A method for constructing a telomere-related gene expression prediction model and its products and applications

Abstract

The present invention discloses a method for constructing a telomere-related gene expression prediction model and its products and applications. The construction method comprises the following steps: 1. Based on the differential expression of gastric cancer and normal tissue genes, screening gastric cancer differentially expressed genes; 2. Based on the telomere-related gene set, screening telomere-related differentially expressed genes associated with gastric cancer prognosis; 3. Based on the prognosis of gastric cancer patients, screening telomere-related differentially expressed genes associated with gastric cancer prognosis; 4. Based on the screened telomere-related differentially expressed genes associated with gastric cancer prognosis, the telomere-related gene expression prediction model is constructed using the LASSO-COX regression algorithm. The present invention establishes a prognosis and drug sensitivity model suitable for gastric cancer patients, which indicates the prognosis of gastric cancer patients and the sensitivity to potential therapeutic drugs, can improve the efficacy of treatment, while reducing material and practical costs, and provide more reliable and convenient decision support for clinicians.

Description

Technical Field

[0001] The present invention relates to a method for constructing a telomere-related gene expression prediction model, a product and an application thereof, and belongs to the technical field of bioinformatics analysis. Background Art

[0002] Gastric cancer is a highly invasive malignant tumor and the fifth most common and third most deadly cancer in the world (Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A (2018) Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 68: 394-424. DOI 10.3322 / caac.21492). Gastric cancer usually originates from the epithelial cells of the gastric mucosa and can spread to surrounding tissues and lymph nodes, and even distant organs through different pathways during development. The early symptoms of gastric cancer are not obvious, which leads to delayed diagnosis, and most patients are discovered and diagnosed in the late stage. Therefore, patients with gastric cancer usually have a low chance of cure, poor prognosis, and limited survival prospects. The 5-year survival rate of such patients is only 10%-30%. Patients with distant metastasis have a significantly poorer prognosis, with a median survival of less than 10 months.

[0003] Traditional treatments for gastric cancer include surgical resection, radiotherapy, and chemotherapy, however, these treatments have limited efficacy. Therefore, there is an urgent need to develop novel, effective, and highly specific biomarkers to improve prognosis and recommend potential treatments.

[0004] Telomeres are repetitive nucleotide sequences present at the ends of linear chromosomes that bind to the shelter protein complex and maintain chromosome stability during DNA replication. Telomeres shorten by approximately 55 base pairs with each cell division. Therefore, telomeres determine the number of times a cell can divide before apoptosis occurs, thereby protecting chromosome ends from erosion and illegal recombination. However, under abnormal circumstances, telomeres can be activated, leading to unlimited replication of telomeric DNA, a key process associated with tumorigenesis.

[0005] Previous studies have reported the potential utility of telomeres and telomere-related genes as useful biomarkers for diagnosis, treatment selection, and prognosis prediction of various malignancies, including glioma, breast cancer, bladder cancer, and pancreatic cancer. In gastric cancer, elevated telomerase activity is associated with accelerated tumor progression and invasiveness. In addition, it has been reported that the expression level of human telomerase reverse transcriptase in gastric cancer tissues is associated with the degree of invasion, the presence of lymph node metastasis, TNM stage, and distant metastasis. This suggests that telomere-related genes may play an important role in the occurrence and development of gastric cancer.

[0006] Therefore, it is necessary to further study the expression pattern and prognostic significance of telomere-related genes in gastric cancer and establish relevant prognostic and drug sensitivity models, so as to understand the expression changes of these genes and their relationship with the prognosis of gastric cancer patients, and provide important basis for treatment selection and prognostic evaluation of gastric cancer. Summary of the invention

[0007] Purpose of the invention: The technical problem to be solved by the present invention is to provide a method for constructing a telomere-related gene expression prediction model and its products and applications, which can more accurately evaluate the prognosis and drug sensitivity of gastric cancer patients.

[0008] Technical solution: To solve the above technical problems, the present invention provides a method for constructing a telomere-related gene expression prediction model, comprising the following steps:

[0009] (1) Screening of differentially expressed genes in gastric cancer based on the differential gene expression between gastric cancer and normal tissues;

[0010] (2) Based on the telomere-related gene set, screen for telomere-related differentially expressed genes;

[0011] (3) Based on the prognosis of gastric cancer patients, screen telomere-related differentially expressed genes associated with gastric cancer prognosis;

[0012] (4) Based on the screened telomere-related differentially expressed genes associated with gastric cancer prognosis, the telomere-related gene expression prediction model was constructed using the LASSO-COX regression algorithm.

[0013] Among them, the telomere-related differentially expressed genes associated with gastric cancer prognosis in step (3) include MYOM2, RAD54L, TEAD4, BRIP1, TUBB3, CDC45, CHEK2, FEN1, GEN1, PALM, CAMKV, EZH2, FANCD2, ECT2, PNCK, PDE1B, RAB9B, EPS8L1, RAD51, LRRN1, RPS6KA5, ATP2B3, CLSPN, PDK4, TCEAL7, TFAP2B, SNCG, KLF9, AR, CDC25A, MSH2, CALD1, NFE2L3, PSAT1 or GAMT.

[0014] Among them, the model predicts the sensitivity of all anticancer drugs.

[0015] The present invention also provides a device based on the telomere-related gene expression prediction model.

[0016] Among them, the telomere-related gene expression prediction model is Score=(0.02940246)*LRRN1+(0.04641864)*SNCG+(0.06776380)*GAMT+(0.19100883)*PDE1B, wherein LRRN1, SNCG, GAMT, and PDE1B are the expression levels of the corresponding genes.

[0017] The present invention also provides the use of the device based on the telomere-related gene expression prediction model in predicting the survival status and / or drug sensitivity of gastric cancer patients.

[0018] The following steps are included: predicting the drug sensitivity of gastric cancer patients based on the model score calculated by the telomere-related gene expression prediction model; when the model score is greater than the threshold, the sample is judged to be high risk, otherwise it is low risk. The model can simultaneously output drug sensitivity IC according to the input gene expression characteristics. 50 Prediction score, based on which drug sensitivity can be assessed.

[0019] Among them, oncoPredict was used to calculate the drug sensitivity score.

[0020] The present invention also provides an electronic device, comprising: a processor and a memory; the processor and the memory are connected; the memory is used to store a computer program, and the processor is used to call the computer program to execute the described method.

[0021] The present invention also provides a computer storage medium, wherein the computer storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the method described is executed.

[0022] The present invention is the first to establish a prognosis and drug sensitivity prediction model for gastric cancer patients based on telomere-related genes using the LASSO-COX algorithm, which is used to evaluate the patient's prognosis and sensitivity to tumor drugs. According to the model score, gastric cancer patients can be divided into high-risk and low-risk groups, with patients in the high-risk group having a poor prognosis and patients in the low-risk group having a good prognosis. Furthermore, compared with the high-risk group, the low-risk group is sensitive to axitinib, cytarabine, gemcitabine, oxaliplatin, sorafenib, and ribociclib, while the high-risk group is more sensitive to niraparib.

[0023] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages:

[0024] 1) The present invention establishes a prognosis and drug sensitivity model suitable for gastric cancer patients: By establishing a prognosis and drug sensitivity model, the prognosis and drug sensitivity of gastric cancer patients can be more accurately evaluated, which helps doctors to more accurately judge patients when formulating individualized treatment strategies, thereby improving the efficacy and effectiveness of treatment;

[0025] 2) The present invention improves convenience and time saving, while reducing costs: Through the model tool of the present invention, sensitive drugs for the population can be quickly recommended for efficient and accurate treatment, saving time for diagnosis and decision-making, improving the convenience and efficiency of clinical work, while reducing treatment costs and alleviating the economic burden on patients;

[0026] In summary, the prognosis and drug sensitivity model for gastric cancer patients established by the present invention is based on telomere-related genes and is modeled using the LASSO-COX algorithm. It can indicate the prognosis of gastric cancer patients and their sensitivity to potential therapeutic drugs, improve the efficacy of treatment, reduce material and practical costs, and provide clinicians with more reliable and convenient decision support. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 The data set and implementation process used in the present invention;

[0028] Figure 2 To establish a prognostic model for gastric cancer based on telomere-related genes: (A) Differential expression analysis between gastric cancer and normal tissues in the TCGA-STAD cohort; (B) Venn diagram of differentially expressed genes (DEGs) and telomere-related genes (TRGs); (C) LASSO-Cox regression analysis; (D) 10-fold cross validation;

[0029] Figure 3Figure 2 shows the relationship between high-risk and low-risk groups and prognosis and gene expression in the training set TCGA-STAD cohort: (A) KM curve of high-risk and low-risk groups; (B) risk scores of high-risk and low-risk groups; (C) survival time of high-risk and low-risk groups; (D) expression levels of 4 prognostic genes in high-risk and low-risk groups;

[0030] Figure 4 Prognostic differences between high-risk and low-risk groups in different clinical feature subgroups in the training set TCGA-STAD cohort: (A) node-negative (N0), (B) node-positive (N+), (C) T stage T1-2, (D) T stage T3-4, (E) TNM stage I-II, (F) TNM stage III-IV, (G) differentiation level G1-2, (H) differentiation level G3;

[0031] Figure 5 Figure 3. The relationship between high-risk and low-risk groups and prognosis and gene expression in the validation set GSE62254 cohort: (A) KM curve of high-risk and low-risk groups; (B) risk scores of high-risk and low-risk groups; (C) survival time of high-risk and low-risk groups; (D) expression levels of 4 prognostic genes in high-risk and low-risk groups;

[0032] Figure 6 ROC curves for the model distinguishing patient prognosis in the TCGA-STAD cohort (A) and the GSE62254 cohort (B);

[0033] Figure 7 High- and low-risk groups and the degree of immune cell infiltration in the TCGA-STAD cohort;

[0034] Figure 8 Analysis of high- and low-risk groups and drug sensitivity in the TCGA-STAD cohort: sensitivity of high- and low-risk groups to (A) cisplatin, (B) axitinib, (C) cytarabine, (D) docetaxel, (E) gemcitabine, (F) oxaliplatin, (G) niraparib, (H) sorafenib, and (I) ribociclib. DETAILED DESCRIPTION

[0035] The technical solution of the present invention is further described below in conjunction with the accompanying drawings.

[0036] In the present invention, "telomere-related genes" refer to 2093 genes that play a role in telomere biology, including genes obtained from telomere chromatin analysis of telomerase-positive cells, genes associated with telomere shortening, genes potentially related to telomere extension alternative mechanism pathways, etc. The gene list comes from the TelNet database (https: / / malone2.bioquant.uni-heidelberg.de / fmi / webd / TelNet).

[0037] In the present invention, "tumor-infiltrating immune cells" refers to immune cells that leave the bloodstream and migrate into the tumor. These immune cells include T cells and B cells, as well as natural killer cells, macrophages, neutrophils, dendritic cells, mast cells, eosinophils, basophils, etc., which can be present in tumors in different proportions. "The level of immune cell infiltration in the tumor" refers to the relative content of tumor-infiltrating immune cells.

[0038] In the present invention, the "drug" in "drug sensitivity" includes potential drugs that have been approved for marketing, are in clinical trials and preclinical studies. These drugs include Camptothecin, Vinblastine, Cisplatin, Cytarabine, Docetaxel, Gefitinib, Navitoclax, Vorinostat, Nilotinib, Olaparib, Axitinib, AZD7762, SB216763, KU-55933, Afatinib, Staurosporine, PLX-4720, NU7441, Doramapimod, Wee1, Inhibitor, Nutlin-3a, (-), Mirin, PD1 73074, ZM447439, Alisertib, RO-3306, MK-2206, Palbociclib, Dactolisib, Pictilisib, AZD8055, PD0325901, Obatoclax, Mesylate, 5-Fluor ouracil, Dasatinib, Paclitaxel, Crizotinib, Rapamycin, Sorafenib, BI-2536, Irinotecan, Oxaliplatin, BMS-536924, GSK1904529A, Tozas ertib, PF-4708671, PRIMA-1MET, Erlotinib, Niraparib, MK-1775, Dinaciclib, Gemcitabine, Bortezomib, GSK269962A, SB505124, Tamoxife n, Fulvestrant, EPZ004777, YK-4-279, Daporinad, BMS-345541, AZ960, Talazoparib, XAV939, Trametinib, Dabrafenib, Temozolomide, AZD54 38. IAP_5620, AZD2014, AZD1208, AZD1332, Ruxolitinib, Linsitinib, Epirubicin, Cyclophosphamide, Pevonedistat, Sapitinib, Uproserti b. LCL161, Lapatinib, Luminespib, Alpelisib, Taselisib, EPZ5676, SCH772984, IWP-2, Leflunomide, Entinostat, OSI-027, LGK974, VE-822,WZ4003、CZC24832、AZD5582、GSK2606414、PFI3、PCI-34051、Wnt-C59、I-BET-762、RVX-208、OTX015、GSK343、ML323、Entospletinib、PRT062607、Ribociclib、AGI-6780、Picolinici-acid、AZD5153、CDK9_5576、CDK9_5038、Eg5_9814、ERK_2440、ERK_6604、IRAK4_4710、JAK1_8709、AZD5991、PAK_5339、TAF1_5496、ULK1_4989、VSP34_8731、Selumetinib、IGF1R_1738、JAK_1739、AZD4547、Ibrutinib、Zoledronate、Acetalax、Oxaliplatin、Carmustine、Topotecan、Teniposide、Mitoxantrone、Dactinomycin、Fludarabine、Nelarabine、Fulvestrant、Vincristine、Docetaxel、Podophyllotoxin、bromide、Dihydrorotenone、Gallibiscoquinazole、Elephantin、Sinularin、Sabutoclax、LY2109761、OF-1、MN-64、KRAS、(G12C)、Inhibitor-12、MG-132、BDP-00009066、Buparlisib、Ulixertinib、Venetoclax、ABT737、Dactinomycin、Afuresertib、AGI-5198、AZD3759、AZD5363、AZD6738、AZD8186、Osimertinib、Cediranib、Ipatasertib、GDC0810、GNE-317、GSK2578215A、I-BRD9、Telomerase、Inhibitor、IX、MIRA-1、NVP-ADW742、P22077、Savolitinib、UMI-77、WIKI4、Sepantronium、bromide、MIM1、WEHI-539、BPD-00008900、Foretinib、BIBR-1532、Pyridostatin、AMG-319、MK-8776、Ulixertinib、Vinorelbine, VX-11e, Uprosertib, LJI308, AZ6102, GSK591, VE821, AZD6482, AT13148, BMS-754807, JQ1. ,

[0039] Example 1 Model Construction

[0040] 1 Dataset

[0041] This paper uses 3 data sets ( Figure 1 ), as follows:

[0042] (1) The training cohort included 334 gastric cancer patients in the TCGA database. The gene transcriptome expression and clinical information data of GDC TCGA Stomach Cancer (STAD) were downloaded from the UCSC Xena database (https: / / xenabrowser.net / datapages / ), and gastric cancer samples with clear clinical stage and survival prognosis information were screened.

[0043] (2) Validation cohort, including 300 gastric cancer patients in the GEO database, and the data were downloaded from the GSE62254 database (https: / / www.ncbi.nlm.nih.gov / geo / query / acc.cgi?acc=GSE62254).

[0044] (3) Telomere-related gene set, including 2093 telomere-related genes, which were downloaded from the TelNet platform (https: / / malone2.bioquant.uni-heidelberg.de / fmi / webd / TelNet).

[0045] 2Technical Tools

[0046] The present invention adopts LASSO-COX regression algorithm for modeling. This is an algorithm that combines LASSO technology and COX proportional hazard model, which is mainly used for variable selection and sparse modeling in survival analysis and is more suitable for predictive model construction of high-dimensional data. The calculation formula is as follows:

[0047]

[0048] Among them, N is the total number of patients, p is the number of genes included, β0 is the constant coefficient, β is the variable coefficient matrix, x is the gene expression matrix, and t is the penalty term to control the complexity of the model and avoid the model being too complex and having poor generalization ability. This formula is published in the article (Tibshirani, R. (1996), 'Regression shrinkage and selection via the lasso', J. Royal. Statist. Soc. B. 58, 267-288.)

[0049] The present invention uses the CIBERSORT method to calculate the tumor infiltrating immune cell score. This method is commonly used to estimate the proportion of cell composition from tissue gene expression data. A total of 22 different types of tumor infiltrating immune cells can be estimated, including T cells and B cells, as well as natural killer cells, macrophages, neutrophils, dendritic cells, mast cells, eosinophils, basophils, etc.

[0050] The present invention uses the oncoPredict method to calculate drug sensitivity scores. This method can use publicly available drug sensitivity data and gene expression data to predict the effect of specific drugs for patients through machine learning methods.

[0051] 3 The specific method of constructing the prediction model is as follows:

[0052] (1) The present invention uses the training set data and DESeq2 software to analyze the differentially expressed genes between gastric cancer tumor and normal. The threshold value is set as the adjusted P value less than 0.01 and the absolute value of the difference multiple greater than 2. As a result, 9267 differentially expressed genes were found, of which 5713 genes were up-regulated and 3554 genes were down-regulated (see Figure 2 A).

[0053] (2) Screening telomere-related differentially expressed genes. The intersection of the 9267 differentially expressed genes found in the previous step and the telomere-related genes was taken, and a total of 328 telomere-related differentially expressed genes were found (see Figure 2 B).

[0054] (3) Single factor COX regression was used to screen genes related to gastric cancer prognosis. The 328 telomere-related differentially expressed genes found in the previous step were evaluated, and 35 core genes related to gastric cancer prognosis were screened.

[0055] (4) LASSO-COX regression analysis was used for feature selection to evaluate the 35 core genes in the previous step, and four important genes were screened, namely PDE1B, LRRN1, SNCG, and GAMT (see Figure 2C, 2D). A model was established based on LASSO-COX regression analysis, and the model was as follows:

[0056] Score=(0.02940246)*LRRN1+(0.04641864)*SNCG+(0.06776380)*GAMT+(0.19100883)*PDE1B.

[0057] Among them, LRRN1, SNCG, GAMT, and PDE1B are the expression levels of the corresponding genes.

[0058] Example 2 Confirmation of the effectiveness of the model in gastric cancer prognosis in the training set

[0059] To further confirm the actual effectiveness of the model in Example 1 in distinguishing the prognosis of gastric cancer, this example uses the model constructed in the previous step in the training set, and uses the median model score (Score = 0.32457) as the threshold to divide the patients into high-risk and low-risk groups. According to the Kaplan-Meier curve, the OS of low-risk patients was significantly higher (P < 0.0001, Appendix Figure 3 A). The risk scores for the two risk groups can be found in the attached Figure 3 B. In the scatter plot (attached Figure 3 C), the mortality rate of the high-risk group (49.1%) was higher than that of the low-risk group (31.7%). The expression heat map showed that the expression levels of these four genes were higher in the high-risk group (see Appendix Figure 3 D).

[0060] Example 3 Confirmation of the prognostic efficacy of the model in various clinical subgroups of gastric cancer in the training set

[0061] To further confirm the actual effectiveness of the model in Example 1 in distinguishing the prognosis of gastric cancer in each clinical subgroup, this example uses the model constructed in the first step in the training set to perform Kaplan-Meier survival analysis on each clinical subgroup. The results showed that the OS of low-risk patients in the N0 (P = 0.038), N + (P = 0.0028), T1-2 (P = 0.002), T3-4 (P = 0.0036), I-II stage (P = 0.00014), III-IV stage (P = 0.039), G1-2 (P = 0.0026) and G3 (P = 0.018) subgroups was significantly better (see Appendix). Figure 4 ).

[0062] Example 4: Validation of the model's efficacy in gastric cancer prognosis in the validation set

[0063] To further verify the effectiveness of the model in Example 1 in distinguishing the prognosis of gastric cancer, this example uses the model constructed in the first step in the validation set, and uses the median model score (Score = 0.5051) as the threshold to divide the group into high-risk and low-risk groups. According to the Kaplan-Meier curve, the OS of low-risk patients was significantly higher (P = 0.0075, Appendix Figure 5 A). The risk scores for the two risk groups can be found in the attached Figure 5 B. In the scatter plot (attached Figure 5 C), the mortality rate in the high-risk group was higher than that in the low-risk group. The expression heat map showed that the expression levels of these four genes were higher in the high-risk group (see Appendix Figure 5 D).

[0064] Example 5 Model performance evaluation in training and validation sets

[0065] To further verify the effectiveness of the model in Example 1 in distinguishing the prognosis of gastric cancer, this example uses the model constructed in Example 1 in the training set and the validation set, draws the ROC curve, and calculates the AUC value. The results show that in the training set, AUC = 0.604, and in the validation set, AUC = 0.608 (see Appendix Figure 6 ). This indicates that the model has the efficacy of distinguishing the prognosis of gastric cancer.

[0066] Example 6 Differences in immune cell infiltration between high-risk and low-risk groups in the training set model

[0067] This example analyzes the differences in immune cell infiltration between the high- and low-risk groups in the model in Example 1. This example uses the CIBERSORT method to evaluate the infiltration of 22 immune cell types such as T cells and B cells. Significant differences were observed in 14 immune cell types. The low-risk group had higher scores for CD8+T cells, resting NK cells, activated NK cells, CD4+ memory activated T cells, activated dendritic cells, activated mast cells, and resting mast cells. In contrast, the high-risk group had significantly higher scores for B cell memory, CD4+ memory resting T cells, plasma cells, follicular helper T cells, regulatory T cells, monocytes, and eosinophils ( Figure 7 ).in, Figure 7 The immune cells include initial B cells, memory B cells, plasma cells, CD8+T cells, CD4+ initial T cells, CD4+ memory resting T cells, CD4+ memory activated T cells, follicular helper T cells, regulatory T cells, γδT cells, resting NK cells, activated NK cells, monocytes, macrophages M0, macrophages M1, macrophages M2, resting dendritic cells, activated dendritic cells, resting mast cells, activated mast cells, eosinophils, and neutrophils.

[0068] Example 7 Differences in drug sensitivity between high and low risk groups in the training set model

[0069] This example analyzes the differences in drug sensitivity between the high-risk and low-risk groups in the model of Example 1. This example uses the oncoPredict method to retrieve the corresponding drug response data and gene expression profiles from the Genomics of Drug Sensitivity in Cancer (https: / / www.cancerrxgene.org / ). Using TCGA-STAD transcriptome data, the sensitivity score is calculated to predict the IC of drugs in gastric cancer patients. 50 Values ​​(Table 1).

[0070] Table 1

[0071]

[0072]

[0073]

[0074]

[0075]

[0076]

[0077]

[0078]

[0079]

[0080]

[0081]

[0082]

[0083]

[0084]

[0085]

[0086]

[0087]

[0088]

[0089]

[0090]

[0091] Compared with the high-risk group, the low-risk group had significantly higher sensitivity to axitinib, cytarabine, gemcitabine, oxaliplatin, sorafenib, and ribociclib (P<0.05). The high-risk group had a higher sensitivity to niraparib (Appendix Figure 8 ).

Claims

1. A method for constructing a prediction model for gastric cancer telomere-related gene expression, characterized in that: The steps include: (1) Screen gastric cancer samples with clear clinical stage and survival prognosis information, and use DESeq2 software to analyze the differentially expressed genes between gastric cancer tumor and normal. Use the threshold of adjusted P value less than 0.01 and the absolute value of the difference factor greater than 2 to find the differentially expressed genes; (2) Taking the intersection of the differentially expressed genes in step (1) and the telomere-related gene set to find the telomere-related differentially expressed genes; (3) Using univariate COX regression to screen genes associated with gastric cancer prognosis, the differentially expressed genes associated with telomeres described in step (2) were evaluated to screen out core genes associated with gastric cancer prognosis; the core genes associated with gastric cancer prognosis include MYOM2 , RAD54L , TEAD4 , BRIP1 , TUBB3 , CDC45 , CHEK2 , FEN1 , GEN1 , PALM , CAMKV , EZH2 , FANCD2 , ECT2 , PNCK , PDE1B , RAB9B , EPS8L1 , RAD51 , LRRN1 , RPS6KA5 , ATP2B3 , CLSPN , PDK4 , TCEAL7 , TFAP2B , SNCG , KLF9 , AR , CDC25A , MSH2 , CALD1 , NFE2L3 , PSAT1 and GAMT ; (4) Use LASSO-COX regression analysis for feature selection, evaluate the core genes described in step (3), screen out important genes, and establish a model based on LASSO-COX regression analysis. The model is as follows: Score = (0.02940246)* LRRN1 + (0.04641864)* SNCG + (0.06776380)* GAMT + (0.19100883)* PDE1B ;in, LRRN1 , SNCG , GAMT , PDE1B All are the expression levels of the corresponding genes.

2. A device for executing the method of claim 1.

3. An electronic device, characterized in that: include: Processor and memory; The processor is connected to a memory; the memory is used to store a computer program, and the processor is used to call the computer program to execute the method according to claim 1.

4. A computer storage medium, characterized in that: The computer storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the method according to claim 1 is performed.

Images