Construction method of breast cancer lung metastasis prognosis prediction model

Abstract

The invention relates to the technical field of medical informatics, and particularly discloses a construction method of a breast cancer lung metastasis prognosis prediction model. Comprising the following steps: (1) acquiring breast cancer in-situ and lung metastasis single cell RNA sequencing data by a GEO database, and preliminarily screening differential genes; (2) acquiring clinical information and gene expression profile information data of the breast cancer patient by a TCGA database, performing single-factor COX regression analysis, and screening genes related to the total survival rate; and (3) screening genes related to the total survival rate through LASSO regression to obtain six model genes, and calculating a risk scoring formula according to the expression quantity of the six model genes to complete the construction of the breast cancer pulmonary metastasis prognosis prediction model. According to the invention, scRNA-seq technical data from GSE202695 and a TCGA database are combined to develop a breast cancer lung metastasis prognosis prediction model based on six model gene characteristics of breast cancer lung metastasis.

Description

Technical Field

[0001] The present invention relates to the technical field of medical informatics, and in particular to a method for constructing a breast cancer lung metastasis prognosis prediction model. Background Art

[0002] Breast cancer is a widespread health concern and the second leading cause of cancer-related death in women. Its incidence has been on the rise since the mid-2000s, increasing at a rate of 5.2% per year. Metastasis, the spread of cancer cells from the primary site to distant organs, remains a major challenge and the leading cause of breast cancer mortality. Approximately 90% of breast cancer deaths are due to complications related to metastasis. Breast cancer metastasis is organ-specific, with the bone being the most common site of involvement, accounting for 75% of cases. However, the prognosis for bone metastasis is good, with a 5-year disease-free survival (DFS) rate of 89.2% for guideline-recommended anti-cancer therapy combined with RANKL inhibitors. Lung metastasis, on the other hand, is the second most common site of distant breast cancer metastasis. Once it develops, the prognosis is poor, with a 5-year overall survival (OS) rate of only 16.8%. One major reason for this is the current inability to accurately predict the occurrence of lung metastasis in breast cancer, leading to delayed treatment. Current clinical indicators, such as ER, PR, HER2, Ki67, and other pathological features, are not accurate predictors of lung metastasis.

[0003] Studies have shown that multiple factors, such as VEGF (vascular endothelial growth factor), TGFβ (transforming growth factor β), TNF (tumor necrosis factor), CSF3 (colony stimulating factor 3), SPARC (secreted protein acidic and rich in cysteine), MMPs (matrix metalloproteinases), and COX20 (cytochrome C oxidase assembly factor COX20), play a role in the development and progression of breast cancer lung metastasis. However, the detection of these factors can be challenging due to high costs and inconvenient detection methods, limiting their clinical use.

[0004] Single-cell RNA sequencing (scRNA-seq) is a new technology that performs high-throughput sequencing of mRNA at the single-cell level. This method involves amplifying small amounts of mRNA from a single isolated cell, followed by high-throughput sequencing. scRNA-seq overcomes the limitations of tissue samples and can identify heterogeneity at the single-cell level, making it possible to study gene expression patterns in individual cells within a tissue or microenvironment. However, there is currently a lack of research on prognostic prediction models for breast cancer lung metastasis based on single-cell transcriptomics.

[0005] Therefore, it is necessary to provide a method for constructing a prognostic prediction model for breast cancer lung metastasis, thereby providing more effective diagnostic strategies and predictive tools to promote early detection and treatment. Deepening the understanding of the mechanism of organ-specific metastasis of breast cancer will further develop new therapies and personalized treatment strategies, thereby bringing better outcomes for patients. Summary of the Invention

[0006] The present invention aims to address at least one of the technical problems existing in the aforementioned prior art. To this end, the present invention proposes a method for constructing a prognostic prediction model for breast cancer lung metastasis, thereby providing more effective diagnostic strategies and predictive tools to promote early detection and treatment. This will deepen our understanding of the mechanisms of organ-specific breast cancer metastasis and further develop new therapies and personalized treatment strategies, thereby achieving better outcomes for patients.

[0007] A first aspect of the present invention provides a method for constructing a prognostic prediction model for breast cancer lung metastasis.

[0008] Specifically, the following steps are included:

[0009] (1) Obtain single-cell RNA sequencing data of breast cancer in situ and lung metastasis from the GEO database and preliminarily screen differentially expressed genes;

[0010] (2) Clinical information and gene expression profiles of breast cancer patients were obtained from the TCGA database, and univariate COX regression analysis was performed to screen genes associated with overall survival;

[0011] (3) LASSO regression was used to screen genes related to overall survival rate, and 6 model genes were obtained. The risk score formula was calculated based on the expression levels of the 6 model genes to complete the construction of the breast cancer lung metastasis prognosis prediction model.

[0012] Preferably, in step (1), the step of preliminary screening of differentially expressed genes includes performing quality control on primary and metastatic subpopulations, eliminating poor quality data, and screening out variable genes, performing PCA linear dimensionality reduction processing on the data using JackStraw Plot and Eblow Plot methods, as well as t-SNE and UMAP nonlinear dimensionality reduction processing, visualizing cell clusters, and analyzing variable genes using Limma to screen out differentially expressed genes.

[0013] Preferably, in step (3), the LASSO regression screening method is performed using the glmnet package in the R language.

[0014] Preferably, in step (3), the six model genes are MUC16, MAP3K13, NR4A3, FAM111A, ZNF562, and NRP1.

[0015] Preferably, the Gene ID of the MUC16 is 94025; the Gene ID of the MAP3K13 is 9175, the Gene ID of the NR4A3 is 8013, the Gene ID of the FAM111A is 63901, the Gene ID of the ZNF562 is 54811, and the Gene ID of the NRP1 is 8829.

[0016] Preferably, the risk score formula is: risk score = 0.030859*(expression level of MUC16)-0.08401*(expression level of MAP3K13)+0.028720092*(expression level of NR4A3)-0.006563639*(expression level of FAM111A)+0.018508065*(expression level of ZNF562)+0.005311887*(expression level of NRP1).

[0017] Preferably, when the risk score is less than or equal to -0.10465, it is a low risk score; when the risk score is greater than -0.10465, it is a high risk score.

[0018] A second aspect of the present invention provides a molecular marker for detecting lung metastasis of breast cancer.

[0019] Specifically, the detected molecular markers are MUC16, MAP3K13, NR4A3, FAM111A, ZNF562, and NRP1.

[0020] A third aspect of the present invention provides use of a molecular marker for detecting breast cancer lung metastasis in the preparation of a product for detecting breast cancer lung metastasis.

[0021] Preferably, the product includes a kit and a chip.

[0022] Compared with the prior art, the present invention has the following beneficial effects:

[0023] This study combined scRNA-seq data from GSE202695 and the TCGA database to develop a prognostic model for breast cancer lung metastasis based on six model gene signatures. By analyzing the expression levels of these six model genes and their interactions with the tumor environment, we further identified potential gene signatures closely associated with the development of lung metastasis in breast cancer patients, providing a more effective and personalized tool for breast cancer prediction and treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 This is a single factor COX regression analysis diagram of the present invention;

[0025] Figure 2This is a screening analysis diagram of 10 genes related to overall survival rate in the present invention;

[0026] Figure 3 It is the LASSO regression analysis diagram of the present invention;

[0027] Figure 4 It is the ROC curve diagram of the present invention;

[0028] Figure 5 This is a graph showing the predicted performance results of the present invention;

[0029] Figure 6 This is the Kaplan-Meier analysis diagram of the present invention. DETAILED DESCRIPTION

[0030] In order to make the technical solution of the present invention more clearly understood by those skilled in the art, the following examples are given for illustration. It should be noted that the following examples do not limit the scope of protection claimed by the present invention.

[0031] Unless otherwise specified, the raw materials, reagents, or devices used in the following examples can be obtained from conventional commercial sources or by existing known methods.

[0032] Example 1

[0033] Method for constructing a prognostic prediction model for breast cancer lung metastasis.

[0034] The following steps are involved:

[0035] (1) The GSE202695 database was selected, which was used to perform single-cell RNA sequencing (scRNA-Seq) on primary tumor tissues of triple-negative breast cancer in NSG mice and lung metastasis tissues of matched breast cancer patient-derived xenografts (PDX). Quality control of primary and metastatic subpopulations was then performed to remove poor-quality data, and 2,000 genes with high inter-cell variation (variable genes) were screened. The data were processed using the JackStraw Plot method and the Eblow Plot method for PCA linear dimensionality reduction, as well as t-SNE and UMAP nonlinear dimensionality reduction to visualize cell clusters. Limma was further used to analyze the differentially expressed genes among the 2,000 variable genes.

[0036] (2) The clinical information and gene expression profiles of breast cancer patients (n=1056) were obtained from the TCGA database, and 2000 variable genes were analyzed using the “Survival package” in R language, such as Figure 1As shown in Figure 2, univariate COX regression analysis was performed (TCGA survival status: 1 = death, 0 = survival; TCGA survival time: days; gene expression: from GSE202695), and 25 genes associated with overall survival were screened. The 25 genes were further screened (hazard ratio HR value was less than 1, HR ≤ 0.930 / HR ≥ 1.043, p < 0.05), and 10 genes associated with overall survival were selected (such as Figure 2 shown).

[0037] (3) Figure 3 As shown, LASSO regression was performed using the glmnet package in R to screen for genes associated with overall survival. Six model genes were identified: MUC16, MAP3K13, NR4A3, FAM111A, ZNF562, and NRP1. A risk score was calculated based on the expression levels of these six model genes: risk score = 0.030859*(MUC16 expression) - 0.08401*(MAP3K13 expression) + 0.028720092*(NR4A3 expression) - 0.006563639*(FAM111A expression) + 0.018508065*(ZNF562 expression) + 0.005311887*(NRP1 expression). The coefficients were automatically generated by the glmnet package after analysis. This completed the construction of a prognostic prediction model for breast cancer lung metastasis.

[0038] Example 2

[0039] 1. Effectiveness verification.

[0040] The model was further validated: the ROC curve showed that the prognostic model AUC = 82.8% (HR: 67%-98.6%), as shown in Figure 4 As shown in Figure 2, the prediction efficiency of 1 year, 3 years and 5 years was AUC=70.9%, 64.7% and 66.7% respectively, which has good prediction efficiency (such as Figure 5 shown).

[0041] 2. Kaplan-Meier analysis.

[0042] Kaplan-Meier analysis of the model revealed an optimal cutoff value for the risk score of -0.10465, which allowed patients to be divided into a high-risk group and a low-risk group for breast cancer lung metastasis (p = 6.4e-8, HR = 4.13, 95% CI (2.36, 7.22). This model can calculate a risk score formula based on the expression levels of the six model genes, thereby predicting the risk of breast cancer lung metastasis (e.g., Figure 6 shown).

[0043] The above describes in detail the preferred embodiments of the present invention. It should be understood that those skilled in the art can make numerous modifications and variations based on the concepts of the present invention without inventive effort. Therefore, any technical solutions derived from modifications, equivalent substitutions, improvements, etc. made by those skilled in the art based on the concepts of the present invention through logical analysis, reasoning, or limited experimentation on the basis of the prior art shall be within the scope of protection defined by the claims.

Claims

1. A method for constructing a prognostic prediction model for lung metastasis of breast cancer, characterized in that: The following steps are involved: (1) Obtain single-cell RNA sequencing data of breast cancer in situ and lung metastasis from the GEO database and preliminarily screen differentially expressed genes; (2) The clinical information and gene expression profile information of breast cancer patients were obtained from the TCGA database, and univariate COX regression analysis was performed to screen genes related to overall survival; (3) Genes associated with overall survival were screened through LASSO regression to obtain 6 model genes. The risk score formula was calculated based on the expression levels of the 6 model genes to complete the construction of the breast cancer lung metastasis prognosis prediction model.

2. The construction method according to claim 1, characterized in that: In step (1), the preliminary screening of differentially expressed genes includes quality control of primary and metastatic subgroups, elimination of poor quality data, screening of variable genes, performing PCA linear dimensionality reduction processing on the data using JackStraw Plot and Eblow Plot methods, and t-SNE and UMAP nonlinear dimensionality reduction processing, visualizing cell clusters, analyzing variable genes using Limma, and screening of differentially expressed genes.

3. The construction method according to claim 1, characterized in that: In step (3), the LASSO regression screening method is performed using the glmnet package in the R language.

4. The construction method according to claim 1, characterized in that: In step (3), the six model genes are MUC16, MAP3K13, NR4A3, FAM111A, ZNF562, and NRP1.

5. The construction method according to claim 4, characterized in that: The Gene ID of the MUC16 is 94025; the Gene ID of the MAP3K13 is 9175, the Gene ID of the NR4A3 is 8013, the Gene ID of the FAM111A is 63901, the Gene ID of the ZNF562 is 54811, and the Gene ID of the NRP1 is 8829.

6. The construction method according to claim 1, characterized in that: The formula of the risk score is: Risk score = 0.030859*(expression level of MUC16)-0.08401*(expression level of MAP3K13)+0.028720092*(expression level of NR4A3)-0.006563639*(expression level of FAM111A)+0.018508065*(expression level of ZNF562)+0.005311887*(expression level of NRP1).

7. The construction method according to claim 6, characterized in that: When the risk score is less than or equal to -0.10465, it is a low risk score; when the risk score is greater than -0.10465, it is a high risk score.

8. A molecular marker for detecting lung metastasis of breast cancer, characterized in that: The detected molecular markers are MUC16, MAP3K13, NR4A3, FAM111A, ZNF562, and NRP1.

9. Use of the molecular marker for detecting lung metastasis of breast cancer according to claim 8 in the preparation of a product for detecting lung metastasis of breast cancer.

10. The use according to claim 9, characterized in that: The products include test kits and chips.

Images