve diagram of activation score, risk score and immunotherapy sensitivity in the CGGA693 database, C is a bar graph of activation score, risk score and immunotherapy sensitivity in the CGGA325 database, D is a bar graph of activation score, risk score and immunotherapy sensitivity in the CGGA693 database, E is a comparison diagram of immunotherapy responsiveness of patients with high and low immunotherapy score groups in the CGGA325 database, and F is a comparison diagram of immunotherapy responsiveness of patients with high and low immunotherapy score groups in the CGGA693 database. DETAILED DESCRIPTION

[0049] The present invention is further described in detail below by specific examples and accompanying drawings. The following examples are only further described for the present invention and should not be construed as limiting the present invention. The experimental methods in the following examples are conventional methods unless otherwise specified, and the experimental reagents and materials involved are conventional biochemical reagents and materials unless otherwise specified. The database used in the specific examples includes:

[0050] ① Discovery group: 149 samples from TCGA RNA-seq cohort (https: / / tcga-data.nci.nih.gov) database. ② Validation group: 137 samples from CGGA325 RNA-seq cohort (http: / / www.cgga.org.cn) database, 237 samples from CGGA693 RNA-seq cohort (http: / / www.cgga.org.cn) database, 149 samples from GSE16011 microarray (https: / / www.ncbi.nlm.nih.gov / geo / ) database, including clinical characteristics of each sample, such as survival, age, grade, IDH1 mutation status, etc.

[0051] Example 1: Establishing a gene set and system for predicting the immune microenvironment, prognosis and treatment sensitivity of glioblastoma

[0052] By reading the literature and combining the T cell activation gene set in the GSEA database, a T cell activation gene set related to GBM tumors was obtained, which contained 262 genes (as shown in Table 1). The R survival package, Graphpad prism7, was used to perform univariate COX regression analysis on the above 262 genes in the samples of the TCGA and CGGA325 databases, and genes with survival value were selected (P<0.05). The intersection of the two databases of TCGA and CGGA325 was taken to obtain 5 genes with stable prognostic value for glioblastoma: CD276, IL-15 (interleukin-15), SLC11A1 (solute carrier family 11 member 1), TNFSF4 (TNF superfamily member 4), and TREML2 (triggering receptor expressed on myeloid cells like 2). These 5 genes are all risk genes with HR>1 and poor prognosis.

[0053] Table 1 GBM tumor-related T cell activation gene set

[0054]

[0055] A calculation formula for a standardized risk score for glioblastoma associated with T cell activation was established based on these five genes:

[0056] RS * =(RS-min) / (max-min)(I)

[0057] Among them, RS* is the standardized risk score, RS is the T cell activation-related risk score, and RS is calculated using the following formula:

[0058] RS = (0.5258×E CD276 )+(0.1477×E IL-15 )+(0.2178×E SLC11A1 )+(0.2346×E TNFSF4 )+(0.2192×E TREML2 )(II).

[0059] Among them, E CD276 、E IL-15 、E SLC11A1 、E TNFSF4 and E TREML2 The expression levels of genes CD276, IL-15, SLC11A1, TNFSF4 and TREML2 in glioblastoma samples are respectively: transcriptome sequencing (mRNA-seq) is performed on glioblastoma samples to obtain the mRNA expression profile of about 20,000 genes in the sample, and the obtained expression levels are expressed in FPKM format. Then, the expression levels of the five genes CD276, IL-15, SLC11A1, TNFSF4 and TREML2 are found from the expression profile of 20,000 genes. The sequencing process uses existing sequencing instruments or entrusts professional companies to perform sequencing.

[0060] We performed univariate and multivariate COX regression analysis on GBM patients in the TCGA database using R survival package, Graphpad prism 7. We found that the risk score was an independent poor prognostic indicator, different from clinical characteristics such as grade, IDH1 mutation, MGMT promoter methylation, radiotherapy, and age (as shown in Figure 1, B and C).

[0061] In order to evaluate the prognostic value of T cell activation risk score, we calculated the risk score of glioblastoma samples in the TCGA database, and used the median value of the risk score of 10.7567 as the reference value. Samples with risk scores higher than 10.7567 were considered as the high-risk group, and samples with risk scores lower than 10.7567 were considered as the low-risk group. Survival curves were drawn using R survival, and the Log-rank test was used to compare the survival differences between the two groups of patients. By comparison, we found that the overall survival time of GBM patients in the high-risk score group in the TCGA database was significantly lower than that of patients in the low-risk score group (as shown in Figure 1A).

[0062] Similarly, we calculated the standardized risk score RS for GBM patients in the TCGA database. *The 0.5-2 year survival probability of GBM patients was predicted by standardized risk score, and the 0.5-2 year survival rate nomogram of glioblastoma patients was drawn using R nomogram (as shown in Figure 1D). The calibration plot test chart was drawn using R nomogram, and the test results showed that the predicted survival probability was very consistent with the actual survival rate (as shown in Figure 1E).

[0063] Furthermore, a system for predicting the immune microenvironment and prognosis of glioblastoma was established based on the standardized risk score: the system calculated the standardized risk score RS* of glioblastoma by formula I and formula II, and predicted based on the score: samples with a standardized risk score greater than the median value of 0.4506 had a shorter prognostic survival period and a complex immune microenvironment, and samples with a standardized risk score less than the median value of 0.4506 had a longer prognostic survival period and a simple immune microenvironment.

[0064] We obtained the gene set related to the first signal of T cell activation (T cell activation via T cell receptor contact with antigen bound to MHC molecule on antigen presenting cell, GO:0002291) from the GSEA database. Using GO:0002291 as the gene set, the glioblastoma sample sequencing data was calculated using the ssGSEA algorithm in R 3.6.3 to obtain the T cell activation score. We combined the risk score and the T cell activation score to evaluate whether there were differences in the sensitivity of different GBM patients to immunotherapy in the TCGA database. The Spearman correlation analysis and chi-square test in R3.6.3 were used to analyze the correlation between the activation score and the risk score and the relationship between the two and the sensitivity to immunotherapy. As shown in A and B in Figure 2, in the TCGA database, the T cell activation score and the risk score were significantly positively correlated, and the TIDE analysis results showed that the proportion of patients sensitive to immunotherapy was the highest in the group with high activation score and low risk score, while the proportion of patients sensitive to immunotherapy was the lowest in the group with low activation score and high risk score, and the difference was statistically significant. The GO:0002291 gene set includes eight genes, including LILRB1, HLA-DMB, ICAM1, ITGAL, LGALS3, LGALS9, APBB1IP and HAVCR2.

[0065] Based on the above analysis, we subtracted the standardized T cell activation score from the risk score and performed Z-score standardization to establish a calculation formula for the glioblastoma immunotherapy score IS:

[0066] IS=Z-score(x)=(x-μ) / δ(III)

[0067] Where x = PS * -RS * , μ is the mean of the sample data 0.0925, δ is the standard deviation of the sample data 0.1571

[0068] PS * The normalized activation score was calculated using Formula IV

[0069] PS * =(PS-min) / (max-min) (IV)

[0070] Among them, PS is the T cell priming activation score, which was calculated using the ssGSEA algorithm in R 3.6.3 using GO:0002291 as the gene set for the glioblastoma sample sequencing data;

[0071] RS * is the standardized risk score for glioblastoma samples, calculated using Formula I.

[0072] Taking the gene expression data of glioblastoma samples in the TCGA database as input, the immunotherapy score IS of GBM samples in the TCGA database was calculated using formula III and formula IV, and the patients were divided into high and low groups based on the median value of the immunotherapy score of 0.0303 in the TCGA database samples (patients with scores higher than the median were classified as high-groups, and patients with scores lower than the median were classified as low-groups). We found that among GBM patients in the TCGA database, our immunotherapy score can well distinguish patients who are sensitive and insensitive to immunotherapy predicted by TIDE. Patients with high immunotherapy scores are mainly patients who are sensitive to immunotherapy predicted by TIDE, while patients with low immunotherapy scores are mainly patients who are insensitive to immunotherapy predicted by TIDE. That is, the immunotherapy response rate of patients in the high immunotherapy score group is significantly higher than that of the low treatment score group, and the difference is statistically significant (as shown in C in Figure 2).

[0073] A system for predicting the treatment sensitivity of glioblastoma was established based on the immunotherapy score: the system calculates the immunotherapy score IS of glioblastoma by formula III and formula IV, and predicts based on the score: samples with an immunotherapy score greater than the median value of 0.0303 are sensitive to immunotherapy, and samples with an immunotherapy score less than the median value of 0.0303 are resistant to immunotherapy. The median value of 0.0303 is calculated based on the immunotherapy scores of samples in the database TCGA.

[0074] Example 2: Relationship between T cell activation risk score and prognosis in the validation group

[0075] We used R survival package, Graphpad prism 7 to perform univariate and multivariate COX regression analysis on GBM patients in the validation group (CGGA325, CGGA693 and GSE16011). It was found that the risk score was also different from clinical characteristics such as grade, IDH1 mutation, MGMT promoter methylation, radiotherapy, age, etc., which was an independent poor prognostic indicator (as shown in AF in Figure 3). These results show that the risk score related to T cell activation is an independent prognostic indicator. Referring to the method of Example 1, the 0.5-2 year survival curve of the validation group (CGGA325 and CGGA693) was drawn using R nomogram. We obtained the same conclusion that the standardized risk score in the CGGA325 and CGGA693 databases also has a high survival prediction accuracy (as shown in G and H in Figure 3). These results show that the standardized risk score has a high survival prediction accuracy and clinical translation value.

[0076] Example 4: Relationship between T cell activation risk score and immune microenvironment

[0077] R estimate was used to calculate the tumor purity, matrix score, and immune score of GBM samples in the TCGA database, and it was found that the T cell activation risk score was significantly positively correlated with the tumor purity, and negatively correlated with the tumor immune score and matrix score (as shown in AC in Figure 4). It can be seen that for glioblastoma samples, the higher the T cell activation risk score, the lower the purity of the tumor, and the more complex the immune microenvironment in the tumor.

[0078] In Rstudio R 3.6.3, the Metagenes method was used to perform relative quantitative calculations on various immune cells in GBM tumors in the TCGA database, and immune cells were divided into three categories: antitumor effect, tumor-promoting immunosuppression, and intermediate. The ssGSEA analysis using the R GSVA package found that with the increase of the T cell activation risk score, the immune activation and immunosuppression cells in the tumor increased simultaneously (as shown in Figure 4D). The antitumor immune activation score and tumor-promoting immunosuppression score were calculated based on the amount of each type of cell content in each GBM patient. It was found that there was a significant positive correlation between the antitumor immune promotion score and the tumor-promoting immunosuppression score in GBM, and both effects increased with the increase of the T cell activation risk score (as shown in Figure 4E). Then we used the MCP counter immune microenvironment component analysis tool to verify the above results and found that the content of immune cells in the tumor increased with the increase of the T cell activation risk score (as shown in Figure 4F). It can be seen that glioblastoma patients with high T cell activation risk scores have a more complex immune microenvironment. The same analysis was performed on samples in the validation set (CGGA325, CGGA693, and GSE16011) and the same conclusions were obtained.

[0079] Example 5: Evaluation of the sensitivity of GBM immunotherapy using the immunotherapy score IS in the validation group

[0080] The method of Example 1 was used to perform correlation analysis on the activation score and risk score of the validation group (CGGA325, CGGA693), and a chi-square test was performed on the relationship between the two and the sensitivity of immunotherapy. We obtained the same result, that is, the immunotherapy response rate of patients with high activation score and low risk score was significantly higher than that of patients with low activation score and high risk score (as shown in AD in Figure 5). The same analysis was performed on the GSE16011 database. The immunotherapy response rate of patients with high activation score and low risk score in the samples of this database was also significantly higher than that of patients with low activation score and high risk score.

[0081] Taking the gene expression data of glioblastoma samples from the two databases CGGA325 and CGGA693 as input, the immunotherapy scores IS of GBM samples in the three databases were calculated using formula III and formula IV. The median immunotherapy score of TCGA database samples was 0.0303 to divide patients into two groups (patients with scores higher than 0.0303 were classified as high-groups, and patients with scores lower than 0.0303 were classified as low-groups). Our immunotherapy score can well distinguish patients who are sensitive and insensitive to immunotherapy predicted by TIDE, that is, the immunotherapy response rate of patients in the high immunotherapy score group is significantly higher than that of the low treatment score group, and the difference is statistically significant (as shown in E and F in Figure 5). Similarly, the immunotherapy score IS of GSE16011 database samples was calculated, and we obtained the same conclusion that the immunotherapy response rate of patients in the high immunotherapy score group of GSE16011 database samples is significantly higher than that of the low treatment score group. Therefore, the immunotherapy score can well evaluate the immunotherapy sensitivity of glioblastoma, and patients with high immunotherapy scores can benefit more from immunotherapy.

[0082] For those skilled in the art, various other corresponding changes and deformations can be made according to the technical solutions and concepts described above, and all of these changes and deformations should fall within the protection scope of the claims of the present invention.

Claims

1. A gene set for predicting the immune microenvironment, prognosis and treatment sensitivity of glioblastoma, characterized in that: The gene set includes 5 genes in total, namely CD276, IL-15, SLC11A1, TNFSF4 and TREML2.

2. A system for predicting the immune microenvironment and prognosis of glioblastoma, characterized in that: The system comprises an input module, an analysis module and an output module, wherein the analysis module is connected to the input module and the output module in data communication; The input module is used to input transcriptome expression data of glioblastoma samples; The analysis module calculates the standardized risk score RS* based on the gene expression data of the glioblastoma sample using Formula I and predicts the immune microenvironment and prognosis of the glioblastoma sample according to the value of the standardized risk score: RS*=(RS-8.7401) / 4.4757 (I) Among them, RS is the risk score associated with T cell activation, which is calculated using formula II: RS= (0.5258×ECD276)+(0.1477×EIL-15)+(0.2178×ESLC11A1)+(0.2346×ETNFSF4)+ (0.2192×ETREML2) (II) Among them, ECD276, EIL-15, ESLC11A1, ETNFSF4, and ETREML2 are the expression levels of genes CD276, IL-15, SLC11A1, TNFSF4, and TREML2 in glioblastoma samples, respectively; The output module is used to output the immune microenvironment and prognosis of glioblastoma samples.

3. A system for predicting the therapeutic sensitivity of glioblastoma, characterized in that: The system comprises an input module, an analysis module and an output module, wherein the analysis module is connected to the input module and the output module in data communication; The input module is used to input transcriptome expression data of glioblastoma samples; The analysis module calculates the immunotherapy score IS based on the gene expression data of the glioblastoma sample using formula III And predict the treatment sensitivity of glioblastoma samples based on the immunotherapy score: IS=Z-score(x)=(x-μ) / δ(III) Where x = PS*-RS*, μ is 0.0925, and δ is 0.1571; PS* is the normalized activation score, calculated using Formula IV PS*=(PS-0.0502) / 0.4357 (IV) PS is the T cell activation score, GO:0002291 is used as the gene set, and the test results of glioblastoma samples are Ordinal data were calculated using the ssGSEA algorithm in R 3.6.3; RS* is a standardized risk score, calculated using Formula I and Formula II in claim 2; The output module is used to output the treatment sensitivity of the glioblastoma sample.

4. A method for predicting the immune microenvironment, prognosis and therapeutic sensitivity of glioblastoma for non-disease diagnosis and treatment purposes, characterized in that: Use the following steps: A) Obtain glioblastoma samples and perform RNA sequencing on glioblastoma samples to obtain transcriptome expression data of glioblastoma patients; B) Calculating the standardized risk score RS* of glioblastoma samples using Formula I and Formula II in claim 3: C) Calculating the immunotherapy score IS of the glioblastoma sample using formula III and formula IV in claim 3: D) The immune microenvironment, prognosis, and treatment sensitivity of glioblastoma samples were predicted based on the calculation results of the standardized risk score RS* and the immunotherapy score IS.

5. A system for predicting the immune microenvironment and prognosis of glioblastoma according to claim 2, characterized in that: The analysis module predicts the immune microenvironment and prognosis of the glioblastoma sample according to the standardized risk score. When the standardized risk score is greater than 0.4506, it is determined that the glioblastoma sample has a shorter prognosis survival period and a complex immune microenvironment. When the standardized risk score is less than 0.4506, it is determined that the glioblastoma sample has a longer prognosis survival period and a simple immune microenvironment.

6. A system for predicting glioblastoma treatment sensitivity according to claim 3, characterized in that: The analysis module predicts the treatment sensitivity of glioblastoma according to the immunotherapy score. When the immunotherapy score is greater than 0.0303, the sample is determined to be sensitive to immunotherapy, and when the immunotherapy score is less than 0.0303, the sample is determined to be resistant to immunotherapy.

7. A gene set for predicting the immune microenvironment, prognosis and treatment sensitivity of glioblastoma according to claim 1, characterized in that: The gene set is used to predict the immune microenvironment, prognosis and treatment sensitivity of glioma patients.

Images