Triple negative breast cancer tumor cell pathway clustering method, applications, and systems

Abstract

The application discloses a triple-negative breast cancer cell channel grouping method, application and system, and relates to the technical field of tumor treatment, and can be used for treating triple-negative breast cancer. The method comprises the following steps: obtaining triple-negative breast cancer scRNA-seq data of a plurality of target samples; processing the triple-negative breast cancer scRNA-seq data to obtain first data of target tumor cells, and reprocessing the first data to obtain a channel correction enrichment score of a single target tumor cell; grouping the target tumor cells according to the correction enrichment score to obtain target tumor cell groups; and identifying characteristic activation channels of each target tumor cell group. The method can group triple-negative breast cancer tumor cells to obtain cell groups with different gene channel expression modes, and can formulate a treatment scheme according to the channel characteristics of each cell group, which has important reference value for clinical individualized treatment and prognosis prediction. The application also discloses a system for implementing the method and application of the method in finding clinical treatment targets and predicting prognosis.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of biological medicine, in particular to a triple negative breast cancer cell pathway grouping method, application and system. BACKGROUND

[0002] Triple negative breast cancer (TNBC) is a type of breast cancer characterized by negative estrogen receptor, progesterone receptor and human epidermal growth factor receptor, accounting for about 15-20% of all invasive breast cancers. Triple negative breast cancer often occurs in young women and presents rapid progression and high invasiveness. Due to the lack of effective therapeutic targets, the treatment of triple negative breast cancer mainly relies on systemic chemotherapy, and the prognosis is poor with high recurrence rate. Subsequent studies have found that triple negative breast cancer is a highly heterogeneous malignant tumor, and the heterogeneity between tumors leads to differences in survival between patients and different sensitivities to treatment regimens. Therefore, further grouping of triple negative breast cancer patients and finding clinical treatment targets for each group are crucial for improving patient prognosis.

[0003] Single-cell RNA-sequencing (scRNA-seq) is a new technology that can perform unbiased, high-throughput, and high-resolution whole transcriptome analysis at the single-cell level. For tumor tissues and other highly heterogeneous samples, traditional high-throughput sequencing technology can only provide the average value of the transcription level of all cells in the sample, but scRNA-seq can accurately depict the transcriptome characteristics of each cell in the sample. Currently, single-cell transcriptome sequencing technology is the main method for exploring breast cancer heterogeneity, which can help us understand the transcriptome characteristics of different types of cells in breast tumor tissue, analyze intratumoral heterogeneity, and find potential clinical treatment targets.

[0004] Existing studies have shown that the single-cell transcriptome of tumor cells varies greatly among different patients, with both intratumoral and intertumoral heterogeneity. Pathway typing of tumor cells is a method of classifying cells based on the expression pattern of their gene pathways, which can group tumor cells with similar signal pathway activation, biological activity, and treatment sensitivity into the same group. This method not only takes into account intertumoral and intratumoral heterogeneity, but also has higher biological stability and better performance than gene expression-based classification methods.

[0005] Currently, the analysis of breast cancer single-cell transcriptome sequencing data is mostly focused on immune cells such as T lymphocytes, B lymphocytes and macrophages, and the analysis of tumor cells is less. Therefore, it is of great significance to establish a triple-negative breast cancer cell pathway clustering method, reveal the intratumoral and intertumoral heterogeneity of triple-negative breast cancer tumor cells, identify tumor cell groups with similar signal pathway activation and treatment sensitivity, and study the drug treatment sensitivity of tumor cells, find treatment targets and predict patient clinical prognosis. SUMMARY

[0006] The present application provides a triple-negative breast cancer cell pathway clustering method, application and system, which clusters triple-negative breast cancer tumor cells by the method to obtain cell groups with different gene pathway expression patterns, and formulates a treatment plan according to the pathway characteristics of each group of cells, which has important reference value for clinical individualized treatment and prognosis prediction.

[0007] To achieve the above object, the technical scheme adopted by the present application is as follows:

[0008] The present application provides a triple-negative breast cancer cell pathway clustering method, which comprises the following steps:

[0009] Obtain triple-negative breast cancer scRNA-seq data of a plurality of target samples;

[0010] Process the triple-negative breast cancer scRNA-seq data to obtain first data of target tumor cells, and process the first data to obtain a pathway-corrected enrichment score of a single target tumor cell;

[0011] According to the corrected enrichment score, the target tumor cells are clustered to obtain a target tumor cell group;

[0012] Identify the characteristic activation pathway of each target tumor cell group.

[0013] Further, the triple-negative breast cancer scRNA-seq data of a plurality of target samples is obtained, specifically as follows:

[0014] Obtain triple-negative breast cancer scRNA-seq data of a plurality of target samples, and the triple-negative breast cancer scRNA-seq data is a triple-negative breast cancer scRNA-seq gene expression matrix.

[0015] Further, the triple-negative breast cancer scRNA-seq data is processed to obtain first data of target tumor cells, and the first data is processed to obtain a pathway-corrected enrichment score of a single target tumor cell, comprising the following steps:

[0016] After quality control, data filtering, data merging, and batch effect removal of the triple negative breast cancer scRNA-seq gene expression matrix, the epithelial cells are annotated by using the Garnett algorithm;

[0017] The copy number variation of the epithelial cells is evaluated by using the inferCNV algorithm to distinguish normal epithelial cells from target tumor cells, and the first data of the tumor epithelial cells are extracted; the target tumor cells are tumor epithelial cells, and the first data are scRNA-seq gene expression matrices;

[0018] The first data are standardized, and the gene expression values of the first data are converted into rankings;

[0019] Pathway gene sets are selected from a reference database, and the single-cell Mann-Whitney-Wilcoxon gene set test is used to calculate the corrected enrichment scores of each pathway of each target tumor cell.

[0020] Further, the target tumor cells are clustered to obtain target tumor cell groups, including the following steps:

[0021] Each target sample is subjected to consistency clustering, and each target sample is divided into multiple subgroups;

[0022] The subgroups are combined, and the Jaccard distance between each subgroup is calculated;

[0023] According to the Jaccard distance, consistency clustering is performed again, the optimal cluster number is determined by the Calinsky Criteria criterion, the clustering effect is verified by the silhouette clustering, the target tumor cells are clustered, and target tumor cell groups are obtained.

[0024] Further, the characteristic activated pathways of each target tumor cell group are identified, specifically:

[0025] The target tumor cell groups are subjected to two-sided Mann-Whitney-Wilcoxon test, and the characteristic activated pathways of each target tumor cell group relative to other groups are screened out.

[0026] Further, the quality control of the triple negative breast cancer scRNA-seq gene expression matrix is performed, specifically:

[0027] The triple negative breast cancer scRNA-seq gene expression matrix is screened by a data screening standard, and the data screening standard includes:

[0028] (1) The target sample is from a patient who has not received treatment;

[0029] (2) The target sample is a sample that has not been subjected to cell sorting;

[0030] (3) the sequencing method of the target sample is 10X genomics.

[0031] Further, the triple negative breast cancer scRNA-seq data is from C5 ontology gene set of MsigDB database;

[0032] The target sample is downloaded from the GEO database, and the triple negative breast cancer scRNA-seq gene expression matrix of the target sample is subjected to quality control.

[0033] In a second aspect of the present application, an application of a triple negative breast cancer cell pathway clustering method in searching for a clinical treatment target and predicting prognosis is provided, and a tumor cell clustering signature is constructed by the triple negative breast cancer cell pathway clustering method to search for a clinical treatment target and a model for predicting prognosis.

[0034] In a third aspect of the present application, a system for performing a tumor cell pathway clustering method is provided, which is used to implement the triple negative breast cancer cell pathway clustering method, and the system comprises:

[0035] A sequencing data acquisition module is configured to acquire triple negative breast cancer single cell transcriptome sequencing data of a target sample;

[0036] A first processing module is configured to process the triple negative breast cancer single cell transcriptome sequencing data to obtain first data of a target tumor cell, and to process the first data again to obtain a pathway-corrected enrichment score of a single target tumor cell;

[0037] A second processing module is configured to cluster the target tumor cells according to the corrected enrichment score to obtain a target tumor cell cluster;

[0038] A pathway identification module is configured to identify a characteristic activated pathway of each target tumor cell cluster to search for a clinical treatment target.

[0039] In a fourth aspect of the present application, an electronic device is provided, which comprises a processor and a memory, the memory stores a plurality of instructions, and the processor is configured to read the instructions and perform the tumor cell pathway clustering method.

[0040] Compared with the prior art, the present application has the following beneficial effects:

[0041] 1. The present application provides a triple negative breast cancer cell pathway grouping method, which is aimed at triple negative breast cancer scRNA-seq data of tumor epithelial cell pathway enrichment grouping and identification of the characteristic activation pathway of each group, systematically groups and describes the tumor epithelial cells from the aspect of molecular pathways, and provides a new method for the precise treatment of triple negative breast cancer, which has a transformation value.

[0042] 2. The present application provides an application of a triple negative breast cancer cell pathway grouping method in finding clinical treatment targets and predicting prognosis, which constructs a tumor cell grouping signature, finds a model for clinical treatment targets and prognosis prediction, and has great significance for studying the sensitivity of tumor cells to drug treatment, finding treatment targets, and predicting patient clinical prognosis through the above-mentioned triple negative breast cancer cell pathway grouping method.

[0043] 3. The present application provides a system for performing a tumor cell pathway grouping method, which is used to implement the above-mentioned triple negative breast cancer cell pathway grouping method, and the system comprises: a sequencing data acquisition module for acquiring triple negative breast cancer single cell transcriptome sequencing data of a target sample; a first processing module for processing the triple negative breast cancer single cell transcriptome sequencing data to obtain first data of a target tumor cell, and reprocessing the first data to obtain a pathway corrected enrichment score of a single target tumor cell; a second processing module for grouping the target tumor cells according to the corrected enrichment score to obtain a target tumor cell group; and a pathway identification module for identifying the characteristic activation pathway of each target tumor cell group and finding a clinical treatment target. Through the system, the above-mentioned triple negative breast cancer cell pathway grouping method can be implemented. BRIEF DESCRIPTION OF DRAWINGS

[0044] Figure 1 It is a flowchart of the triple negative breast cancer cell pathway grouping method of the present application;

[0045] Figure 2 It is a principle structure diagram of the system for performing a tumor cell pathway grouping method of the present application;

[0046] Figure 3 The present application is based on the technical roadmap of tumor cell pathway grouping of triple negative breast cancer scRNA-seq data;

[0047] Figure 4 It is an annotation and grouping diagram of a target tumor cell in an embodiment of the present application, wherein A diagram is a target tumor cell annotation diagram, B diagram is a tSNE dimension reduction diagram of the pathway enrichment situation of seven groups of target tumor cells, C diagram shows a characteristic activation pathway diagram of each group of target tumor cells, and D diagram shows a Cytoscape enrichment result diagram of the characteristic activation pathway of each group of target tumor cells.

[0048] Figure 5 Figure for tumor cell subpopulation effect in one embodiment of the present application, where A is a consistency clustering prompt that the clustering effect is good when the clustering number is 7; B is a silhouette coefficient test prompt that the clustering effect is good when the clustering number is 7; C figure prompts the proportion of each tumor cell group in all target samples;

[0049] Figure 6 Figure for finding clinical target and prognosis prediction in one embodiment of the present application, where A is a FUSCC TNBC queue, PC3 signature is positively correlated with LSTm, LOH and HRD, and B is a DOP treatment arm pCR group PC3 signature significantly higher than non-pCR group in I-SPY2 queue; C is a RFS survival curve graph of C5 and C7 signature high and low groups in METABRIC queue; D is a RFS survival curve graph of C5 and C7 signature high and low groups in FUSCC queue;

[0050] Figure 7 Figure for electronic device structure in one embodiment of the present application. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme of the embodiments of the present application will be described clearly and completely below in combination with the drawings in the embodiments of the present application. Obviously, the embodiments described below are part of the embodiments of the present application, not all the embodiments. The components of the embodiments of the present application described and shown in the drawings herein can be arranged and designed in various different configurations.

[0052] Therefore, the detailed description of the embodiments of the present application provided below in combination with the drawings is intended to represent only selected embodiments of the present application, not to limit the scope of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without making creative efforts fall within the scope of the present application.

[0053] In the present application, the term "Triple Negative Breast Cancer (TNBC)" is a type of breast cancer characterized by negative estrogen receptor, progesterone receptor and human epidermal growth factor receptor, accounting for about 15-20% of all invasive breast cancers.

[0054] The term "single-cell RNA-sequencing (scRNA-seq)" is a new technology that can perform unbiased, high-throughput, high-resolution whole transcriptome analysis at the single-cell level.

[0055] The term "enrichment" means that the proportion of genes or substances annotated to a certain metabolic pathway in all differential genes or substances is significantly greater than the proportion of genes or substances annotated to a certain metabolic pathway in all background genes or substances.

[0056] The term "Enrichment Score (ES)" reflects the degree of enrichment of members s of a gene set (such as a gene set within a certain pathway) at both ends of a ranked gene set L (such as a differential gene set ranked according to logFC, by default in descending order, so that the up-regulated genes are at the top). The enrichment score ES is finally defined as the maximum peak value. A positive ES indicates that the gene set S is enriched at the top of the gene set L, and a negative ES indicates that the gene set S is enriched at the bottom of the gene set L. Therefore, the enrichment score is dynamic. Because the size of each gene set is different, and the enrichment score is related to the size of the gene set, it needs to be normalized to obtain the corrected enrichment score (Normalized Enrichment Score, NES), which is a constant used to compare the enrichment degree of the expression data set in different functional gene data sets.

[0057] The term "Garnett algorithm" is an interpretable, hierarchical labeling language based on cell type-specific marker genes. Garnett successfully classifies cell types in tissue and whole organism datasets, as well as across species. In simple terms, the Garnett algorithm takes as input a labeling file that defines cell types using marker genes and constructs a cell type hierarchy that can contain cell subtypes. Next, Garnett trains a classifier at each node starting from the root of the tree by comparing cell type representatives using elastic net multinomial regression. Finally, Garnett hierarchically classifies all cells and optionally provides a second extension to classify the clustering results.

[0058] The term "inferCNV algorithm" is a more authoritative single-cell CNV analysis tool developed by the Broad Institute, which analyzes the idea of comparing the gene expression of each tumor cell with the average expression or "normal" reference cell gene expression, and displaying the relative expression of each chromosome in the form of a heat map. It can be seen directly that compared with normal cells, the tumor cell genome will have large-scale overexpression or underexpression.

[0059] The rows of the term "gene expression matrix" represent the expression of a gene under different environmental conditions or at different time points, and the columns represent the expression of all genes under different conditions or samples (such as tissues, experimental conditions, treatment factors, etc.), and the data in each cell represents the expression level of a specific gene in a specific sample.

[0060] The term "consensus clustering" is a method for providing quantitative evidence for determining the number and members of possible clusters in a dataset, such as microarray gene expression. This method has been widely used in cancer genomics, where new molecular subtypes of disease have been discovered. The consensus clustering method includes subsampling from a set of items, such as microarrays, and determining clusters of a specific cluster count (k). Then, the consensus value, which is the number of times two items have the same cluster occurring in the same subsample, is calculated and stored in a symmetric consensus matrix for each k.

[0061] The term "Jaccard Distance" is an index used to measure the difference between two sets, which is the complement of the Jaccard similarity coefficient, defined as 1 minus the Jaccard similarity coefficient. While the Jaccard similarity coefficient, also known as the Jaccard Index, is an index used to measure the similarity between two sets.

[0062] The term "Calinski-Harabasz criterion", sometimes referred to as Variance Ratio Criterion (VRC), can be used to determine the optimal K value of clustering. The Calinski Harabasz index is defined as:

[0063]

[0064] where K is the number of clusters, N is the number of samples, SSB is the sum of squares error between groups, and SSw is the sum of squares error within groups. Therefore, if SSw is smaller and SSB is larger, the clustering effect will be better, that is, the Calinsky criterion value is larger, the clustering effect is better.

[0065] The term "silhouette" refers to a method that reflects the consistency of data clustering results, which can be used to evaluate the dispersion degree between clusters after clustering. The value of silhouette ranges from -1 to 1, if the silhouette of a sample is close to 1, it means that the sample clustering result is reasonable; if it is close to -1, it means that it should be classified into other clusters; if the silhouette is close to 0, it means that the sample is on the boundary of two clusters. The average of all sample silhouettes is called the silhouette coefficient of the clustering result, which is a measure of whether the clustering result is reasonable and effective.

[0066] The term "Mann-Whitney Wilcoxon test", also known as Mann-Whitney U test, is a common statistical test used to compare two samples or groups. It is used in many fields, including economics, biological sciences, and epidemiology. It is particularly useful when you are evaluating differences between two independent groups, with a small number of individuals in each group (usually fewer than 30), not normally distributed, and the data is continuous.

[0067] The term "10x Genomics" is one of the mainstream platforms for single-cell transcriptome sequencing technology, which can realize large-scale single-cell transcriptome sequencing, has the advantages of high cell throughput, high cell capture rate, short project cycle, and is widely used in cell heterogeneity, immune cell population detection, and cell atlas construction research.

[0068] Embodiment one

[0069] Referring to Figure 1 , Figure 3 The present application provides a triple-negative breast cancer cell pathway grouping method, and the present embodiment uses R software (http: / / www.R-project.org) loaded on an Ubuntu 16.04 LTS system (https: / / ubuntu.com / 16-04) for bioinformatics analysis. The method comprises the following steps:

[0070] Step 101, obtaining triple-negative breast cancer scRNA-seq data of a plurality of target samples.

[0071] In the present embodiment, the target samples are 29 patients from 3 triple-negative breast cancer scRNA-seq cohorts selected and downloaded from the GEO database (https: / / www.ncbi.nlm.nih.gov / geo / ).

[0072] Step 102, processing the triple-negative breast cancer scRNA-seq data to obtain first data of target tumor cells, and reprocessing the first data to obtain a pathway-corrected enrichment score of a single target tumor cell.

[0073] Step 103, grouping the target tumor cells according to the corrected enrichment score to obtain a target tumor cell group;

[0074] Step 104, identifying the characteristic activated pathway of each target tumor cell group.

[0075] Optionally, in step 101, the triple-negative breast cancer scRNA-seq data of a plurality of target samples is obtained, specifically:

[0076] Obtaining triple negative breast cancer scRNA-seq data of a plurality of target samples, the triple negative breast cancer scRNA-seq data being a triple negative breast cancer scRNA-seq gene expression matrix.

[0077] Optionally, in step 102, the triple negative breast cancer scRNA-seq data is processed to obtain first data of a target tumor cell, and the first data is processed to obtain a pathway-corrected enrichment score of a single target tumor cell, including the following steps:

[0078] In step 201, after quality control, data filtering, data merging, and removal of batch effects are performed on the triple negative breast cancer scRNA-seq gene expression matrix, Garnett algorithm is used to annotate epithelial cells;

[0079] In step 202, copy number variation of the epithelial cells is evaluated by an inferCNV algorithm to distinguish normal epithelial cells from target tumor cells, and first data of the tumor epithelial cells is extracted; the target tumor cell is a tumor epithelial cell, and the first data is a scRNA-seq gene expression matrix;

[0080] In step 203, the first data is standardized, and gene expression values of the first data are converted into rankings;

[0081] In step 204, a pathway gene set is selected from a reference database, and a single cell Mann-Whitney-Wilcoxon Gene Set test (scMWW-GST) is used to calculate a pathway-corrected enrichment score of each target tumor cell.

[0082] In step 203, the gene expression values are converted into rankings, which generally uses a GSEA ranking method. The specific implementation process of the GSEA ranking method is as follows: first, gene expression profile data is sorted according to the expression level of genes, and then gene sets are sorted according to their relevance to a specific biological process. Next, the enrichment score of the gene set is calculated to determine the relevance of the gene set to the specific biological process. Finally, the gene set is sorted according to its enrichment score to find the gene set related to the specific biological process.

[0083] After cell filtering and cell annotation of the scRNA-seq data of the 29 target samples, 31789 tumor epithelial cells (such as Figure 4After screening, 8799 pathways were retained from 31789 tumor epithelial cells, and the normalized enrichment score (NES) of each pathway of the target tumor cell was calculated to form a normalized enrichment score (NES) matrix for subsequent tumor epithelial cell clustering.

[0084] Optionally, in step 103, the target tumor cells are clustered to obtain a target tumor cell group, including the following steps:

[0085] Step 301, consistent clustering is performed on each target sample, and each target sample is divided into multiple subgroups;

[0086] Step 302, the subgroups are merged, and the Jaccard distance between each subgroup is calculated;

[0087] Step 303, consistent clustering is performed again according to the Jaccard distance, the optimal cluster number is determined by Calinsky Criteria criterion, the clustering effect is verified by silhouette clustering, and the target tumor cells are clustered to obtain a target tumor cell group.

[0088] Optionally, in step 104, the characteristic activated pathways of each target tumor cell group are identified, specifically:

[0089] The target tumor cell group is subjected to bilateral Mann-Whitney-Wilcoxon test, and the characteristic activated pathways of each target tumor cell group are screened out from other groups.

[0090] The normalized enrichment score (NES) of 8799 pathways of each tumor epithelial cell is calculated, and clustering is performed based on this, and the tumor epithelial cells are divided into 7 groups (as shown in Figure 4 B, Figure 5 A, 5B, 5C), named C1, C2, C3, C4, C5, C6, C7. Then the characteristic activated pathways (logFoldChange>0.3, qValue<0.01) of each group of tumor epithelial cells are identified by Mann-Whitney-Wilcoxon test, and the 7 groups of tumor epithelial cells have 172, 82, 1164, 48, 37, 112, 503 characteristic activated pathways respectively, and the top 100 characteristic activated pathways of each group of tumor epithelial cells are subjected to pathway enrichment analysis, and it is found that the neurotransmitter related pathway is activated in C1 subgroup, and the transcription initiation, elongation and termination related pathway is activated in C3 subgroup (as shown in Figure 4 C, 4D).

[0091] Optionally, in step 201, the triple-negative breast cancer scRNA-seq gene expression matrix is subjected to quality control, specifically:

[0092] The triple-negative breast cancer scRNA-seq gene expression matrix is filtered by a data filtering standard, and the data filtering standard includes:

[0093] (1) The target sample is from a patient who has not received treatment;

[0094] (2) The target sample is a sample that has not been subjected to cell sorting;

[0095] (3) The sequencing method of the target sample is 10X genomics.

[0096] Optionally, in step 101, the triple-negative breast cancer scRNA-seq data is from the C5 ontology gene set of the MsigDB database. The gene set selected in this embodiment is from the C5 ontology gene set of the MsigDB database (https: / / www.gsea-msigdb.org), which contains three parts of tumor-related molecular functions, cell components and biological processes. After screening, 8799 pathways are retained for subsequent analysis.

[0097] In step 101, the target sample is downloaded from the GEO database, and the triple-negative breast cancer scRNA-seq gene expression matrix of the target sample is subjected to quality control.

[0098] The data sources in this embodiment: 29 patients from 3 triple-negative breast cancer scRNA-seq cohorts, all downloaded from the GEO database (https: / / www.ncbi.nlm.nih.gov / geo / ). The data filtering standard is: (1) the sample is from a patient who has not received treatment; (2) the sample has not been subjected to cell sorting; (3) the sequencing method is 10X genomics.

[0099] The triple-negative breast cancer cell pathway clustering method used in this embodiment is aimed at triple-negative breast cancer scRNA-seq data, and is used for tumor epithelial cell pathway enrichment clustering and identification of characteristic activation pathways of each group. From the perspective of molecular pathways, the tumor epithelial cells are systematically clustered and described, and different clusters are classified for treatment, providing a new method for precise treatment of triple-negative breast cancer and having a transformative value.

[0100] Embodiment Two

[0101] The embodiment of the present application provides a model for finding a clinical treatment target and predicting prognosis based on a triple negative breast cancer cell pathway grouping method, wherein the model adopts the triple negative breast cancer cell pathway grouping method in the embodiment 1, and tumor cell grouping signatures are constructed according to the method.

[0102] The data source of the embodiment is as follows: (1) METABRIC cohort: 320 TNBC patients are downloaded from the cBio-Portal for Cancer Genomic (https: / / www.cbioportal.org / ) database; (2) FUSCC cohort: 360 TNBC patients are downloaded from The National Omics Data Encyclopedia (NODE) (http: / / www.biosino.org / node) database; (3) I-SPY2 cohort: 362 TNBC patients are downloaded from the GEO (http: / / www.ncbi.nlm.nih.gov / geo / ) database.

[0103] The embodiment adopts R software (http: / / www.R-project.org), and the specific analysis method is as follows:

[0104] The embodiment adopts the triple negative breast cancer cell pathway grouping method in the embodiment 1, and the FindMarkers algorithm is used to identify the differentially expressed genes (logFoldChange>1, p<0.01) of each group. After removing the repeated genes between groups, 50 genes with the largest difference in each group are extracted to construct a group-specific gene set, which is used to calculate the signature.

[0105] The embodiment extracts the RNA-seq data of each cohort, calculates the signature of each sample of each group based on the group-specific gene set by using the GSVA algorithm, and analyzes the relationship between the signature and the survival prognosis. The multi-omics characteristics of the samples with high signature scores are analyzed, and the treatment target is found according to the characteristics.

[0106] In the embodiment, in the FUSCC cohort, the C3 signature score is positively correlated with LSTm, LOH and HRD, which indicates that the patients with high C3 signature scores have homologous recombination repair defects (such as BRCA1 / 2 mutations, as shown in FIG. 8A), and the I-SPY2 cohort verifies that the patients with high C3 signature scores are more likely to benefit from olaparib (as shown in FIG. 8B). Figure 6 Figure 6 ​(As shown in B). In both the FUSCC and METABRIC cohorts, it was found that patients with higher C5 signature scores had worse RFS, while patients with higher C7 signature scores had better RFS (e.g., ...). Figure 6 (as shown in C, 6D).

[0107] In summary, the application of the triple-negative breast cancer cell pathway clustering method in this embodiment for identifying clinical therapeutic targets and predicting prognosis, constructing tumor cell cluster signatures, identifying clinical therapeutic targets, and predicting prognosis through the aforementioned triple-negative breast cancer cell pathway clustering method, is of great significance for studying the sensitivity of tumor cells to drug therapy, identifying therapeutic targets, and predicting the clinical prognosis of patients.

[0108] Example 3

[0109] See Figure 2 This embodiment provides a system for performing tumor cell pathway segmentation, used to implement the above-described triple-negative breast cancer cell pathway segmentation method. The system includes:

[0110] Sequencing data acquisition module 1 acquires single-cell transcriptome sequencing data of the target sample from triple-negative breast cancer.

[0111] The first processing module 2 is used to process the single-cell transcriptome sequencing of the triple-negative breast cancer to obtain the first data of the target tumor cells, and to process the first data again to obtain the pathway-corrected enrichment score of a single target tumor cell.

[0112] The second processing module 3 is used to group the target tumor cells according to the corrected enrichment score to obtain the target tumor cell population.

[0113] Module 4, the pathway identification module, is used to identify the characteristic activation pathways of various target tumor cell populations and to find clinical therapeutic targets.

[0114] See Figure 7 The present invention also provides an electronic device, including a processor 5 and a memory 6, wherein the memory 6 stores multiple instructions, and the processor 5 is used to read the instructions and execute the above-described tumor cell pathway grouping method.

[0115] The foregoing merely illustrates the principles of the application and application of its leading features. This application is not limited to the illustrative embodiments shown and described herein. Rather, this application is capable of operating within a further range of conditions and environments than those specifically described herein, and further modifications can be made without departing from the spirit or scope of the application. Accordingly, the description is to be construed as illustrative only and not restrictive of the broad disclosure or application of the application. The specification and drawings are, accordingly, to be regarded simply as an illustration of the broadest aspects of the application and in no way limiting of its scope. Any reference to claims in the specification is intended to refer to the claims as they are presented in the application, and not as the claims may be amended in future prosecution of the application. Thus, the scope of the application is indicated by the appended claims, rather than by the foregoing description.

[0116] Furthermore, it should be understood that although the description has been set forth in the context of certain embodiments, the present application is not limited to these embodiments alone. Rather, the description is intended to embrace all alternatives, modifications and variations of the present application, including those that are presently known or that are later developed.

Claims

1. A method for pathway clustering of triple-negative breast cancer cells, characterized in that, Includes the following steps: Obtain scRNA-seq data from multiple target samples of triple-negative breast cancer; The scRNA-seq data of the triple-negative breast cancer is processed to obtain the first data of the target tumor cells. The first data is then processed again to obtain the pathway-corrected enrichment score of a single target tumor cell. Based on the corrected enrichment fraction, the target tumor cells are grouped to obtain the target tumor cell population. Identify the characteristic activation pathways of each target tumor cell population; The acquisition of scRNA-seq data from multiple target samples of triple-negative breast cancer specifically involves: Obtain scRNA-seq data of triple-negative breast cancer from multiple target samples, wherein the triple-negative breast cancer scRNA-seq data is a triple-negative breast cancer scRNA-seq gene expression matrix; The process of processing the triple-negative breast cancer scRNA-seq data to obtain first data of target tumor cells, and processing the first data to obtain the pathway-corrected enrichment score of a single target tumor cell, includes the following steps: After quality control, data filtering, data merging, and batch effect removal of the scRNA-seq gene expression matrix of triple-negative breast cancer, epithelial cells were annotated using the Garnett algorithm. The copy number variation of the epithelial cells was evaluated using the inferCNV algorithm to distinguish normal epithelial cells from the target tumor cells, and the first data of the tumor epithelial cells was extracted; the target tumor cells are tumor epithelial cells, and the first data is the scRNA-seq gene expression matrix. The first data is standardized, and the gene expression values ​​of the first data are converted into a ranking. The pathway gene set was selected from the reference database, and the corrected enrichment score of each pathway in a single target tumor cell was calculated using the single-cell Mann-Whitney-Wilkerson test. The process of grouping the target tumor cells to obtain a target tumor cell population includes the following steps: Perform consistent clustering on each target sample to divide each target sample into multiple subgroups; The subgroups are merged, and the Jaccard distance between each subgroup is calculated. Consistent clustering is performed again based on the Jaccard distance. The optimal number of clusters is determined by the Calinsky Criteria criterion. The clustering effect is tested by contour clustering. The target tumor cells are then grouped to obtain the target tumor cell population.

2. The method for clustering triple-negative breast cancer cells according to claim 1, wherein the identification of characteristic activation pathways of each target tumor cell population is characterized in that, Specifically: The Mann-Whitney-Wilkerson test was used to screen out the characteristic activation pathways of each target tumor cell population compared to other populations.

3. The method for clustering triple-negative breast cancer cells according to claim 1, wherein the triple-negative breast cancer cells are... Quality control of cancer scRNA-seq gene expression matrix is ​​characterized by, Specifically: The triple-negative breast cancer scRNA-seq gene expression matrix was screened using data screening criteria, including: (1) The target samples were obtained from untreated patients; (2) The target sample is a sample that has not undergone cell sorting; (3) The sequencing method for the target sample is 10X genomics.

4. The method for pathway grouping of triple-negative breast cancer cells according to claim 1, characterized in that: The triple-negative breast cancer scRNA-seq data were obtained from the C5 ontology gene set in the MsigDB database; All target samples were downloaded from the GEO database, and the scRNA-seq gene expression matrices of triple-negative breast cancer samples were all subject to quality control.

5. An application of a triple-negative breast cancer cell pathway segmentation method in identifying clinical therapeutic targets and predicting prognosis, characterized in that, Using the triple-negative breast cancer cell pathway segmentation method described in any one of claims 1-4, a tumor cell segmentation signature is constructed to identify clinical treatment targets and predict prognosis.

6. A system for performing tumor cell pathway segmentation, used to implement the triple-negative breast cancer cell pathway segmentation method according to any one of claims 1-4, characterized in that, include: Sequencing data acquisition module (1) acquires single-cell transcriptome sequencing data of triple-negative breast cancer from the target sample; The first processing module (2) is used to process the single-cell transcriptome sequencing of the triple-negative breast cancer to obtain the first data of the target tumor cell, and to process the first data again to obtain the pathway-corrected enrichment score of a single target tumor cell. The second processing module (3) is used to group the target tumor cells according to the corrected enrichment fraction to obtain the target tumor cell population. The pathway identification module (4) is used to identify the characteristic activation pathways of each target tumor cell population and to find clinical therapeutic targets.

7. An electronic device, characterized in that, It includes a processor (5) and a memory (6), the memory (6) storing multiple instructions, and the processor (5) being used to read the instructions and execute the triple-negative breast cancer cell pathway clustering method as described in any one of claims 1-4.

Images