Identification method and system for multi-omics analysis of lung cancer brain metastasis driving gene

Abstract

The invention provides an identification method and system for multi-omics analysis of lung cancer brain metastasis driving genes. The method comprises the following steps: data quality control: acquiring single-cell RNA sequencing data from a public data source, carrying out initial cell screening after standardization processing, removing low-quality data and ensuring the reliability of the data; a batch effect removing algorithm is adopted, technical deviation between batches is eliminated, and data consistency is ensured; the cells are grouped, the cells are divided into different subgroups through dimensionality reduction and clustering analysis, and the heterogeneity and diversity of the cells in the tumor microenvironment are revealed; inferring copy number variation, and inferring genome instability and possible canceration characteristics of the cells by analyzing the copy number variation in the single cell data; performing quasi-time sequence analysis, reconstructing a cell development or transformation process by utilizing the quasi-time sequence analysis, and exploring the change of cell fate in the brain metastasis process of the lung cancer; function enrichment analysis: carrying out function annotation on different cell subgroups, and identifying key pathways and molecular markers related to brain metastasis of lung cancer; and drug sensitivity analysis before and after radiotherapy: analyzing the sensitivity difference of different cell subgroups to drugs before and after radiotherapy, and providing a basis for an individualized treatment scheme.

Description

Technical Field

[0001] This invention relates to the fields of metabolomics and gene analysis technology, specifically to a method and system for identifying driver genes of lung cancer brain metastasis through multi-omics analysis. Background Technology

[0002] 47% of lung cancer patients have brain metastases at diagnosis. Brain metastases include parenchymal metastases and meningeal metastases. The most common site of parenchymal metastases is the cerebral hemispheres, followed by the cerebellum and brainstem. Meningeal metastases are less common than parenchymal metastases, but have a worse prognosis.

[0003] Patients often receive systematic anti-tumor treatment after diagnosis. Brain metastases occur in 20%–65% of lung cancer patients during treatment, making it the most common type of brain metastatic tumor. Common symptoms of brain metastases include headache, vomiting, double vision, blurred vision, and decreased visual acuity. However, some non-small cell lung cancer patients have occult brain metastases, which are difficult to detect due to the absence of obvious symptoms. Furthermore, patients with brain metastases have a poor prognosis, with a median survival of less than one year. Therefore, elucidating the mechanisms of lung cancer brain metastases and developing new treatment strategies targeting key targets is of great significance for prolonging the survival time of patients with brain metastases. Summary of the Invention

[0004] This invention provides a method and system for identifying driver genes of lung cancer brain metastasis through multi-omics analysis. Its purpose is to accurately analyze the characteristics of various cell subpopulations in lung cancer brain metastasis organoids and their roles in the tumor microenvironment, providing strong support for the study of the mechanism of lung cancer brain metastasis and precision medicine.

[0005] To achieve the above objectives, the present invention provides a method for identifying driver genes of lung cancer brain metastasis through multi-omics analysis, comprising:

[0006] Data quality control involves obtaining single-cell RNA sequencing data from public data sources, performing initial cell screening after standardization, removing low-quality data, and ensuring data reliability.

[0007] Batch effect removal: A batch effect removal algorithm is used to eliminate technical deviations between batches and ensure data consistency.

[0008] Cell clustering, through dimensionality reduction and cluster analysis, divides cells into different subpopulations, revealing their heterogeneity and diversity in the tumor microenvironment;

[0009] Copy number variation inference: By analyzing copy number variations in single-cell data, we can infer genomic instability and potential carcinogenic features of cells.

[0010] Pseudo-temporal analysis: Using pseudo-temporal analysis, we can reconstruct the process of cell development or transformation and explore changes in cell fate during lung cancer brain metastasis.

[0011] Functional enrichment analysis, through functional annotation of different cell subpopulations, identifies key pathways and molecular markers associated with lung cancer brain metastasis;

[0012] Drug sensitivity analysis before and after radiotherapy: This study analyzes the differences in drug sensitivity among different cell subpopulations before and after radiotherapy, providing a basis for individualized treatment plans.

[0013] Furthermore, in the data quality control step, single-cell RNA sequencing data is obtained based on the public data port, and mitochondrial, ribosome, and erythrocyte genes are removed.

[0014] Furthermore, the cell clustering step includes:

[0015] Batch response was removed using the Harmony package, followed by dimensionality reduction and clustering to obtain 32 cell subpopulations;

[0016] Then, using classic cell markers, the cell subpopulations were annotated, resulting in 11 cell subpopulations.

[0017] After extracting brain-derived cells, they were re-clustered with reduced dimensionality to obtain nine major cell subpopulations, which were then used to construct a brain immune cell atlas.

[0018] Furthermore, the nine cell subpopulations include: microglia, oligodendrocytes, endothelial cells, fibroblasts, tumor cells, T cells, B cells, NK cells, and neutrophils.

[0019] Furthermore, the step of inferring copy number variation also includes:

[0020] Furthermore, step S300 also includes:

[0021] After re-dimension reduction and clustering of epithelial cells, 12 subgroups were obtained. Then, the CNV level in all epithelial cells was assessed using the inferCNV package.

[0022] Malignant epithelial cells with significantly higher CNV levels than normal epithelial cells were extracted from the CNV level distribution map, and each reclassified epithelial cell subpopulation was evaluated using CNV scoring.

[0023] Based on the assessment results, the cells were reclassified into four subpopulations: low copy number tumor cells, stem cell-like tumor cells, stromal-like tumor cells, and neutrophil-like tumor cells, and the proportion of each subpopulation in different samples was calculated.

[0024] Furthermore, in the pseudo-time series analysis step, the monocle3 package was used to infer the differentiation trajectory of cells. It was found that stem cell-like tumor cell populations differentiated into neutrophil-like and stromal-like tumor cell populations through multiple pathways, while low-copy-number tumor cell populations exhibited lower differentiation characteristics. The cytoTRACE2 package was used to further infer the differentiation potential of the four subpopulations.

[0025] Furthermore, in the functional enrichment analysis step, the metabolic characteristics of each cell subpopulation were assessed using the scMetabolism package and visualized using dot plots.

[0026] Furthermore, the drug sensitivity analysis steps before and after radiotherapy also include:

[0027] The pseudobulk package was used to compare differentially expressed genes in tumor cells before and after radiotherapy.

[0028] Furthermore, the prognostic impact of these genes on adenocarcinoma patients in the TCGA database was analyzed. KEGG analysis showed that related pathways included cytokine-receptor interactions, lipid metabolism, and atherosclerosis, while GSEA analysis showed statistically significant differences in drug metabolism pathways such as cytochrome P450 and phenylalanine metabolism.

[0029] Furthermore, the Pseudobulk algorithm was used to compare the differences in transcriptional levels between brain metastatic tumor cells and primary lung tumor cells. By comparing brain metastatic and primary lung lesion samples with ordinary transcriptomes, significantly upregulated genes were obtained. The intersection of the upregulated genes was taken to obtain a key gene S100B.

[0030] On the other hand, the present invention also provides a risk identification system for adverse events in hospitals based on causal reasoning, comprising:

[0031] The data quality control module obtains single-cell RNA sequencing data from public data sources, performs initial cell screening after standardization processing, removes low-quality data, and ensures data reliability.

[0032] The batch effect removal module uses a batch effect removal algorithm to eliminate technical deviations between batches and ensure data consistency.

[0033] The cell clustering module, through dimensionality reduction and cluster analysis, divides cells into different subpopulations, revealing their heterogeneity and diversity in the tumor microenvironment;

[0034] The copy number variation inference module infers genomic instability and potential carcinogenic features of cells by analyzing copy number variations in single-cell data.

[0035] The pseudo-temporal analysis module uses pseudo-temporal analysis to reconstruct cell development or transformation processes and explore changes in cell fate during lung cancer brain metastasis.

[0036] The functional enrichment analysis module identifies key pathways and molecular markers associated with lung cancer brain metastases by performing functional annotation on different cell subpopulations.

[0037] The pre- and post-radiotherapy drug sensitivity analysis module analyzes the differences in drug sensitivity among different cell subpopulations before and after radiotherapy, providing a basis for individualized treatment plans.

[0038] This invention provides a method and system for identifying driver genes in lung cancer brain metastasis using multi-omics analysis, comprising: data quality control, acquiring single-cell RNA sequencing data from public data sources, performing initial cell screening after standardization processing to remove low-quality data and ensure data reliability; batch effect removal, employing batch effect removal algorithms to eliminate technical biases between batches and ensure data consistency; cell population segmentation, using dimensionality reduction and cluster analysis to divide cells into different subpopulations, revealing their heterogeneity and diversity in the tumor microenvironment; copy number variation inference, analyzing copy number variations in single-cell data to infer genomic instability and possible carcinogenic features of cells; pseudo-temporal analysis, using pseudo-temporal analysis to reconstruct cell development or transformation processes and explore changes in cell fate during lung cancer brain metastasis; functional enrichment analysis, identifying key pathways and molecular markers related to lung cancer brain metastasis through functional annotation of different cell subpopulations; and pre- and post-radiotherapy drug sensitivity analysis, analyzing the differences in drug sensitivity of different cell subpopulations before and after radiotherapy to provide a basis for individualized treatment plans. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1 This is a flowchart of a method for identifying driver genes of lung cancer brain metastasis using multi-omics analysis, according to an embodiment of the present invention.

[0041] Figure 2 This is a system architecture diagram of a multi-omics analysis system for identifying driver genes of lung cancer brain metastases, according to an embodiment of the present invention.

[0042] Figure 3 This is a schematic diagram illustrating the grouping and interaction of cell subpopulations in brain metastases and primary tumor tissues according to an embodiment of the present invention.

[0043] Figure 4 This is a diagram illustrating the copy number variation of cell subpopulations in an embodiment of the present invention.

[0044] Figure 5 This is a diagram showing the analysis results of four cell subpopulations in an embodiment of the present invention.

[0045] Figure 6 This is a graph showing the analysis results of the cell proportion in an embodiment of the present invention.

[0046] Figure 7 This is a diagram showing the results of cell pseudo-time series analysis in an embodiment of the present invention.

[0047] Figure 8 This is a volcano diagram of differentially expressed genes before and after radiotherapy, according to an embodiment of the present invention.

[0048] Figure 9 This is a box plot of drug sensitivity analysis according to an embodiment of the present invention.

[0049] Figure 10 This is a schematic diagram of multi-cohort data differential analysis for screening upregulated target genes according to an embodiment of the present invention. Detailed Implementation

[0050] 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, and 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.

[0051] In the following description, references are made to “some embodiments,” which describe a subset of all possible embodiments. However, it is understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0052] If the application documents contain similar descriptions such as "first / second", the following explanation shall be added: In the following description, the terms "first / second / third" are used only to distinguish similar objects and do not represent a specific order of objects. It is understood that "first / second / third" may be interchanged in a specific order or sequence where permitted, so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein.

[0053] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0054] Example 1

[0055] Figure 1 This is a flowchart illustrating a method for identifying driver genes of lung cancer brain metastases using multi-omics analysis, according to an embodiment of the present invention. Figure 1 As shown in this embodiment, a method for identifying driver genes of lung cancer brain metastasis through multi-omics analysis includes:

[0056] S101, according to quality control, single-cell RNA sequencing data are obtained from public data sources, and initial cell screening is performed after standardization to remove low-quality data and ensure data reliability;

[0057] S102, De-batch effect: This algorithm eliminates technical deviations between batches, ensuring data consistency.

[0058] Specifically, in this embodiment, single-cell transcriptome data is read and examined in the R environment. The doubletFinder package is used to remove double cells, and samples with a mitochondrial ratio greater than 25% and a red blood cell ratio greater than 5% are screened out.

[0059] S103, cell clustering, uses dimensionality reduction and cluster analysis to divide cells into different subpopulations, revealing their heterogeneity and diversity in the tumor microenvironment.

[0060] Specifically, such as Figure 3 As shown, this embodiment uses the Harmony package to remove batch effects, followed by dimensionality reduction clustering to obtain 32 cell subpopulations. Annotation of these subpopulations using classic cell markers yielded 11 subpopulations. After extracting brain-derived cells, dimensionality reduction clustering was performed again, resulting in 9 major cell subpopulations, including microglia, oligodendrocytes, endothelial cells, fibroblasts, tumor cells, T cells, B cells, NK cells, and neutrophils, constructing a brain immune cell atlas. Intercellular communication analysis revealed the interactions between cell subpopulations within brain metastases, and it was found that tumor cells and microglia exhibited the strongest communication quantity and intensity.

[0061] S104, inferring copy number variations, inferring genomic instability and potential carcinogenic features of cells by analyzing copy number variations in single-cell data.

[0062] Specifically, such as Figure 4 As shown, after tumor cell identification, the cells were reclassified into four subpopulations based on markers of different cell subpopulations: low copy number tumor cells, stem cell-like tumor cells, stromal-like tumor cells, and neutrophil-like tumor cells, and the proportion of each subpopulation in different samples was calculated.

[0063] S105, pseudo-temporal analysis, uses pseudo-temporal analysis to reconstruct cell development or transformation processes and explore changes in cell fate during lung cancer brain metastasis.

[0064] Specifically, such as Figure 5 and Figure 6 As shown, the monocle3 package was used to infer the differentiation trajectory of cells, revealing that stem cell-like tumor cell populations differentiate into neutrophil-like and stromal-like tumor cell populations through multiple pathways, while low copy number tumor cell populations exhibited lower differentiation characteristics. The cytoTRACE2 package was used to further infer the differentiation potential of four subpopulations.

[0065] S106, functional enrichment analysis, identifies key pathways and molecular markers associated with lung cancer brain metastases by functionally annotating different cell subpopulations.

[0066] Specifically, such as Figure 7 As shown, the metabolic characteristics of each cell subpopulation were assessed using the scMetabolism package and visualized using dot plots. The results showed that low copy number tumor cells exhibited active multiple metabolic pathways, such as the interconversion of pentoses and glucuronides, oxidative phosphorylation, and glycolysis; neutrophil-like tumor cells primarily focused on the interconversion of pentoses and glucuronides; stem cell-like tumor cells showed moderate metabolic activity in multiple pathways; and stromal-like tumor cells exhibited high signal intensity in pathways such as oxidative phosphorylation and glycolysis.

[0067] In another embodiment, such as Figure 10 As shown, the Pseudobulk algorithm was used to compare the transcriptional levels of 23,530 brain metastasis tumor cells from 28 brain metastasis samples and 7,919 primary lung tumor cells from 16 primary lung lesion samples, resulting in 1,790 significantly upregulated genes. A comparison of the transcriptomes of 28 brain metastasis and primary lung lesion samples (22 of which were paired samples) yielded 100 significantly upregulated genes. Spatial transcriptome data from 44 lung cancer brain metastasis patients (55 brain metastasis samples and 45 primary lung tumor samples) were collected and analyzed, resulting in 25 differentially upregulated genes. The intersection of these upregulated genes yielded a key gene, S100B.

[0068] S107, Drug Sensitivity Analysis Before and After Radiotherapy: This study analyzes the differences in drug sensitivity among different cell subpopulations before and after radiotherapy, providing a basis for individualized treatment plans.

[0069] Specifically, such as Figure 8As shown, the pseudobulk package was used to compare differentially expressed genes in tumor cells before and after radiotherapy, identifying 12 significantly upregulated genes. The expression changes of these genes were displayed using a volcano plot. Further analysis of the prognostic impact of these genes on adenocarcinoma patients in the TCGA database revealed that high expression of KRT17 and RAET1L was associated with a poorer prognosis in lung adenocarcinoma patients. Functional enrichment analysis of the differentially upregulated genes mainly involved biological processes such as humoral immunity and lymphocyte chemotaxis. KEGG analysis showed that related pathways included cytokine-receptor interactions, lipid metabolism, and atherosclerosis, while GSEA analysis showed statistically significant differences in drug metabolism pathways such as cytochrome P450 and phenylalanine metabolism. Figure 9 As shown, the oncoPredic package was used to predict the sensitivity of tumor cells to different drugs before and after radiotherapy. Significant differences were found in the IC50 values ​​of 12 drugs, including the common lung cancer chemotherapy drug docetaxel. The differences in drug sensitivity of samples before and after radiotherapy were shown by box plots.

[0070] Example 2

[0071] Figure 2 This invention provides a risk identification system for hospital adverse events based on causal reasoning. The invention also provides a risk identification system for hospital adverse events based on causal reasoning, comprising:

[0072] Data quality control module 1 obtains single-cell RNA sequencing data from public data sources, performs initial cell screening after standardization processing, removes low-quality data, and ensures data reliability;

[0073] Batch Effect Removal Module 2 employs a batch effect removal algorithm to eliminate technical deviations between batches and ensure data consistency.

[0074] Cell clustering module 3 uses dimensionality reduction and cluster analysis to divide cells into different subpopulations, revealing their heterogeneity and diversity in the tumor microenvironment;

[0075] The copy number variation inference module 4 infers genomic instability and possible carcinogenic features of cells by analyzing copy number variations in single-cell data.

[0076] Module 5, pseudo-temporal analysis, uses pseudo-temporal analysis to reconstruct cell development or transformation processes and explore changes in cell fate during lung cancer brain metastasis.

[0077] Functional enrichment analysis module 6 identifies key pathways and molecular markers associated with lung cancer brain metastasis by performing functional annotation on different cell subpopulations.

[0078] Module 7, which analyzes the differences in drug sensitivity among different cell subpopulations before and after radiotherapy, provides a basis for individualized treatment plans.

[0079] Although embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for identifying driver genes of lung cancer brain metastasis using multi-omics analysis, characterized in that, include: Data quality control involves obtaining single-cell RNA sequencing data from public data sources, performing initial cell screening after standardization, removing low-quality data, and ensuring data reliability. Batch effect removal: A batch effect removal algorithm is used to eliminate technical deviations between batches and ensure data consistency. Cell clustering, through dimensionality reduction and cluster analysis, divides cells into different subpopulations, revealing their heterogeneity and diversity in the tumor microenvironment; Copy number variation inference: By analyzing copy number variations in single-cell data, we can infer genomic instability and potential carcinogenic features of cells. Pseudo-temporal analysis: Using pseudo-temporal analysis, we can reconstruct the process of cell development or transformation and explore changes in cell fate during lung cancer brain metastasis. Functional enrichment analysis, through functional annotation of different cell subpopulations, identifies key pathways and molecular markers associated with lung cancer brain metastasis; Drug sensitivity analysis before and after radiotherapy: This study analyzes the differences in drug sensitivity among different cell subpopulations before and after radiotherapy, providing a basis for individualized treatment plans.

2. The method for identifying driver genes of lung cancer brain metastasis using multi-omics analysis as described in claim 1, characterized in that, In the data quality control step, single-cell RNA sequencing data is obtained based on the public data port, and mitochondrial, ribosome and erythrocyte genes are removed.

3. The method for identifying driver genes of lung cancer brain metastasis using multi-omics analysis as described in claim 2, characterized in that, The cell clustering process includes: Batch response was removed using the Harmony package, followed by dimensionality reduction and clustering to obtain 32 cell subpopulations; Then, using classic cell markers, the cell subpopulations were annotated, resulting in 11 cell subpopulations. After extracting brain-derived cells, they were re-clustered with reduced dimensionality to obtain nine major cell subpopulations, which were then used to construct a brain immune cell atlas.

4. The method for identifying driver genes of lung cancer brain metastasis using multi-omics analysis as described in claim 3, characterized in that, The nine cell subpopulations include: microglia, oligodendrocytes, endothelial cells, fibroblasts, tumor cells, T cells, B cells, NK cells, and neutrophils.

5. The method for identifying driver genes of lung cancer brain metastasis using multi-omics analysis as described in claim 4, characterized in that, Also includes: After re-dimension reduction and clustering of epithelial cells, 12 subgroups were obtained. Then, the CNV level in all epithelial cells was assessed using the inferCNV package. Malignant epithelial cells with significantly higher CNV levels than normal epithelial cells were extracted from the CNV level distribution map, and each reclassified epithelial cell subpopulation was evaluated using CNV scoring. Based on the assessment results, the cells were reclassified into four subpopulations: low copy number tumor cells, stem cell-like tumor cells, stromal-like tumor cells, and neutrophil-like tumor cells, and the proportion of each subpopulation in different samples was calculated.

6. The method for identifying driver genes of lung cancer brain metastasis using multi-omics analysis as described in claim 5, characterized in that, In the pseudo-time series analysis step, the monocle3 package was used to infer the differentiation trajectory of cells. It was found that stem cell-like tumor cell populations differentiated into neutrophil-like and stromal-like tumor cell populations through multiple pathways, while low-copy-number tumor cell populations exhibited lower differentiation characteristics. The cytoTRACE2 package was used to further infer the differentiation potential of four subpopulations.

7. The risk identification method for adverse events in hospitals based on causal reasoning as described in claim 6, characterized in that, In the functional enrichment analysis step, the metabolic characteristics of each cell subpopulation were assessed using the scMetabolism package and visualized using dot plots.

8. The risk identification method for adverse events in hospitals based on causal reasoning as described in claim 7, characterized in that, The pre- and post-radiotherapy drug sensitivity analysis steps also include: The pseudobulk package was used to compare differentially expressed genes in tumor cells before and after radiotherapy. Further analysis of the prognostic impact of these genes on adenocarcinoma patients in the TCGA database revealed that KEGG analysis showed related pathways including cytokine-receptor interactions, lipid metabolism, and atherosclerosis, while GSEA analysis showed statistically significant differences in drug metabolism pathways such as cytochrome P450 and phenylalanine metabolism.

9. The risk identification method for adverse events in hospitals based on causal reasoning as described in claim 7, characterized in that, It also includes: then using the Pseudobulk algorithm to compare the differences in transcriptional levels between brain metastatic tumor cells and primary lung tumor cells, comparing brain metastasis and primary lung lesion samples through ordinary transcriptome analysis to obtain significantly upregulated genes, and taking the intersection of upregulated genes to obtain a key gene S100B.

10. A risk identification system for adverse events in hospitals based on causal reasoning, characterized in that, include: The data quality control module obtains single-cell RNA sequencing data from public data sources, performs initial cell screening after standardization processing, removes low-quality data, and ensures data reliability. The batch effect removal module uses a batch effect removal algorithm to eliminate technical deviations between batches and ensure data consistency. The cell clustering module, through dimensionality reduction and cluster analysis, divides cells into different subpopulations, revealing their heterogeneity and diversity in the tumor microenvironment; The copy number variation inference module analyzes copy number variations in single-cell data to infer genomic instability and potential carcinogenic features of cells. The pseudo-temporal analysis module uses pseudo-temporal analysis to reconstruct cell development or transformation processes and explore changes in cell fate during lung cancer brain metastasis. The functional enrichment analysis module identifies key pathways and molecular markers associated with lung cancer brain metastases by performing functional annotation on different cell subpopulations. The pre- and post-radiotherapy drug sensitivity analysis module analyzes the differences in drug sensitivity among different cell subpopulations before and after radiotherapy, providing a basis for individualized treatment plans.

Images