Ovarian cancer prognosis analysis method, system and device

Abstract

The invention discloses an ovarian cancer prognosis analysis method, system and device, and the method comprises the following steps: 1, obtaining case data of an ovarian cancer patient, and removing case data without patient survival information, the case data comprising patient clinical data and ovarian cancer gene expression data; 2, carrying out clustering analysis on the gene expression data and inflammation genes of the ovarian cancer patients meeting the requirements in the step 1, and screening out differentially expressed inflammation related genes; 3, performing univariate Cox regression analysis on the differential gene expression level obtained in the step 2 and the survival rate of the ovarian cancer patient, and screening inflammation-related genes with prognosis value; through multivariable Cox regression analysis, inflammation-related genes with remarkable ovarian cancer prognosis are screened; 4, establishing an ovarian cancer prognosis model based on the ovarian cancer prognosis significant inflammation related genes screened in the step 3; and 5, based on the prognosis model constructed in the step 4, performing prognosis analysis on the case data of the ovarian cancer patient. Through verification, the prognosis model constructed by the invention can be used for prognosis of ovarian cancer patients.

Description

Technical Field

[0001] This invention relates to the field of medical assessment technology, specifically to a method, system, and apparatus for prognostic analysis of ovarian cancer based on inflammation-related genes. Background Technology

[0002] Ovarian cancer is one of the three major gynecological malignancies. Due to its extremely high malignancy and difficulty in timely diagnosis, it is often referred to as the "silent killer." CA Cancer J Clin . 2022;72(1):7-33). Furthermore, ovarian cancer presents three heartbreaking "70%" characteristics in clinical practice: 70% of ovarian cancer patients are diagnosed at an advanced stage; 70% of ovarian cancer patients experience a recurrence rate within three years of treatment; and 70% of patients with advanced ovarian cancer have a five-year mortality rate ( Chin Med J (Engl) . 2022;135(5):584-590; Int J Cancer . 2017;140(11):2451-2460). Therefore, early diagnosis and treatment are extremely important for improving the survival rate of ovarian cancer patients.

[0003] CA125 is currently the most widely used diagnostic marker for ovarian cancer in clinical practice. However, CA125 levels are often negative in patients with borderline, early-stage, and non-epithelial ovarian malignancies, and benign conditions such as menstruation, pregnancy, endometriosis, and pelvic inflammatory disease can also cause elevated CA125 levels. Gynecol Oncol 2022;165(3):650-663; Cytokine (2012;58(2):133-47), therefore, CA125-based ovarian cancer screening is not recommended for the general population. Therefore, it is essential to find biomarkers that can be used for early diagnosis and prognostic evaluation of ovarian cancer.

[0004] It is well known that after a mild infection, the body's immune system can clear the foreign substance with almost no typical immune response. However, when the foreign invasion is numerous and potent, the body must recruit or activate immune cells to the infected area through chemokines or inflammatory factors to eliminate the foreign invasion, thus leading to local inflammation. Gynecol Oncol 2022;165(3):650-663; Cytokine . 2012;58(2):133-47). Studies have shown that uncontrolled inflammation may be an important factor in tumor development. This is because a persistent inflammatory microenvironment may trigger certain specific gene mutations to induce tumor formation ( ). Gynecol Oncol 2022;165(3):650-663; Cytokine. 2012;58(2):133-47). Therefore, inflammation may be one of the factors in the occurrence and progression of ovarian cancer. Many studies have reported that inflammation-related factors are important for the progression of ovarian cancer, including ovulation, endometriosis, and pelvic inflammatory disease ( ). Gynecol Oncol 2022;165(3):650-663; Cytokine . 2012;58(2):133-47). The continuous ovulation theory is considered a major hypothesis for the carcinogenesis of epithelial ovarian cancer, mainly due to the production of cytokines / chemokines and matrix remodeling enzymes, such as growth factors, TNF-α, collagenase, and interleukins, during ovulation and repair. Int J Cancer .2003;104(2):228-32). IL-6 is a pro-inflammatory cytokine that has been shown to be an important immunomodulatory cytokine involved in many stages of tumorigenesis by activating the JAK / STAT pathway and inducing the transformation of ovarian cancer epithelium into mesenchyme. Cancer Manag Res (2018;10:6685-6693). Therefore, inflammation is a major factor contributing to ovarian cancer, including tumorigenesis, promotion, progression, and metastasis. Identifying prognostic markers for ovarian cancer based on inflammatory responses is a potential strategy that could provide a useful model for predicting prognosis. To date, there have been no reports on the use of inflammation-related genes as biomarkers for the diagnosis and prognostic evaluation of ovarian cancer.

[0005] To better serve the auxiliary diagnosis and prognosis of ovarian cancer patients, it is essential to develop a method, system, and device for more accurate prognostic estimation of ovarian cancer based on inflammation-related genes, which can be used as an auxiliary reference for the selection of treatment options and prognosis of ovarian cancer patients. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method, system, and apparatus for prognostic analysis of ovarian cancer based on inflammation-related genes. It screens out inflammation-related genes that are significantly associated with the prognosis of ovarian cancer and constructs a prognostic model to provide effective and accurate prognostic assessment.

[0007] To overcome the aforementioned technical deficiencies, this invention discloses a method for prognostic analysis of ovarian cancer based on inflammation-related genes. The method includes the following steps: First, acquiring case data of ovarian cancer patients, excluding case data lacking patient survival information; the case data includes clinical data and gene expression data of ovarian cancer patients. Second, performing consistency cluster analysis on the gene expression data and inflammatory genes of ovarian cancer patients meeting the criteria in the first step to identify differentially expressed genes. Third, performing univariate Cox regression analysis on the differential gene expression levels obtained in the second step and the survival rate of ovarian cancer patients to screen for genes with prognostic value; and using multivariate Cox regression analysis to screen for inflammation-related genes with significant prognostic value for ovarian cancer. Fourth, establishing an ovarian cancer prognostic model based on the inflammation-related genes with significant prognostic value screened in the third step. Fifth, performing prognostic analysis on the case data of ovarian cancer patients based on the prognostic model constructed in the fourth step.

[0008] Preferably, the number of repeated samplings in the cluster analysis is 1000.

[0009] Preferably, the proportion of samples extracted in each iteration of the cluster analysis is 80%.

[0010] Preferably, the correlation and false discovery rate thresholds between the expression of each prognosis-related gene in ovarian cancer patients and the overall survival rate of ovarian cancer patients are both set to be less than 0.05.

[0011] Preferably, the correlation and false discovery rate thresholds for the clustering correlation of prognosis-related genes and inflammation-related genes in ovarian cancer patients are both set to be less than 0.05.

[0012] Preferably, the risk score calculation formula in the ovarian cancer prognostic model based on inflammatory-related genes with significant ovarian cancer prognosis is as follows: .

[0013] Preferably, ovarian cancer patients are divided into high-risk and low-risk groups based on the threshold of the average risk score, and the critical value of the correlation P-value is set to be less than 0.05.

[0014] Preferably, the differentially expressed inflammation-related genes in ovarian cancer include any one or a combination of the following genes: TNFRSF13B, CD1B, AMPD1, FCRL1, FDCSP, CPN2, ALLC, FGF10, SPIC, FCRL4, IL21, GPR33, TMEM244, IL26, GPR21, AC034228.3, ABCC12, OR1N1, OTX2, and ASCL4. OTOR, SPATS1, AFM, KCNK16, PABPC1L2A, CFC1, TNP1, MAGEA9B, TCP11X2, KRT33B OR4C6, C9orf57, MBL2, CLEC2A, NXF2B, GNAT3, SFTA3, OR5BS1P, SPO11, GC, OR51M1, RHOXF2B, PSG7, MIA2, OR1B1, OR10AG1, NXF2, SYCN, MAS1L, BSPH1, OR2M2, CRYGA, GSX1, CTXN3, OR2J2, OR4E2, CLCA1, RESP18, NEUROG1, BHLHE23, GUCA1C, OR51C1P, TRIM64B, BIVM-ERCC5.

[0015] Preferably, inflammation-related genes with prognostic value for ovarian cancer include any one or a combination of the following genes: IL25, TNFR8F13B, CLEC2A, OTR, PSG7, OR1B1, and OR5BS1P.

[0016] Preferably, the inflammation-related genes that are significantly associated with the prognosis of ovarian cancer include TNFRSF13B, CLEC2A, and OTR.

[0017] Preferably, the prognostic analysis includes 1-10 year survival analysis.

[0018] Preferably, the prognostic model includes signaling pathway and immune infiltration analysis.

[0019] The present invention also provides a system for implementing the above method, including a data acquisition and preprocessing module, a cluster analysis module, a Cox regression analysis module, a prognostic model construction module, and an analysis module.

[0020] The data acquisition and preprocessing module is used to acquire case data of ovarian cancer patients and remove samples that do not meet the requirements.

[0021] The clustering analysis module is used to perform differential analysis on gene expression data from eligible ovarian cancer patients to screen for differentially expressed inflammation-related genes in ovarian cancer.

[0022] The Cox regression analysis module is used to perform univariate Cox regression analysis on the expression levels of differentially expressed inflammation-related genes in ovarian cancer and the survival rate of ovarian cancer patients to screen for ovarian cancer inflammation-related genes with prognostic value. Then, multivariate Cox regression analysis is performed on the ovarian cancer inflammation-related genes with prognostic value to screen for ovarian cancer inflammation-related genes with significant prognostic value.

[0023] The prognostic model construction module is used to establish a prognostic model based on the gene expression level of the significant gene and clinical information.

[0024] The analysis module is used to perform prognostic analysis on the gene expression levels and clinical information of ovarian cancer patients based on the prognostic model.

[0025] Preferably, the analysis module includes signaling pathways and immune analysis of ovarian cancer inflammation-related genes with significant prognostic effects in the prognostic model.

[0026] The present invention also provides an ovarian cancer prognostic analysis device, including a processor and a memory, the memory being used to store a program and the above-mentioned prognostic model; the program includes instructions for performing prognostic analysis on case data of ovarian cancer patients based on the prognostic model.

[0027] The present invention also provides an ovarian cancer prognostic analysis device, including a processor and a memory, wherein the memory is used to store a program and the above-mentioned prognostic model; the program includes instructions for performing prognostic analysis on ovarian cancer case data according to the prognostic model.

[0028] Furthermore, under the lowest configuration operating environment, the prognostic model of the ovarian cancer prognostic analysis device requires that the account management login / logout time for ovarian cancer patients be less than 3 seconds.

[0029] Furthermore, under the lowest configuration operating environment, the ovarian cancer prognostic analysis device has a running time of less than 120 minutes when the prognostic analysis of the ovarian cancer prognostic model is run at full load; less than 50 minutes when the prognostic analysis of the ovarian cancer prognostic model is run at half load; and less than 8 minutes when the prognostic analysis of the colorectal cancer prognostic model is run with a single sample.

[0030] Furthermore, under the lowest configuration operating environment, the ovarian cancer prognostic analysis device can retrieve and run the ovarian cancer prognostic model's prognostic analysis results report in less than 3 seconds.

[0031] Furthermore, the hardware and operating system configuration of the ovarian cancer prognostic analysis device is divided into a prognostic model front-end configuration and a back-end configuration.

[0032] Furthermore, the front-end configuration of the device described in this embodiment includes 4 CPUs, 16 cores / core; a main frequency of 2.5GHz or higher; 1TB or higher memory; 1TB or higher hard disk; a display resolution of 1366*768; a VGA Asus Turbo GTX 1080TI graphics card; a Windows Server 2012 R2 operating system; Office 2010; Firefox 65; Chrome 73 and above; a B / S architecture (browser / server); a local area network; and a bandwidth of gigabit network or higher.

[0033] Furthermore, the backend configuration of the device described in this embodiment includes 250GB of RAM and a 64-core CPU; the main frequency is 2.5GHz or higher, the hard drive is 10TB or higher, and it supports / dev / shm as shared memory in the form of a RAM disk. Pre-allocation of memory prevents swapping to the hard drive, and the memory read / write speed is 10-100GB per second. Simultaneously, through memory address mapping, inter-process file communication (IPC) is supported using / dev / shm, with a parallel process count of 10. The display resolution is 1366*768, the graphics card is a VGA Asus Turbo GTX 1080TI, the network architecture is a B / S architecture (browser / server), the network type is a local area network, and the bandwidth is gigabit or higher.

[0034] Compared with the prior art, the beneficial effects of the present invention are as follows: the selected inflammation-related genes are significant genes for prognostic analysis of ovarian cancer patients, also known as independent risk factors. After verification and detection analysis, the prognostic model constructed by the present invention can be used for prognostic assessment of ovarian cancer patients. Attached Figure Description

[0035] Figure 1 This is a flowchart of the ovarian cancer prognostic analysis method based on inflammation-related genes of the present invention.

[0036] Figure 2 This is a Kaplan-Meier curve of the survival probability of 378 ovarian cancer patients from the TCGA database, based on the prognostic model.

[0037] Figure 3 This is a Kaplan-Meier curve of the survival probability of 194 ovarian cancer patients from the GEO database / GSE49997, based on the prognostic model.

[0038] Figure 4 It is the KEGG enrichment curve of the prognostic model.

[0039] Figure 5 This is a prognostic model of immune cell infiltration curves from 378 ovarian cancer patients from the TCGA database.

[0040] Figure 6 This is a prognostic model of immune cell infiltration curves from 194 ovarian cancer patients from the GEO database / GSE49997.

[0041] Figure 7 This is the system logic block diagram of the present invention. Implementation

[0042] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0043] The present invention will now be described in further detail with reference to the accompanying drawings.

[0044] A method for prognostic analysis of ovarian cancer, such as Figure 1 As shown, the method includes: Step 101: Obtain case data of ovarian cancer patients. This case data includes gene expression data and clinical information, which can be obtained from the databases of hospitals or third-party clinical medical testing laboratories, or downloaded from public platforms (such as the UCSCXena browser, https: / / xenabrowser.net / ). Preprocessing of the gene expression data and clinical information of ovarian cancer patients is existing technology and will not be elaborated upon in this application, such as deduplication and deletion of lost data.

[0045] Step 102: Perform a consistency cluster analysis on the gene expression data and inflammatory genes of the eligible ovarian cancer patients selected in Step 101 to screen out the differentially expressed inflammatory-related genes in ovarian cancer.

[0046] Step 103: Univariate Cox regression analysis was performed on the differentially expressed levels of inflammation-related genes obtained in Step 102 and the survival rate of ovarian cancer patients to screen for inflammation-related genes with prognostic value; multivariate Cox regression analysis was then used to screen for inflammation-related genes with significant prognostic value in ovarian cancer.

[0047] Step 104: Establish an ovarian cancer prognostic model based on inflammation-related genes with significant prognostic significance identified in the third step of screening; the prognostic model includes signaling pathway and immune infiltration analysis.

[0048] Step 105: Based on the prognostic model constructed in step 4, perform prognostic analysis on the case data of ovarian cancer patients.

[0049] Statistical analysis revealed that inflammation-related genes TNFRSF13B, CLEC2A, and OTR are significant genes, i.e., independent risk factors, in the prognostic analysis of ovarian cancer. After verification and detection analysis, the prognostic model constructed in this invention can be used for prognostic assessment of ovarian cancer patients.

[0050] Example 1 Step S1: Data from 379 ovarian cancer patients was downloaded from the most recently updated UCSC Xena browser (https: / / xenabrowser.net / ), including important clinicopathological data and gene expression data. After excluding patients without survival information, the data was standardized and preprocessed. Ensembl IDs in the expression profiles were converted to gene symbols, encoding gene expression profiles were extracted, and the expression values ​​of the same genes were averaged to extract expression data for 200 inflammatory genes. Simultaneously, survival status was binary classified: 0 for survival and 1 for death. Normal tissue samples were removed, while gene expression levels from tumor tissue samples, such as RNA expression matrices, were retained. However, data sources are not limited to these. In this embodiment, all data is publicly available. In other words, this is ethically sound. Furthermore, this embodiment complies with the UCSC and TCGA access and publication policies for RNA-seq data and clinical follow-up data (https: / / xenabrowser.net / datapages / ).

[0051] Step S2: For the case data of eligible ovarian cancer patients, clustering and typing were performed using the R package ConsensusClusterPlus. The similarity distance between samples was calculated using Euclidean distance, and K-means clustering was used. The number of repeated samplings was 1000, and the sample proportion in each iteration was 80%. The optimal number of clusters was determined to be 2 using the cumulative distribution function (CDF). Subsequently, 378 tumor samples were assigned to the two categories. The results showed that 64 genes exhibited significant differential expression in the two transcriptional profiles, with 41 genes upregulated and 23 genes downregulated.

[0052] Sixty-four differentially expressed inflammation-related genes associated with ovarian cancer include: TNFRSF13B, CD1B, AMPD1, FCRL1, FDCSP, CPN2, ALLC, FGF10, SPIC, FCRL4, IL21, GPR33, TMEM244, IL26, GPR21, AC034228.3, ABCC12, OR1N1, OTX2, ASCL4, OTOR, SPATS1, AFM, KCNK16, PABPC1L2A, CFC1, TNP1, MAGEA9B, TCP11X2, KRT33B, and OR4C. 6, C9orf57, MBL2, CLEC2A, NXF2B, GNAT3, SFTA3, OR5BS1P, SPO11, GC, OR51M1, RHOXF2B, PSG7, MIA2, OR1B1, OR10AG1, NXF2, SYCN, MAS1L, BSPH1, OR2M2, CRYGA, GSX1, CTXN3, OR2J2, OR4E2, CLCA1, RESP18, NEUROG1, BHLHE23, GUCA1C, OR51C1P, TRIM64B, BIVM-ERCC5. But not limited to these.

[0053] Using the Coxph function in the R package `survival` to analyze 64 differentially expressed inflammation-related genes and survival data of ovarian cancer patients, univariate Cox proportional hazards regression analysis was performed with p < 0.05 as the threshold. This identified seven ovarian cancer inflammation-related genes with prognostic value: IL25, TNFR8F13B, CLEC2A, OTR, PSG7, OR1B1, and OR5BS1P. Multivariate Cox proportional hazards regression analysis was then performed using the Coxph function in `survival`, again with p < 0.05 as the threshold, and the seven identified prognostic ovarian cancer inflammation-related genes and survival data. This identified three ovarian cancer inflammation-related genes with significant prognostic value: TNFRSF13B, CLEC2A, and OTR. TNFRSF13B is a member of the TNF receptor superfamily and is highly expressed on activated B cells. It participates in the activation of B cell and T cell functions by binding to TNF ligands. Front Oncol 2021;11:642170). Furthermore, CLEC2A is a member of the C-type lectin domain family and is expressed on various immune cells, including T cells and NK cells, and can promote T cell proliferation. Immunogenetics 2007;59(12):903-12; Cancer Immunol Res2020;8(3):334-344). Meanwhile, OTR, as a cochlear gene of Otoraplin, is an effective prognostic predictor of breast cancer ( Biochem Cell Biol . 2019;97(6):750-757.). Due to limited research, how these three genes affect the prognosis and progression of ovarian cancer remains unclear.

[0054] Step S3: Establish a prognostic model.

[0055] Inflammation-related genes with significant prognostic effects in ovarian cancer were identified as independent risk factors. A risk scoring model was established using the β coefficient from multivariate Cox analysis as weights and their expression levels. The formula for calculating the risk score for ovarian cancer patients is as follows: .

[0056] Based on the median risk score, the model of this invention classifies patients into high-risk and low-risk individuals according to the input case data of eligible ovarian cancer patients.

[0057] Step S4: Application validation of the prognostic model.

[0058] To verify the reliability of the ovarian cancer prognostic model constructed in this invention based on inflammatory-related genes with significant prognostic effects, case data from 378 eligible ovarian cancer patients from the TCGA database were input into the model. The model divided the 378 patients into two groups based on the median risk score: a high-risk group (189 patients) and a low-risk group (189 patients). Figure 2 As shown. Figure 2 Kaplan-Meier curves showed that patients in the high-risk group had a worse prognosis compared to the low-risk group. The areas under the ROC curves at 10, 5, and 3 years were 0.66, 0.644, and 0.572, respectively, indicating that the prognostic model for ovarian cancer based on inflammation-related genes constructed in this invention is reliable.

[0059] To verify the model's universality in predicting prognoses for patients from different backgrounds, the reliability of the ovarian cancer prognostic model constructed in this invention was further validated using gene expression levels and clinical information from 194 eligible ovarian cancer patients within the GEO database gene chip GSE49997. During the validation process, the model divided the 194 patients into a high-risk group (n=69) and a low-risk group (n=125) based on median risk scores. The validation results showed that the survival time of patients in the high-risk group was shorter than that in the low-risk group (…). Figure 3(P=0.028). These results further demonstrate that the ovarian cancer prognostic model based on inflammation-related genes constructed in this invention is reliable and has patient-derived universality.

[0060] To further determine the differences in cancer progression between the high-risk and low-risk groups, GSVA enrichment analysis was performed using the "GSVA" R package. GSVA is commonly used in nonparametric and unsupervised methods to estimate changes in pathways and biological processes in expression datasets. The gene set "c2.cp.kegg.v7.4.symbols" was downloaded from the MSigDB database for GSVA analysis. A p-value less than 0.05 was considered statistically significant. The results showed that many signaling pathways based on TNFRSF13B, CLEC2A, and OTR genes differed significantly between the high-risk and low-risk groups. Figure 4 Typical pathways include: Wnt signaling pathway, TGF-β signaling pathway, T cell receptor signaling pathway, JAK-STAT signaling pathway, cytokine-cytokine receptor interaction, and chemokine signaling pathway. The Wnt signaling pathway is very important for the development of ovarian cancer, and inhibiting the Wnt / β-catenin signaling pathway is considered a new therapeutic strategy for ovarian cancer (Front Oncol. 2022;12:852260; Int J Biochem Cell Biol. 2022;145:106191). The Wnt / β-catenin signaling pathway can also regulate T cell infiltration in the tumor microenvironment (TME) (Front Immunol. 2019;10:2293). Abnormal activation of the JAK / STAT3 signaling pathway is closely related to the progression of various cancers, including ovarian cancer, and has an important impact on immune responses (Signal Transduct Target Ther. 2021;6(1):402). T cell receptors are involved in the recognition of antigens presented by the MHC complex, and T cell activation is mainly mediated through the T cell receptor signaling pathway (Exp Mol Med. 2020;52(5):750-761). Various cytokines and chemokines, such as IL-10, TGF-β, CCL2, and CXCL12, contribute to immunosuppressive TME and tumor progression (Front Immunol. 2020;11:594609; Clin Exp Metastasis. 2021;38(2):139-161).

[0061] Furthermore, the infiltration of immune cells in the tumor microenvironment is closely related to cancer progression. This is confirmed by the significant differences in the infiltration of immune cells such as T cells, B cells, NK cells, dendritic cells, macrophages, monocytes, and neutrophils between high-risk and low-risk groups. Figure 5 Compared to the low-risk group, the high-risk group showed less T cell infiltration, including CD4+ T cells and CD8+ T cells. Figure 6 It is well known that T cell infiltration is the most important factor in the response to cancer immunotherapy (Nat Rev Immunol. 2020;20(11):651-668). CD4+ T cells are mainly responsible for helper functions, among which Th1 and Th2 are typical types (Methods Mol Biol. 2018;1803:335-351). Figure 6 As shown, Th1 and Th2 infiltration were significantly higher in the low-risk group than in the high-risk group. Th1 cells can secrete pro-inflammatory cytokines, including IFN-γ and IL-12, into active macrophages, DCs, and NK cells (Nat Rev Immunol. 2018;18(10):635-647). The antitumor effect of Th2 cells is partly mediated by IL-4 production and eosinophil recruitment. Both Th1 and Th2 cells can differentiate into cytotoxic CD4+ T cells with antitumor cytotoxic effector functions (Immunity. 2021;54(12):2701-2711). CD8+ T cells are cytotoxic T cells that can directly kill tumor cells by producing cytokines, releasing granzymes, and inducing apoptosis through Fas-FasL interactions (Immunity. 2011;35(2):161-168). Furthermore, the expression level of the T-cell marker CD3E was negatively correlated with the risk score, further demonstrating the importance of inflammation-related genes in the construction of prognostic models for ovarian cancer. In other words, inflammation may dominate the progression of ovarian cancer, and alterations in the expression of inflammation-related genes are an important potential strategy for early diagnosis and prognosis prediction.

[0062] The present invention also provides a system for implementing the above method, such as... Figure 7 As shown, it includes a data acquisition and preprocessing module 101, a cluster analysis module 102, a Cox regression analysis module 103, a prognostic model construction module 104, and an analysis module 105.

[0063] The data acquisition and preprocessing module 101 is used to acquire case data of ovarian cancer patients and remove samples that do not meet the requirements.

[0064] The clustering analysis module 102 is used to perform clustering analysis on gene expression data and inflammatory genes from eligible ovarian cancer patients to screen for differentially expressed inflammatory-related genes in ovarian cancer.

[0065] The Cox regression analysis module 103 is used to perform univariate Cox regression analysis on the expression levels of differentially expressed inflammation-related genes in screened ovarian cancer and the survival rate of ovarian cancer patients to screen for ovarian cancer inflammation-related genes with prognostic value. Then, multivariate Cox regression analysis is performed on the ovarian cancer inflammation-related genes with prognostic value to screen for ovarian cancer inflammation-related genes with significant prognostic value.

[0066] The prognostic model construction module 104 is used to establish a prognostic model based on the gene expression levels of the ovarian cancer inflammation-related genes with significant prognosis and clinical information.

[0067] The analysis module 105 is used to perform prognostic analysis on the gene expression level and clinical information of ovarian cancer patients based on the prognostic model.

[0068] Preferably, the analysis module includes signaling pathways and immune infiltration analysis of inflammation-related genes in the prognostic model.

[0069] The present invention also provides an ovarian cancer prognostic analysis device, including a processor and a memory, the memory being used to store a program and the above-mentioned prognostic model; the program includes instructions for performing prognostic analysis on case data of ovarian cancer patients based on the prognostic model.

[0070] Furthermore, under the lowest configuration operating environment, the prognostic model of the ovarian cancer prognostic analysis device requires that the account management login / logout time for ovarian cancer patients be less than 3 seconds.

[0071] Furthermore, under the lowest configuration operating environment, the ovarian cancer prognostic analysis device has a running time of less than 120 minutes when the prognostic analysis of the ovarian cancer prognostic model is run at full load; less than 50 minutes when the prognostic analysis of the ovarian cancer prognostic model is run at half load; and less than 8 minutes when the prognostic analysis of the colorectal cancer prognostic model is run with a single sample.

[0072] Furthermore, under the lowest configuration operating environment, the ovarian cancer prognostic analysis device can retrieve and run the ovarian cancer prognostic model's prognostic analysis results report in less than 3 seconds.

[0073] Furthermore, the hardware and operating system configuration of the ovarian cancer prognostic analysis device is divided into a prognostic model front-end configuration and a back-end configuration.

[0074] Furthermore, the front-end configuration of the device described in this embodiment includes 4 CPUs, 16 cores / core; a main frequency of 2.5GHz or higher; 1TB or higher memory; 1TB or higher hard disk; a display resolution of 1366*768; a VGA Asus Turbo GTX 1080TI graphics card; a Windows Server 2012 R2 operating system; Office 2010; Firefox 65; Chrome 73 and above; a B / S architecture (browser / server); a local area network; and a bandwidth of gigabit network or higher.

[0075] Furthermore, the backend configuration of the device described in this embodiment includes 250GB of RAM and a 64-core CPU; the main frequency is 2.5GHz or higher, the hard drive is 10TB or higher, and it supports / dev / shm as shared memory in the form of a RAM disk. Pre-allocation of memory prevents swapping to the hard drive, and the memory read / write speed is 10-100GB per second. Simultaneously, through memory address mapping, inter-process file communication (IPC) is supported using / dev / shm, with a parallel process count of 10. The display resolution is 1366*768, the graphics card is a VGA Asus Turbo GTX 1080TI, the network architecture is a B / S architecture (browser / server), the network type is a local area network, and the bandwidth is gigabit or higher.

[0076] Application of the ovarian cancer prognostic analysis device described in this invention in actual clinical practice The clinical data of 200 ovarian cancer patients were entered into the ovarian cancer prognostic analysis device described in this invention for clinical auxiliary application. Through patient follow-up and treatment tracking, the results showed that the ovarian cancer prognostic analysis device can effectively group ovarian cancer patients by clinical risk, providing doctors with adjuvant therapy suggestions, thereby significantly improving the prognosis of ovarian cancer patients. Specifically, the clinical auxiliary application results showed that the prognosis of the control group (ovarian cancer patients who did not receive adjuvant therapy according to the prognostic suggestions of the ovarian cancer prognostic analysis device described in this invention) was nearly 100% consistent with the prognostic analysis of the ovarian cancer prognostic analysis device described in this invention. Meanwhile, the overall survival rate of ovarian cancer patients in the group receiving adjuvant therapy according to the prognostic suggestions given by the ovarian cancer prognostic analysis device described in this invention was significantly improved; the clinical auxiliary application experimental data showed that the average overall survival rate of ovarian cancer patients increased by 75%. Therefore, the ovarian cancer prognostic analysis device described in this invention not only provides effective suggestions for auxiliary diagnosis and treatment for hospital doctors but also brings benefits to ovarian cancer patients using the device, demonstrating good market prospects and application potential.

[0077] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for prognosis of ovarian cancer, characterized by, The method includes the following steps: First, acquiring case data of ovarian cancer patients, excluding case data without patient survival information; the case data includes patient clinical data and ovarian cancer gene expression data; Second, performing cluster analysis on the gene expression data and inflammatory genes of ovarian cancer patients meeting the requirements in the first step, and screening out differentially expressed inflammation-related genes; Third, performing univariate Cox regression analysis on the differential gene expression levels obtained in the second step and the survival rate of ovarian cancer patients, and screening for inflammation-related genes with prognostic value; and using multivariate Cox regression analysis to screen for inflammation-related genes with significant prognostic value for ovarian cancer; Fourth, establishing an ovarian cancer prognostic model based on the inflammation-related genes with significant prognostic value for ovarian cancer screened in the third step; Fifth, performing prognostic analysis on the case data of ovarian cancer patients based on the prognostic model constructed in the fourth step.

2. The method for prognostic analysis of ovarian cancer according to claim 1, characterized in that, Also included is a method for risk stratification of ovarian cancer patients: normalizing the gene expression data to obtain normalized expression values; obtaining the coefficients of the genes based on a regression analysis method; and calculating the risk score of the ovarian cancer patient according to the normalized expression values and the β coefficients in the multivariate Cox analysis: .

3. The ovarian cancer prognosis analysis method according to claim 1, characterized by, Differentially expressed inflammation-related genes in ovarian cancer include any one or a combination of the following genes: TNFRSF13B, CD1B, AMPD1, FCRL1, FDCSP, CPN2, ALLC, FGF10, SPIC, FCRL4, IL21, GPR33, TMEM244, IL26, GPR21, AC034228.3, ABCC12, OR1N1, OTX2, ASCL4, OTOR, SPATS1, AFM, KCNK16, PABPC1L2A, CFC1, TNP1, MAGEA9B, TCP11X2, KRT33B. The following genes are considered prognostic inflammation-related genes: OR4C6, C9orf57, MBL2, CLEC2A, NXF2B, GNAT3, SFTA3, OR5BS1P, SpO11, GC, OR51M1, RHOXF2B, PSG7, MIA2, OR1B1, OR10AG1, NXF2, SYCN, MAS1L, BSPH1, OR2M2, CRYGA, GSX1, CTXN3, OR2J2, OR4E2, CLCA1, RESP18, NEUROG1, BHLHE23, GUCA1C, OR51C1P, TRIM64B, and BIVM-ERCC5. Prognostic inflammation-related genes include any one or a combination of the following genes: IL25, TNFR8F13B, CLEC2A, OTR, PSG7, OR1B1, and OR5BS1P.

4. The ovarian cancer prognosis analysis method according to claim 1, characterized by, The ovarian cancer prognostic significantly inflammatory expression-related genes include TNFRSF13B, CLEC2A, and OTR; a prognostic model is established based on the expression levels of the aforementioned prognostic significantly inflammatory expression-related genes; the prognostic analysis includes 1-10 year survival analysis.

5. The ovarian cancer prognosis analysis method according to claim 1, characterized by, The prognostic model includes signaling pathway and immune infiltration analysis.

6. A system for performing the ovarian cancer prognosis analysis method according to any one of claims 1 to 5, characterized in that, The system includes a data acquisition and preprocessing module, a cluster analysis module, a Cox regression analysis module, a prognostic model construction module, and an analysis module. The data acquisition and preprocessing module acquires case data from ovarian cancer patients and removes unsuitable samples. The cluster analysis module performs cluster analysis on gene expression data and inflammation-related genes from eligible ovarian cancer patients to screen for differentially expressed inflammation-related genes in ovarian cancer. The Cox regression analysis module performs univariate Cox regression analysis on the expression levels of the screened differentially expressed inflammation-related genes in ovarian cancer and the survival rate of ovarian cancer patients to screen for ovarian cancer inflammation-related genes with prognostic value. Then, multivariate Cox regression analysis is further performed on the ovarian cancer inflammation-related genes with prognostic value to screen for inflammation-related genes with significant prognostic value in ovarian cancer. The prognostic model construction module is used to establish a prognostic model based on the gene expression levels of inflammation-related genes with significant prognostic effects in ovarian cancer and clinical information; the analysis module is used to perform prognostic analysis on the case data of ovarian cancer patients based on the prognostic model.

7. The system of claim 7, wherein, The significant genes include TNFRSF13B, CLEC2A, and OTR; the prognostic model includes signaling pathway and immune infiltration analysis.

8. A colorectal cancer prognosis analysis device characterized by comprising: The device includes a processor and a memory, the memory being used to store a program and a prognostic model established according to any one of claims 1-5; the program includes instructions for performing prognostic analysis on case data of ovarian cancer patients based on the prognostic model.

Images