Tumor senescence gene-based renal clear cell carcinoma molecular subtype and prognosis model construction method and device

Abstract

The invention discloses a tumor senescence gene-based renal clear cell carcinoma molecular subtype and prognosis model construction method and device. The device comprises a data set; acquiring a tumor aging gene from the database; extracting an expression data box of a tumor aging gene from the data set, and performing unsupervised hierarchical clustering analysis to obtain two subtypes; comparing gene expressions of the two obtained subtypes, and carrying out pathway enrichment analysis on differentially expressed genes; verifying by using the data set; carrying out single-factor COX analysis in the data set by utilizing tumor aging genes to obtain remarkably expressed genes of each gene set, and constructing a subtype classifier; verifying in the data set based on a subtype classifier; and constructing a prognosis prediction column graph on the verified random forest model. By subdividing ccRCC subtypes according to tumor cell aging characteristics, heterogeneity within a patient population can be identified; the most effective treatment method can be selected, the curative effect can be improved to the greatest extent, and unnecessary side effects can be reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of renal cell carcinoma prognosis model construction, and in particular to a method and device for constructing a renal clear cell carcinoma molecular subtype and prognosis model based on tumor senescence genes. Background Art

[0002] Renal cell carcinoma (RCC) is one of the most lethal malignancies in the urinary system, accounting for approximately 2% to 3% of all malignant tumors worldwide. It is usually a metabolic disease that begins in the renal cortex or tubular epithelial cells. The known subtypes are clear cell RCC (ccRCC), papillary RCC, and chromaffin cell RCC, of ​​which ccRCC is the most common subtype and one of the most lethal urinary system cancers. Abnormal lipid metabolism in ccRCC cells, with free cholesterol and esterified cholesterol aggregated to form large intracellular lipid droplets, is a unique disease feature. Due to the lack of typical clinical symptoms, approximately 35% of ccRCC patients have metastasis at the time of diagnosis. Therefore, the mortality rate of patients with metastatic or locally advanced ccRCC is high. However, with the continuous improvement of treatment strategies, the 5-year survival rate of patients with this disease has increased to 50% to 69%. Despite this, the mortality rate of patients with locally advanced or metastatic ccRCC is still high, and understanding the heterogeneity of tumors is crucial to improving the prognosis prediction and precision treatment of the disease.

[0003] Cellular senescence is a stable cellular state characterized by cessation of cell division that serves as a mechanism to suppress tumorigenesis. Cellular senescence is a powerful barrier to metabolic changes that drive cancer initiation and progression as precancerous or cancerous cells become resistant to proliferative stimuli. Cellular senescence is induced by various stress signals, such as telomere shortening, oncogene activation, cell cycle inhibition, and DNA damage. Therefore, studying the role of cellular senescence in cancer can provide new insights into the underlying biological mechanisms and new strategies for treating cancer.

[0004] There is evidence that some senescence markers are associated with cancer prognosis, immunity, and therapeutic response. For example, inhibition of the glutamine transporter SLC1A5, whose expression is enhanced in multiple cancers, can induce cellular senescence in ccRCC and inhibit tumor growth. In addition, depletion of CMTM6, which is upregulated in ccRCC, can promote DNA damage-induced cellular senescence and anti-tumor immunity. However, there is still a lack of systematic studies on senescence-related genes to reveal their evidence-based associations with ccRCC prognosis and therapeutic response. In addition, there are signs that cellular senescence affects the tumor microenvironment. The senescence-associated secretory phenotype refers to the secretion of a large number of inflammatory cytokines, immunomodulatory factors, growth factors, and proteases by senescent cells, which can regulate macrophages, fibroblasts, and other cells that play a role in innate immunity, allowing tumor cells to escape immune surveillance. However, how cellular senescence affects the tumor microenvironment of ccRCC is still unclear.

[0005] This study aims to identify ccRCC subtypes based on tumor aging gene signatures and explore their associations with clinical and therapeutic responses. Through comprehensive analysis of genetic and clinical data of 977 ccRCC patients in four independent public datasets, two distinct aging subtypes were identified, and the characteristics of the tumor microenvironment of the two subtypes and their significant clinical correlation were described. In addition, a subtype classifier was developed based on the characteristics of the two tumor subtypes, and the effectiveness of these subtypes was verified in three other datasets, that is, the aging subtypes under different tumor microenvironments showed similar prognostic and biological function trends in each dataset, and there were differences in treatment response between different subtypes. Finally, a nomogram model was constructed to predict the prognosis of patients. Summary of the invention

[0006] The purpose of the present invention is to provide a method and device for constructing a molecular subtype and prognosis model of renal clear cell carcinoma based on tumor senescence genes, so as to overcome the deficiencies in the prior art.

[0007] To achieve the above object, the present invention provides the following technical solutions: The present application discloses a method for constructing a molecular subtype and prognosis model of renal clear cell carcinoma based on tumor senescence genes, which specifically comprises the following steps: S1, obtain TCGA-KIRC dataset, GSE22541 dataset, E-MTAB-1980 dataset and immunotherapy dataset; S2. Obtain several tumor aging genes from the Cell-Age database; S3, extracting the expression data frame of tumor senescence genes in step S2 from the TCGA-KIRC dataset, and performing unsupervised hierarchical clustering analysis to obtain two subtypes; S4. Compare the gene expression of the two subtypes obtained in step S3 using the "limma" software package, and perform KEGG pathway enrichment analysis on the differentially expressed genes to obtain pathway expression characteristics; validate using the E-MTAB-1980 dataset; S5. Perform univariate COX analysis in TCGA-KIRC, GSE22541, E-MTAB-1980, and immunotherapy cohort datasets using the tumor aging genes from step S2 to obtain significantly expressed genes in each gene set. After screening, the genes were included in the random forest model to construct a subtype classifier in TCGA-KIRC. S6. The subtype classifier established based on step S5 was validated in GSE22541, E-MTAB-1980, and immunotherapy cohort datasets; S7. Construct a prognosis prediction nomogram based on the random forest model verified in step S6, combining patient age, tumor type and grade information as predictors.

[0008] Preferably, step S1 specifically includes the following sub-steps: S11. Download the genetic data and clinical information of ccRCC patients in the TCGA-KIRC dataset from the Xena portal. S12. Download the gene expression profiles and clinical information of ccRCC metastases and primary tumors in the GSE22541 dataset from the GEO database; S13. Download the gene expression profiles and clinical information of 101 ccRCC patients in the E-MTAB-1980 dataset from the ArrayExpress database; S14. Download the immunotherapy cohort dataset from publicly available sources.

[0009] Preferably, in step S2, 96 tumor aging genes are obtained from the Cell-Age database; Preferably, the specific operation of verifying using the E-MTAB-1980 dataset in step S4 is as follows: repeating steps S3 and S4 in the E-MTAB-1980 dataset, if the same two subtypes as in step 3 and the same pathway expression features as in step 4 are obtained, it is proved that the classification is valid.

[0010] Preferably, step S4 specifically includes the following sub-steps: S41, using the "limma" software package to compare the gene expression of the two subtypes obtained in step S3, setting the screening threshold of significantly dysregulated genes to corrected P<0.05, differential expression fold |log2FC|>0.5, and sorting by |log2FC| to determine the differentially expressed genes to obtain a gene list for enrichment analysis; S42. Use the "ClusterProfiler" software package to perform enrichment analysis on the gene list obtained in step S41 to obtain pathway expression characteristics.

[0011] Preferably, step S5 specifically includes the following sub-steps: S51, performing univariate Cox analysis on the tumor senescence genes obtained in step S2 in the datasets of TCGA-KIRC, GSE22541, E-MTAB-1980 and immunotherapy cohort, respectively, to obtain the significantly expressed genes in each gene set; S52, remove genes that are significantly expressed only in a single dataset, and retain genes that are significantly expressed in at least two datasets; S53. Incorporate the genes retained in step S52 into the random forest model, and construct a subtype classifier in TCGA-KIRC based on the random forest model.

[0012] The present invention also discloses a device for constructing a molecular subtype and prognosis model of renal clear cell carcinoma based on tumor aging genes, comprising a memory and one or more processors, wherein the memory stores executable code, and when the one or more processors execute the executable code, they are used to implement the above-mentioned method for constructing a molecular subtype and prognosis model of renal clear cell carcinoma based on tumor aging genes.

[0013] The present invention also discloses a computer-readable storage medium on which a program is stored. When the program is executed by a processor, the method for constructing a molecular subtype and prognosis model of renal clear cell carcinoma based on tumor aging genes is implemented.

[0014] Beneficial effects of the present invention: 1. By subdividing ccRCC subtypes based on tumor cell aging characteristics, the heterogeneity within the patient population can be more accurately identified, thereby developing personalized treatment plans. This helps to select the most effective treatment method, maximize efficacy, and reduce unnecessary side effects; 2. The prognostic prediction model established by subdividing ccRCC subtypes according to tumor cell aging characteristics can help predict the prognosis of patients, provide more accurate survival estimates and recurrence risk assessments, and thus guide clinical decision-making.

[0015] 3. By subdividing ccRCC subtypes according to tumor cell aging characteristics, we can better match patients with the most appropriate therapeutic drugs and improve treatment efficacy. At the same time, it helps to identify patients who are resistant to standard therapies and explore alternative treatment options; 4. Subdividing ccRCC subtypes according to tumor cell aging characteristics will help understand the heterogeneity of tumor aging gene characteristics, promote the discovery of new drug targets and the development of new treatments, and improve the success rate and efficiency of new drug research and development; 5. The results of this subtype classification have been verified by multiple data sets, so it has high accuracy and can be used in clinical practice to risk stratify ccRCC patients and guide more personalized clinical treatment.

[0016] The features and advantages of the present invention will be described in detail through embodiments in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 and tumor aging gene mutations; Figure 2 is the distribution of copy number alterations of all identified tumor senescence genes; Figure 3 It is a volcano map of tumor aging genes; Figure 4 It is the expression correlation diagram of tumor senescence genes; Figure 5 It is the volcano plot of Cox proportional hazards regression analysis and the schematic diagram of the correlation between gene expression and ESTIMATE results; Figure 6 is a schematic diagram showing hierarchical clustering; Figure 7 is a schematic diagram showing the expression pattern of tumor senescence genes; Figure 8 is a schematic diagram showing the differences in matrix, immune, and ESTIMATE scores between C1 and C2 clusters; Fig. 9 is a schematic diagram showing the difference in microenvironment infiltrate content between C1 and C2 groups; Fig.10 is the gene set enrichment map of up-regulated and down-regulated KEGG pathways; Fig.11 represents the proportion of aging-related isoforms of hemoglobin, platelets, and serum calcium in clinical association; Fig.12 The proportions of aging-related subtypes in terms of histological grade, tumor grade, and cancer status are shown in the clinical associations; Fig.13 is the Kaplan-Meier survival curve of overall survival and disease-free survival in the TCGA-KIRC dataset; Fig.14 It is the forest plot of univariate Cox analysis and multivariate analysis; Fig.15 It is the genome enrichment analysis graph and disease-free survival of the GSE22541 dataset; Fig.16 It is the E-MTAB-1980 dataset genome enrichment analysis graph and overall survival; Fig.17is the genomic enrichment analysis plot and overall survival of the Braun et al. dataset; Fig.18 is the Kaplan-Meier survival curve of progression-free survival for the Braun et al. dataset; Fig.19 is a schematic diagram showing the differences in predicted drug sensitivity between senescence-associated subtypes of renal clear cell carcinoma in 4 independent datasets; Fig. 20 is a schematic diagram of the proportion of subtypes in different immunotherapy response groups; Fig.21 It is a nomogram for predicting survival in patients with aging-related subtype of renal clear cell carcinoma; Fig. 22 is a schematic diagram of the hierarchical clustering of genes and the similarity of tumor senescence gene expression patterns between datasets; Fig.23 is a schematic diagram showing the up-regulated and down-regulated KEGG pathways between the datasets; Fig.24 Schematic diagram of the association between the C2 subtype of clear cell renal cell carcinoma and low content of tumor-associated endothelial cells; Fig.25 is the enrichment score of the C1 / C2 upregulated gene set; Fig.26 It is a schematic structural diagram of a device for constructing a molecular subtype and prognosis model of renal clear cell carcinoma based on tumor aging genes according to the present invention. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical scheme and advantages of the present invention clearer, the present invention is further described in detail below through the accompanying drawings and embodiments. However, it should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the scope of the present invention. In addition, in the following description, the description of known structures and technologies is omitted to avoid unnecessary confusion of the concept of the present invention.

[0019] The specific steps of the method and device for constructing a renal clear cell carcinoma molecular subtype and prognosis model based on tumor senescence genes of the present invention are as follows: Data acquisition: The gene data of ccRCC patients in the Cancer Genome Atlas Renal Clear Cell Carcinoma (TCGA-KIRC) dataset were downloaded from the Xena portal (https: / / xenabrowser.net / hub / ). The gene expression profiles of ccRCC metastases and primary tumors in the GSE22541 dataset were downloaded from the GEO database (Gene Expression Omnibus, https: / / www.ncbi.nlm.nih.gov / geo / ). The gene expression profiles of 101 ccRCC patients in the E-MTAB-1980 dataset were downloaded from the ArrayExpress database (https: / / www.ebi.ac.uk / arrayexpress). The immunotherapy (PD-1 blockade) cohort dataset was downloaded from publicly available data; 96 tumor aging genes were obtained from the Cell-Age database; Establishment and validation of subtype classification model: 1. Analyze tumor-related cellular senescence genes using the TCGA-KIRC dataset, including gene mutations, copy number changes, and immune cell infiltration levels. Use these data to draw a genomic map of tumor senescence genes in renal clear cell carcinoma. In addition, compare tumor and normal tissue samples to evaluate the expression of tumor senescence genes.

[0020] 2. Unsupervised hierarchical clustering analysis was performed using the expression of tumor aging genes in the TCGA-KIRC dataset to identify two subtypes of renal clear cell carcinoma and immune analysis and enrichment analysis were performed on the subtype groups; the results were verified using an external dataset E-MTAB-1980, indicating that this classification is effective.

[0021] 3. Tumor aging genes were used to perform univariate Cox analysis in TCGA-KIRC, GSE22541, E-MTAB-1980, and immunotherapy cohort datasets. 46 genes with significant Cox analysis in at least two datasets were included in the random forest algorithm to construct a subtype classifier in TCGA-KIRC. 4. The model will be applied to external validation data (GSE22541, E-MTAB-1980, immunotherapy cohort datasets) to verify its effectiveness. The stability and independence of the classification model will be further evaluated, and its potential significance in prognostic risk assessment will be explored.

[0022] Subtype characteristic analysis: 1. Use the TCGA-KIRC dataset and expression data and clinical data to explore the relationship between C1 and C2 subtypes and clinical manifestations, including hemoglobin, platelets, serum calcium levels, histological grade, tumor grade, cancer stage, and overall survival and disease-free survival. Analyze the differences between different subtype groups and the correlation between different subtypes and clinical patients.

[0023] 2. Obtain ccRCC single-cell sequencing data (GSE178481) from the GEO database (https: / / www.ncbi.nlm.nih.gov / geo / ) for single-cell analysis (including 15 samples), and exclude low-quality cells and double cells by excluding cells with less than 500 total UMIs, more than 30,000 reads, or a mitochondrial gene ratio greater than 15%. After quality control, 98,481 cells were obtained, and Seurat (v5.0.3) was used to normalize and integrate the scRNA-seq data sets of multiple samples, and the enrichment scores of the upregulated gene sets were calculated using the AUCell analysis method to evaluate the enrichment of C1 and C2 subtypes in endothelial cells.

[0024] 3. Analysis of association between drug response and immunotherapy response based on the aging-related subtype of renal clear cell carcinoma: The "oncoPredict" software package was used to predict the sensitivity of different subtypes to more than 400 targeted drugs. The drug sensitivity data in the "oncoPredict" software package was used as a training set to build a drug sensitivity prediction model. After evaluating the performance of the model, the gene expression profile data of different subtypes in TCGA-KIRC were input to obtain the scores of different drugs in these subtypes, and then the second version of the Cancer Treatment Response Portal (known phenotype information) was used for ridge regression to calculate the effect of gene expression on the phenotype. The association between PD-1 blocking immunotherapy and aging-related subtypes of renal clear cell carcinoma in the dataset of Braun et al. was analyzed using the chi-square test. The characteristics of endothelial cells and endothelial markers were explored, and the expression of endothelial markers was analyzed in 4 datasets.

[0025] Construction of prognostic prediction nomogram: Based on the multi-validated random forest subtype classifier, combined with patient age, tumor category and grade information as predictors, the prognosis of ccRCC patients can be more confidently predicted. ROC analysis showed that the area under the curve (AUC) of 1-year, 3-year and 5-year OS was 0.82, 0.79 and 0.78, respectively, indicating that the nomogram showed good predictive performance. In addition, through calibration curve analysis, it was observed that the actual survival status was very close to the predicted survival status.

[0026] The specific experimental results are as follows: Delineating genomic profiles of tumor senescence genes in renal clear cell carcinoma First, we analyzed the tumor aging genes in the TCGA-KIRC dataset and found that the mutation frequencies of these genes were relatively low (see Figure 1 ); Among 534 ccRCC patients, only 70 patients had non-silent mutations in 30 genes. The three genes with the highest mutation rates in patients were tumor protein p53 (TP53), PDZ domain-containing 2 (PDZD2), and Notch receptor 3 (NOTCH3). In addition, most genes maintained a normal diploid state, with only a few genes undergoing low-level changes (see Figure 2 , where colors indicate different copy number states), indicating that copy number variation is rare. The exception is casein kinase 1alpha 1 (CSNK1A1), which showed relatively high copy number amplification in patients. Next, tumor and normal tissue samples were compared to evaluate the expression of tumor senescence genes. The results showed that 10 genes were significantly upregulated and 9 genes were downregulated (adjusted P < 0.05; |log2FC|>1) (see Figure 3 , where genes with an absolute fold change value greater than 1 are significant. The size of the dot represents the AUC value for distinguishing tumor samples from normal samples). These genes also show positive or negative correlations between samples (see Figure 4 , where red indicates positive correlation and blue indicates negative correlation), indicating potential co-regulation and unique gene expression patterns at the transcriptome level. Cox proportional hazards regression analysis was performed, and the results supported the bidirectional prognostic value of tumor senescence genes (see Figure 5 E, red indicates statistical significance (P < 0.01). Since the tumor microenvironment is composed of a large number of immune cells and stromal cells, the correlation between the expression of tumor senescence genes and the levels of immune cells and stromal cells was also investigated, and the infiltration level of immune cells in tumor tissues was evaluated ( Figure 5 (F, where the X-axis represents the correlation between gene expression and ESTIMATE immune score, and the Y-axis represents the correlation between gene expression and ESTIMATE stromal score). In summary, the expression of tumor senescence genes reveals multiple potential senescence phenotypes in ccRCC.

[0027] Different expression patterns of tumor senescence genes reveal distinct subtypes of ccRCC To understand the heterogeneity of tumor senescence genes, unsupervised hierarchical cluster analysis was performed based on expression data from ccRCC samples (n = 534) from the TCGA-KIRC dataset. Two major gene clusters (C1 and C2) were found (see Figure 6 ) and plotted their expression patterns as heat maps (see Figure 7). As senescent cells play an important role in the tumor microenvironment by secreting various proteins, stromal, immune, and ESTIMATE scores were calculated for ccRCC samples in different cohorts, see Figure 8 , C2 had higher scores than C1, with *p<05; **p<01; ***p<001; ****p<0001. In terms of immune infiltration, C2 had higher levels of tumor-associated fibroblasts, myeloid dendritic cells, and tumor-infiltrating lymphocytes (B, T, and CD8+T cells) than C1, but lower levels of endothelial cells and neutrophils than C1 (see Fig. 9 , where *p<05; **p<01; ***p<001; ****p<0001); pathway enrichment analysis was performed on the gene sets based on Kyoto Encyclopedia of Genes and Genomes (KEGG) to further investigate the biological functions between gene clusters. Some signaling pathways were upregulated in C2 but downregulated in C1: NF-kappaB, p53, IL-17, and cell cycle pathways (see Fig.10 In contrast, the insulin, Notch, Rap1, and renin secretion pathways were downregulated in C2 compared with C1 (see Fig.10 The above results were confirmed by repeating the analysis on samples from the E-MTAB-1980 dataset (n=101). It is worth noting that the expression patterns and biological functions of the two clusters in the TCGA-KIRC dataset and the E-MTAB-1980 dataset are consistent (see Fig. 22 and Fig.23 ). These results suggest that the expression patterns of tumor senescence genes distinguish two different subtypes of ccRCC.

[0028] Clinical association with tumor subtypes By analyzing the clinical information of ccRCC patients, the association between ccRCC subtypes and clinical characteristics can be clarified. Therefore, the chi-square test was used to study the association between the clinical characteristics of patients in the TCGA-KIRC dataset and ccRCC subtypes. Blood tests such as complete blood count and basic metabolic tests are routine tests in clinical practice. Hemoglobin and platelet counts are basic parameters of complete blood counts. Studies have found that anemia can lead to reduced overall survival in patients. It was found that the proportion of patients with the C2 subtype was low in the low hemoglobin group (P=0.0065) (see Fig.11 A), while it was higher in the group with elevated platelets and serum calcium (P=0.0072; P=0.0051) ( Fig.11 B and C, where B is platelets and C is serum calcium). The tumor grade increased from low to high (P < 0.0001) (see Fig.12D), or the tumor stage gradually increased from early to late stage (P < 0.0001) (see Fig.12 This proportion was also high in the tumor recurrence group (P < 0.0001) (see Fig.12 These results suggest that ccRCC patients with C2 subtype have a poor prognosis. Overall survival and disease-free survival analyses were performed to validate the results of the TCGA-KIRC dataset. In fact, overall survival was significantly higher in patients with C2 subtype (P < 0.0001) (see Fig.13 G) and disease-free survival (P < 0.0001) (see Fig.13 Multivariate Cox regression analysis of overall survival and disease-free survival also confirmed that the C2 subtype was a significantly associated risk factor (see Fig.14 I and J, where I is univariate Cox analysis and J is multivariate analysis).

[0029] Validation of senescence-associated subtypes Considering the impact of ccRCC aging-related subtypes on the clinical characteristics of patients, a random forest classification model was constructed based on the TCGA-KIRC dataset to evaluate the subtypes. The classification model was applied to other public datasets: GSE22541, E-MTAB-1980, and the datasets of Braun et al. For patients in these datasets, ccRCC subtypes were predicted. Subsequently, submapping analysis was performed by comparing the subtypes identified in the TCGA-KIRC dataset with the other 3 datasets. It is worth noting that the subtypes between datasets are consistent. To verify their biological functions, GSEA analysis was performed using the upregulated pathways of the C2 subtype identified in the TCGA-KIRC dataset. The upregulated pathways of the C2 subtype are consistent in the 3 datasets ( Fig.15 A. Fig.16 C and Fig.17 E), confirming the biological function of this tumor subtype. In addition, the level of infiltrating endothelial cells in the C2 subtype was relatively low in all datasets (P < 0.001, all datasets) ( Fig.24 A, Box plots show that the number of cancer-associated endothelial cells in the C2 subtype was lower than that in the C1 subtype in 4 independent datasets). Single-cell RNA-seq analysis also confirmed that endothelial cells expressed fewer C2-upregulated genes (see Fig.25 ), which is consistent with the above results. To verify the clinical association, survival analysis was performed based on the patients' clinical information. The results confirmed that the C2 subtype was associated with poor overall survival (P<0.0001) ( Fig.16 D and Fig.17 F), disease-free survival (P=0.011) ( Fig.15 B) and progression-free survival (P=0.0097) ( Fig.18 Therefore, the consistent findings across the datasets imply that a senescence-associated subtype of ccRCC exists in the patient population.

[0030] Relationship with drug response and immunotherapy As the aging-associated subtype of ccRCC may have clinical significance, it was further explored by drug response and immunotherapy association analysis. The sensitivity of each dataset to more than 400 targeted drugs was predicted using the "oncoPredict" package. Fig.19 The results of partial drug sensitivity were shown, indicating that C1 was more sensitive to drugs than C2. Since immunotherapy has become a core strategy for the treatment of ccRCC, it is crucial to design immunotherapy-based treatment methods. Therefore, the association between PD-1 blockade immunotherapy and aging-related tumor subtypes in a recent article by Braun et al. was analyzed using a chi-square test. Patients with the C1 subtype had a higher chance of benefiting from treatment, while patients with the C2 subtype had a lower chance of benefiting (P=0.0129) ( Fig. 20 Tumor-associated high endothelial veins are the main site of lymphocyte entry into the tumor during anti-PD-1 therapy (17). Their endothelial cells are characterized by upregulation of selectin P (SELP) and selectin E (SELE) and downregulation of fucosyltransferase 7 (FUT7) and glucosyl (N-acetyl) transferase 1 (GCNT1). Therefore, the expression of these endothelial markers was investigated in 4 datasets (see Fig.24 B, Expression levels of endothelial cell marker genes in 4 independent datasets (****P < 0.001, TA-HECs, tumor-associated high endothelial cells; TA-HEVs, tumor-associated high endothelial veins) were used to infer their levels in tumor subtypes. The C1 subtype showed induced levels of SELP and SELE markers, but suppressed levels of FUT7 and GCNT1. The number of infiltrating endothelial cells was low in the C2 subtype ( Fig.24 A), low expression of endothelial markers ( Fig.24 B), thus explaining why patients with C2 subtype ccRCC have a poor response to treatment.

[0031] Constructing a Nomogram See also Figure 6 , a nomogram model was established by combining the aging-related subtypes of ccRCC with patient age, tumor type, and grade information ( Figure 6 A) as a predictor, more confidently predicting the prognosis of ccRCC patients. ROC analysis showed that the area under the curve (AUC) of 1-year, 3-year and 5-year OS ( Figure 6The BD of the nomogram was 0.82, 0.79 and 0.78, respectively, indicating that the nomogram showed good predictive performance. In addition, through calibration curve analysis, it was observed that the actual survival status was very close to the predicted survival status ( Figure 6 The total score performed well in predicting prognosis.

[0032] The invention provides two subtypes of ccRCC identified based on aging gene characteristics. These subtypes are significantly related to clinical conditions, can further understand the heterogeneity of tumors, effectively improve the prognosis prediction of tumors and increase the possibility of precision treatment. The subtype classification is based on a comprehensive analysis of the gene expression and clinical data of 977 ccRCC samples, unsupervised hierarchical clustering to identify tumor subtypes, and then drug sensitivity prediction, immune cell distribution estimation, survival analysis, and model verification are performed on different subtype patient data to further describe the tumor microenvironment characteristics of different subtypes and their correlation with clinical and treatment responses.

[0033] An embodiment of a device for constructing a molecular subtype and prognosis model of renal clear cell carcinoma based on tumor aging genes according to the present invention can be applied to any device with data processing capabilities, and the device with data processing capabilities can be a device or apparatus such as a computer. The device embodiment can be implemented through software, or through hardware or a combination of software and hardware. Taking software implementation as an example, as a device in a logical sense, the processor of any device with data processing capabilities reads the corresponding computer program instructions in the non-volatile memory into the internal memory and runs them. From a hardware perspective, if Fig.26 As shown, it is a hardware structure diagram of any device with data processing capability where a device for constructing a molecular subtype and prognosis model of renal clear cell carcinoma based on tumor aging genes of the present invention is located, except Fig.26 In addition to the processor, memory, network interface, and non-volatile memory shown, any device with data processing capabilities in the embodiment of the device may also include other hardware according to the actual function of the device with data processing capabilities, which will not be described in detail. The implementation process of the functions and effects of each unit in the above device is specifically detailed in the implementation process of the corresponding steps in the above method, which will not be described in detail here.

[0034] For the device embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The device embodiment described above is only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of the present invention. Ordinary technicians in this field can understand and implement it without paying creative work.

[0035] An embodiment of the present invention also provides a computer-readable storage medium having a program stored thereon. When the program is executed by a processor, a method for constructing a molecular subtype and prognosis model of renal clear cell carcinoma based on tumor aging genes in the above embodiment is implemented.

[0036] The computer-readable storage medium may be an internal storage unit of any device with data processing capability described in any of the aforementioned embodiments, such as a hard disk or a memory. The computer-readable storage medium may also be an external storage device of any device with data processing capability, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), an SD card, a flash card (Flash Card), etc. equipped on the device. Furthermore, the computer-readable storage medium may also include both an internal storage unit and an external storage device of any device with data processing capability. The computer-readable storage medium is used to store the computer program and other programs and data required by any device with data processing capability, and may also be used to temporarily store data that has been output or is to be output.

[0037] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modification, equivalent substitution or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for constructing a molecular subtype and prognostic model of renal clear cell carcinoma based on tumor senescence genes, characterized in that: The specific steps include: S1, obtain TCGA-KIRC dataset, GSE22541 dataset, E-MTAB-1980 dataset and immunotherapy dataset; S2. Obtain several tumor aging genes from the Cell-Age database; S3, extracting the expression data frame of the tumor senescence genes in step S2 from the TCGA-KIRC dataset, and performing unsupervised hierarchical clustering analysis to obtain two subtypes; S4. Using the "limma" software package, compare the gene expressions of the two subtypes obtained in step S3, and perform KEGG pathway enrichment analysis on the differentially expressed genes to obtain pathway expression characteristics; The E-MTAB-1980 dataset was used for validation; S5. Perform univariate Cox analysis using the tumor aging genes from step S2 in the TCGA-KIRC, GSE22541, E-MTAB-1980, and immunotherapy cohort datasets to obtain significantly expressed genes in each gene set, screen them, and include them in the random forest model to construct a subtype classifier in TCGA-KIRC. S6. The subtype classifier established based on step S5 was validated in GSE22541, E-MTAB-1980, and immunotherapy cohort datasets; S7. Construct a prognosis prediction nomogram based on the random forest model verified in step S6, combining patient age, tumor type and grading information as predictors.

2. The method for constructing a molecular subtype and prognostic model of renal clear cell carcinoma based on tumor senescence genes according to claim 1, characterized in that: Step S1 specifically includes the following sub-steps: S11. Download the genetic data and clinical information of ccRCC patients in the TCGA-KIRC dataset from the Xena portal. S12. Download the gene expression profiles and clinical information of ccRCC metastases and primary tumors in the GSE22541 dataset from the GEO database; S13. Download the gene expression profiles and clinical information of 101 ccRCC patients in the E-MTAB-1980 dataset from the ArrayExpress database; S14. Download the immunotherapy cohort dataset from publicly available sources.

3. The method for constructing a molecular subtype and prognostic model of renal clear cell carcinoma based on tumor senescence genes according to claim 1, characterized in that: In step S2, 96 tumor aging genes are obtained from the Cell-Age database.

4. The method for constructing a molecular subtype and prognostic model of renal clear cell carcinoma based on tumor senescence genes according to claim 1, characterized in that: The specific operation of verification using the E-MTAB-1980 dataset in step S4 is as follows: repeat steps S3 and S4 in the E-MTAB-1980 dataset. If the same two subtypes as in step 3 and the same pathway expression characteristics as in step 4 are obtained, the classification is proven to be valid.

5. The method for constructing a molecular subtype and prognostic model of renal clear cell carcinoma based on tumor senescence genes according to claim 1, characterized in that: Step S4 specifically includes the following sub-steps: S41, using the "limma" software package to compare the gene expression of the two subtypes obtained in step S3, setting the screening threshold of significantly dysregulated genes to corrected P<0.05, differential expression fold |log2FC|>0.5, and sorting by |log2FC| to determine the differentially expressed genes to obtain a gene list for enrichment analysis; S42. Use the "ClusterProfiler" software package to perform enrichment analysis on the gene list obtained in step S41 to obtain pathway expression characteristics.

6. The method for constructing a renal clear cell carcinoma molecular subtype and prognosis model based on tumor senescence genes according to claim 1, characterized in that: Step S5 specifically includes the following sub-steps: S51, performing univariate cox analysis on the tumor aging genes obtained in step S2 in the data sets of TCGA-KIRC, GSE22541, E-MTAB-1980 and immunotherapy cohort, respectively, to obtain the significantly expressed genes in each gene set; S52, remove genes that are significantly expressed only in a single dataset, and retain genes that are significantly expressed in at least two datasets; S53. Incorporate the genes retained in step S52 into the random forest model, and construct a subtype classifier in TCGA-KIRC based on the random forest model.

7. A device for constructing a molecular subtype and prognosis model of renal clear cell carcinoma based on tumor senescence genes, characterized in that: The method comprises a memory and one or more processors, wherein the memory stores executable codes, and when the one or more processors execute the executable codes, the method is used to implement a method for constructing a molecular subtype and prognosis model of renal clear cell carcinoma based on tumor aging genes as described in any one of claims 1 to 6.

8. A computer-readable storage medium, characterized in that: A program is stored thereon, and when the program is executed by a processor, a method for constructing a molecular subtype and prognosis model of renal clear cell carcinoma based on tumor aging genes as described in any one of claims 1 to 6 is implemented.

Images