Method and system for identification of rare macrophage subsets and disease markers in idiopathic pulmonary fibrosis

Abstract

The present invention provides a method and system for identifying rare macrophage subpopulations and disease markers in idiopathic pulmonary fibrosis, by obtaining a public data set of human lung tissue, preprocessing the data set, fusing the preprocessed data set and removing batch effects, clustering the fused data set and marking the cell type of each group of cells and screening macrophages, identifying rare macrophage subtypes and differentially expressed genes in the macrophages, the method and system for identifying rare macrophage subpopulations and disease markers in idiopathic pulmonary fibrosis provided by the present application, avoid the existing single-cell data analysis task of only identifying cell subtypes from tissue samples, and further analyze specific cell types on this basis to identify rare macrophage subpopulations, and further identify IPF disease markers, providing a reference for disease diagnosis.

Description

Technical Field

[0001] The present invention relates to the field of bioinformatics, and in particular to a method and system for identifying rare macrophage subpopulations and disease markers in idiopathic pulmonary fibrosis. Background Art

[0002] Idiopathic pulmonary fibrosis (IPF) is a chronic, progressive, irreversible interstitial lung disease with lung scarring. The cause of the disease is still unclear, there is no effective treatment, and it is accompanied by high morbidity and mortality. Previous pathological analyses of IPF were based on traditional tissue sequencing data, but it can only provide the average expression signal of all cells in the tissue and cannot finely analyze the expression differences between different cells in IPF lung tissue. The emergence of single-cell technology provides a higher resolution for the study of IPF. Data analysis of single-cell sequencing data can more deeply explore the expression of different cell types.

[0003] Mapping genotype to phenotype is one of the long-standing challenges in biology and medicine. An effective strategy to address this problem is to perform transcriptome analysis. However, although all cells in our body have almost the same gene types, the transcriptome information of any one cell only reflects the activity of a portion of genes. In addition, because many different types of cells in our body express a unique transcriptome, traditional tissue sequencing can only provide an average expression signal for all cells in the tissue. And there is increasing evidence that even in the same cell type, gene expression is different. Based on this, a more precise understanding of the transcriptome in individual cells is crucial to elucidating their role in cell function and understanding how gene expression promotes beneficial or harmful states. With the introduction of single-cell technology in 2009, this problem has been effectively solved. Single-cell technology has brought the resolution of tissue and disease research to a higher level by extracting, sequencing and analyzing individual cells in tissues, providing a new dimension for the study of tissues and diseases.

[0004] Single-cell technology is currently widely used in IPF research, and the main research goals are to identify new cell subtypes and study the mechanism of drug treatment of diseases. For example, in the study of Reyfman et al., rare cell populations including respiratory stem cells and senescent cells were found to appear in the process of fibrosis, and two unique macrophage subpopulations were identified in IPF samples. In the study of Xu et al., atypical alveolar epithelial cell subpopulations expressing respiratory-related genes were found in the explant tissues of patients undergoing end-stage IPF transplantation. The study of Adams et al. found changes in endothelial cells in IPF lungs, and they found a group of vascular endothelial cells expressing COL15A1 in IPF lung tissue. In the study of Kwapiszewska et al., single-cell technology was used to analyze lung tissue homogenates and lung tissue fibroblasts with and without pirfenidone, and it was found that pirfenidone exerted beneficial effects on multiple pathways of fibroblasts and other lung tissue cells. In their study, Sheu et al. analyzed IPF fibroblasts from patients taking nintedanib and found more than 100 abnormally upregulated genes, which were mainly involved in cell cycle pathways and inhibition of fibroblast proliferation.

[0005] Many current single-cell studies are aimed at identifying unique cell subtypes within a tissue, but rarely focus on in-depth studies of a specific cell type. Summary of the invention

[0006] The embodiments of the present invention provide a method and device for identifying rare macrophage subpopulations and disease markers in idiopathic pulmonary fibrosis, so as to at least solve the defect that the existing single-cell data analysis task only stays at identifying cell subtypes from tissue samples, but rarely conducts in-depth research on a specific cell type.

[0007] According to one embodiment of the present invention, a method for identifying rare macrophage subpopulations and disease markers in idiopathic pulmonary fibrosis comprises the following steps:

[0008] Step S10: obtaining a public dataset of human lung tissue and preprocessing the dataset;

[0009] Step S20: performing data fusion on the preprocessed data set and removing batch effects;

[0010] Step S30: clustering the fused data set, marking the cell type of each group of cells, and screening macrophages;

[0011] Step S40: Identifying rare macrophage subtypes and differentially expressed genes in the macrophages.

[0012] In some embodiments, in step S10, the step of obtaining a public dataset of human lung tissue and preprocessing the dataset specifically includes the following steps:

[0013] Step S11: selecting 4 data sets from the GEO database, namely GSE136831, GSE135893, GSE128033 and GSE122960, all of which extracted human lung tissue and separated single cells for sequencing;

[0014] Step S12: performing data cleaning on each of the above data sets, wherein the data cleaning includes removing genes that are not expressed in any cells, removing cells that express less than 200 genes, and removing cells that express more than 25% of mitochondrial-related genes;

[0015] Step S13: performing data preprocessing on the cleaned data set, the preprocessing including LogNormalization data normalization, data scaling, removing cell cycle signals, and selecting 8000 highly variable genes for each data set for subsequent data integration.

[0016] In some embodiments, in step S20, the step of performing data fusion on the preprocessed data set and removing batch effects specifically includes the following steps:

[0017] Step S21: finding projections of the plurality of data sets by a canonical correlation analysis algorithm so as to maximize the correlation between all data sets;

[0018] Step S22: using a dynamic time planning adjustment algorithm to determine the best mapping between the data sets.

[0019] In some embodiments, in step S21, the step of finding projections of the plurality of data sets by a canonical correlation analysis algorithm so as to maximize the correlation between all data sets specifically includes the following steps:

[0020] Step S211: performing singular value decomposition on the highly variable gene expression matrix of each data set to obtain an initial canonical correlation vector CCV;

[0021] Step S212: Then, the typical correlation vector W corresponding to each data set is updated by performing singular value decomposition on each data set and other data sets until the difference ratio of W before and after the update is greater than a given threshold.

[0022] In some embodiments, in step S22, the step of using a dynamic time planning adjustment algorithm to determine the optimal mapping between the data sets specifically includes the following steps:

[0023] In a single-cell dataset, the dynamic time planning adjustment algorithm is used to calculate a warp path between the datasets to minimize the distance between the datasets, wherein W(w1, w2…wk) is a warp matrix, each vector of which corresponds to a point in the warp path, which maps an element in dataset X to dataset Y and minimizes the distance between them; the typical correlation vector is mapped to the warp matrix to align the two datasets into a low-dimensional space;

[0024] For more than two datasets, the dataset with the largest number of cells is used as the reference dataset; the other datasets are aligned with the reference dataset respectively, and finally the typical correlation vector of each dataset is normalized to the common calibration space defined by the reference dataset.

[0025] In some embodiments, in step S30, the step of clustering the fused data set, marking the cell type of each group of cells, and screening macrophages specifically includes the following steps: clustering the fused data set, marking the cell type of each group of cells, and screening macrophages using a clustering analysis algorithm based on community detection, the steps are as follows:

[0026] Step S31: using the data after dimensionality reduction as input data and cells as nodes, the Euclidean distance between each node is calculated by the KNNG algorithm to determine the K-nearest neighbors of each node;

[0027] Step S32: Use the following formula to calculate the neighborhood overlap between each node and its K nearest neighbor nodes to construct a shared nearest neighbor matrix, where A and B represent the number of "neighbor" nodes of the two nodes): J(A,B)=(|A∩B|) / (|A∪B|)=(|A∩B|) / (|A|+|B|-|A∩B|);

[0028] Step S33: The Louvain algorithm is used to scan nodes and their neighboring nodes, calculate the modularity of each node and measure the module benefit after adding the module, select the neighboring node with the maximum benefit to add to the module, repeat the iteration, and finally form a node cluster;

[0029] Step S34: Wilcoxon rank sum test was used to perform differential analysis, and the cell type was determined and macrophages were screened based on the genes differentially upregulated in each group of cells.

[0030] In some embodiments, in step S33, Louvain algorithm is used to scan nodes and their neighboring nodes, calculate the modularity of each node and measure the module benefit after adding the module, select the neighboring node with the maximum benefit to add the module, repeat the iteration, and finally form a node cluster, which specifically includes the following steps:

[0031] The Louvain algorithm is used to scan the node and its neighboring nodes, and the modularity of each node is calculated using the following formula, where m is the sum of the weights of the edges in the graph, i and j represent two nodes, Aij represents the weight between the two nodes, which is calculated by SNN, ki and kj represent the sum of the weights of all edges of nodes i and j, ci and cj represent the group to which nodes i and j belong, and δ is the Kronecker delta function):

[0032]

[0033] In order to maximize the Q value, the Louvain algorithm calculates the modularity gain and iterates repeatedly, as shown in the following formula, where Σin is the sum of the internal node weights of the community that node i enters, Σtot is the sum of the weights of the edges connecting node i and all communities in the graph, and ki,in is the sum of the weights of the edges between node i and all nodes in the community that i wants to move into:

[0034]

[0035] In some embodiments, in step S40, the step of identifying rare macrophage subtypes and differentially expressed genes in the macrophages specifically includes the following steps:

[0036] Three rare macrophage subsets composed of IPF samples were identified and recorded as Cluster 0, Cluster 1, and Cluster 2. The three groups of cells all expressed IPF markers, namely SPP1, CCL2, FABP4, and CHIT1, among which: SPP1 is a gene encoding osteopontin, which can significantly promote the migration and proliferation of fibroblasts and epithelial cells, and can be used as a disease marker for IPF; CCL2 is a chemokine 2 that can recruit mononuclear macrophages and promote fibrosis through various mechanisms involving inflammation, angiogenesis, and myofibroblast accumulation, and can be used as an IPF disease marker; FABP4 encodes fatty acid binding protein, which is a cytoplasmic fatty acid chaperone protein expressed in adipocytes and myeloid cells, and can promote ATP polarization of macrophages to M1 type, and participates in the occurrence of IPF by promoting the activation of M1 macrophages; CHIT1 encodes chitintriosidase, which has a pro-fibrotic effect and is significantly expressed in the lungs of IPF patients.

[0037] According to another embodiment of the present invention, there is provided a system for identifying rare macrophage subsets and disease markers in idiopathic pulmonary fibrosis, comprising:

[0038] A data set acquisition unit: acquiring a public data set of human lung tissue and preprocessing the data set;

[0039] Preprocessing unit: performing data fusion on the preprocessed data set and removing batch effects;

[0040] Screening unit: clustering the fused data set, marking the cell type of each group of cells, and screening macrophages;

[0041] Identification unit: identifying rare macrophage subtypes and their differentially expressed genes in the macrophages.

[0042] A storage medium storing a program file capable of implementing any of the above-mentioned methods for identifying rare macrophage subpopulations and disease markers in idiopathic pulmonary fibrosis.

[0043] A processor is used to run a program, wherein when the program is run, any one of the methods for identifying rare macrophage subpopulations and disease markers in idiopathic pulmonary fibrosis is executed.

[0044] The method and system for identifying rare macrophage subpopulations and disease markers in idiopathic pulmonary fibrosis in the embodiments of the present invention obtain a public data set of human lung tissue, preprocess the data set, fuse the preprocessed data set and remove batch effects, cluster the fused data set and mark the cell type of each group of cells and screen macrophages, identify rare macrophage subtypes in the macrophages and their differentially expressed genes. The method and system for identifying rare macrophage subpopulations and disease markers in idiopathic pulmonary fibrosis provided in the present application avoid the existing single-cell data analysis task of only identifying cell subtypes from tissue samples. On this basis, further analysis of specific cell types can identify rare macrophage subpopulations, and further identify IPF disease markers, providing a reference for disease diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0046] Figure 1 A flowchart of the steps of the method for identifying rare macrophage subsets and disease markers in idiopathic pulmonary fibrosis provided in Example 1 of the present invention;

[0047] Figure 2 A flowchart of the steps of obtaining a public dataset of human lung tissue and preprocessing the dataset provided in Example 1 of the present invention;

[0048] Figure 3A flowchart of the steps of performing data fusion and removing batch effects on the preprocessed data set provided in Example 1 of the present invention;

[0049] Figure 4 A flow chart of the steps of clustering the fused data set, marking the cell type of each group of cells, and screening macrophages provided in Example 1 of the present invention.

[0050] Figure 5 This is the Heatmap of differentially upregulated gene expression of Cluster 0, 1, and 2 provided in Example 1 of the present invention.

[0051] Figure 6 This is the expression distribution diagram of the key markers SPP1, CCL2, FABP4, and CHIT1 in three groups of macrophages provided in Example 1 of the present invention.

[0052] Figure 7 This is a schematic diagram of the structure of the identification system for rare macrophage subsets and disease markers in idiopathic pulmonary fibrosis provided in Example 2 of the present invention. DETAILED DESCRIPTION

[0053] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0054] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0055] Example 1

[0056] See also Figure 1According to one embodiment of the present invention, a method for identifying rare macrophage subsets and disease markers in idiopathic pulmonary fibrosis is provided, comprising the following steps S10 to S40, and the implementation scheme of each step is described in detail below.

[0057] Step S10: Acquire a public dataset of human lung tissue and preprocess the dataset.

[0058] See also Figure 2 , which is a flowchart of the steps of obtaining a public dataset of human lung tissue and preprocessing the dataset provided in this embodiment, including the following steps S11 to S13, and the implementation scheme of each step is described in detail below.

[0059] Step S11: 4 data sets were selected from the GEO database, namely GSE136831, GSE135893, GSE128033 and GSE122960. The 4 data sets all extracted human lung tissues and isolated single cells for sequencing.

[0060] Please refer to Table 1 below, which provides 4 data sets for this embodiment.

[0061] Table 1 shows the four data sets provided in the embodiment.

[0062]

[0063] Step S12: performing data cleaning on each of the above data sets, wherein the data cleaning includes removing genes that are not expressed in any cell, removing cells that express less than 200 genes, and removing cells that express mitochondrial-related genes at a proportion greater than 25%.

[0064] Step S13: performing data preprocessing on the cleaned data set, the preprocessing including LogNormalization data normalization, data scaling, removing cell cycle signals, and selecting 8000 highly variable genes for each data set for subsequent data integration.

[0065] Step S20: performing data fusion on the preprocessed data set and removing batch effects.

[0066] It is understandable that after summarizing the highly variable gene expression data of each dataset in the above steps, all datasets are integrated into one dataset for subsequent analysis. Due to the differences in abiotic factors of these datasets, such as experimental platforms, technologies, personnel, dates, reagents, and sample selection, batch effects may be introduced in the process of integrating these datasets.

[0067] See also Figure 3, is a flow chart of the steps of performing data fusion and removing batch effects on the preprocessed data set in this embodiment, specifically including the following steps S21 to S22. The implementation scheme of each step is described in detail below.

[0068] Step S21: finding projections of the plurality of data sets through a canonical correlation analysis algorithm so as to maximize the correlation between all data sets.

[0069] It can be understood that the core of the canonical correlation analysis algorithm (CCA) is to maximize the correlation between all data sets by finding the projections of multiple data sets. The specific formula is as follows, where n represents different data sets and w corresponds to the canonical correlation vector CCV in each data set. i , X i , w i , w j Represents two different data sets and their corresponding CCVs:

[0070]

[0071] The typical correlation vector W(w1,w2…wn) corresponding to each data set is solved by an iterative algorithm.

[0072] Specifically, the step of finding projections of the plurality of data sets through a canonical correlation analysis algorithm to maximize the correlation between all data sets includes the following steps S211 to S212, and the implementation scheme of each step is described in detail below.

[0073] Step S211: performing singular value decomposition on the highly variable gene expression matrix of each data set to obtain an initial canonical correlation vector CCV;

[0074] Step S212: Then, the typical correlation vector W corresponding to each data set is updated by performing singular value decomposition on each data set and other data sets until the difference ratio of W before and after the update is greater than a given threshold.

[0075] The specific formula is as follows, where n represents different data sets, w represents the typical correlation vector CCV, Γ and Δ represent the projection vectors of different data sets, Λ represents the eigenvector, and SVD is the singular value decomposition of the data set X, and the result is used to initialize w. When the difference ratio of w is less than the threshold 103, w is updated by performing singular value decomposition on the data set X and other data sets k until the difference ratio is greater than the threshold:

[0076] initialize n ←Δ n ,[Γ n ,Λ n ,Δ n ]←SVD(Xn )

[0077]

[0078] Step S22, using a dynamic time planning adjustment algorithm to determine the optimal mapping between the data sets, specifically includes the following steps:

[0079] In a single-cell dataset, the dynamic time planning adjustment algorithm is used to calculate a warp path between the datasets to minimize the distance between the datasets, wherein W(w1, w2…wk) is a warp matrix, each vector of which corresponds to a point in the warp path, which maps an element in dataset X to dataset Y and minimizes the distance between them; the typical correlation vector is mapped to the warp matrix to align the two datasets into a low-dimensional space;

[0080] For more than two datasets, the dataset with the largest number of cells is used as the reference dataset; the other datasets are aligned with the reference dataset respectively, and finally the typical correlation vector of each dataset is normalized to the common calibration space defined by the reference dataset, and the common calibration space is defined as follows, where W is the distortion matrix, w k Represents the distortion vector:

[0081] DTW(X,Y)=min W [∑ k (w k )].

[0082] Step S30, clustering the fused data set, marking the cell type of each group of cells, and screening macrophages.

[0083] See also Figure 4 , which is a flow chart of the steps of clustering the fused data set and marking the cell type of each group of cells and screening macrophages provided in this embodiment, specifically including the following steps S31 to S34, and the specific implementation scheme of each step is described in detail below.

[0084] Step S31: Using the data after dimensionality reduction as input data and cells as nodes, the KNNG algorithm is used to calculate the Euclidean distance between each node to determine the K-nearest neighbors of each node.

[0085] In this embodiment, a single-cell graph is constructed by the KNNG (K-Nearest Neighbor graph) algorithm, with the data after dimensionality reduction as input data and cells as nodes, and the Euclidean distance between each node is calculated to determine the K-nearest neighbors (KNN, K defaults to 20) of each node.

[0086] Step S32: Use the following formula to calculate the neighborhood overlap between each node and its K nearest neighbor nodes to construct a shared nearest neighbor matrix, where A and B represent the number of "neighbor" nodes of the two nodes): J(A,B)=(|A∩B|) / (|A∪B|)=(|A∩B|) / (|A|+|B|-|A∩B|).

[0087] In this embodiment, after the KNNG is constructed, the neighborhood overlap (Jaccard index) between each node and its K nearest neighbor nodes is calculated to construct a shared nearest neighbor (SNN) matrix. The Jaccard index calculates the ratio of the "neighbors" shared by two nodes to all "neighbors". The larger the ratio, the more similar the two nodes are.

[0088] Step S33: The Louvain algorithm is used to scan nodes and their neighboring nodes, calculate the modularity of each node and measure the module benefit after joining the module, select the neighboring node with the maximum benefit to join the module, repeat the iteration, and finally form a node cluster.

[0089] It can be understood that the number of clusters is determined by the Louvain algorithm as a modularity function. The Louvain algorithm is a method for extracting modules from a network and a greedy optimization method. The Louvain algorithm scans nodes and their neighboring nodes, calculates the modularity of each node and measures the module benefit after adding the module, selects the neighboring node with the maximum benefit to join the module, repeats the iteration, and finally forms a node cluster.

[0090] Specifically, the Louvain algorithm is used to scan the node and its neighboring nodes, and the modularity of each node is calculated using the following formula, where m is the sum of the weights of the edges in the graph, i and j represent two nodes, Aij represents the weight between the two nodes, which is calculated by SNN, ki and kj represent the sum of the weights of all edges of nodes i and j, ci and cj represent the group to which nodes i and j belong, and δ is the Kronecker delta function)

[0091]

[0092] In order to maximize the Q value, the Louvain algorithm calculates the modularity gain and iterates repeatedly, as shown in the following formula, where Σin is the sum of the internal node weights of the community that node i enters, Σtot is the sum of the weights of the edges connecting node i and all communities in the graph, and ki,in is the sum of the weights of the edges between node i and all nodes in the community that i wants to move into:

[0093]

[0094] Table 2 shows the total number of cells and the number of screened macrophages in the four data sets provided in this example

[0095]

[0096]

[0097] Step S40: Identifying rare macrophage subtypes and differentially expressed genes in the macrophages.

[0098] In this example, after screening macrophages, three rare macrophage subpopulations consisting of IPF samples were identified by secondary clustering and differential analysis of macrophages (the three groups of cells were recorded as Cluster 0, Cluster 1, and Cluster 2, and the differential gene expression Heatmap was as follows: Figure 5 These three groups of cells all expressed IPF markers, namely SPP1, CCL2, FABP4, and CHIT1 (the expression dot plots of the four markers are shown in Figure 6 As shown). SPP1 is a gene encoding osteopontin, which is highly associated with IPF and can be used as a disease marker for IPF. It can significantly promote the migration and proliferation of fibroblasts and epithelial cells. CCL2 is a chemokine 2 that can recruit mononuclear macrophages and basophils and can be used as a disease marker for IPF. FABP4 encodes fatty acid binding protein, a cytoplasmic fatty acid chaperone protein that is mainly expressed in adipocytes and myeloid cells. FABP4 has a potential role in the disease progression of IPF. The fatty acid oxidation (FAO) in which it participates produces a large amount of ATP that is believed to promote the polarization of macrophages to the M1 type. When activated, M1 macrophages produce pro-fibrotic mediators such as TGF-β1 (activating fibroblasts and ECM accumulation). CHIT1 is significantly expressed in the lungs of IPF patients and has been shown to have a pro-fibrotic effect and can become a potential new target for the treatment of IPF.

[0099] The method for identifying rare macrophage subpopulations and disease markers in idiopathic pulmonary fibrosis in an embodiment of the present invention obtains a public data set of human lung tissue, preprocesses the data set, fuses the preprocessed data set and removes batch effects, clusters the fused data set and marks the cell type of each group of cells and screens macrophages, identifies rare macrophage subtypes in the macrophages and their differentially expressed genes. The above identification method avoids the existing single-cell data analysis task of only identifying cell subtypes from tissue samples. On this basis, further analysis of specific cell types can identify rare macrophage subpopulations, and further identify IPF disease markers, providing a reference for disease diagnosis.

[0100] Example 2

[0101] See also Figure 7 According to another embodiment of the present invention, a schematic diagram of a system for identifying rare macrophage subsets and disease markers in idiopathic pulmonary fibrosis is provided, comprising:

[0102] The data set acquisition unit 110 is configured to acquire a public data set of human lung tissue and pre-process the data set;

[0103] Preprocessing unit 120: performing data fusion on the preprocessed data set and removing batch effects;

[0104] Screening unit 130: clustering the fused data set, marking the cell type of each group of cells, and screening macrophages;

[0105] Identification unit 140: identifying rare macrophage subtypes and differentially expressed genes in the macrophages.

[0106] The detailed implementation scheme of the identification system of rare macrophage subsets and disease markers in idiopathic pulmonary fibrosis provided in this embodiment has been described in detail in Example 1 and will not be repeated here.

[0107] The identification system of rare macrophage subpopulations and disease markers in idiopathic pulmonary fibrosis in the embodiment of the present invention obtains a public data set of human lung tissue, preprocesses the data set, fuses the preprocessed data set and removes batch effects, clusters the fused data set and marks the cell type of each group of cells and screens macrophages, identifies rare macrophage subtypes in the macrophages and their differentially expressed genes. The above identification system avoids the existing single-cell data analysis task of only identifying cell subtypes from tissue samples. On this basis, further analysis of specific cell types can identify rare macrophage subpopulations, and further identify IPF disease markers, providing a reference for disease diagnosis.

[0108] Example 3

[0109] A storage medium stores a program file capable of implementing any of the above-mentioned methods for identifying rare macrophage subpopulations and disease markers in idiopathic pulmonary fibrosis.

[0110] Example 4

[0111] A processor is used to run a program, wherein when the program is run, any one of the above-mentioned methods for identifying rare macrophage subpopulations and disease markers in idiopathic pulmonary fibrosis is executed.

[0112] The technical advantage of the embodiments of the present invention is at least that it avoids the existing single-cell data analysis task of only identifying cell subtypes from tissue samples. Further analysis of specific cell types on this basis can identify rare macrophage subpopulations and further identify IPF disease markers, providing a reference for disease diagnosis.

[0113] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0114] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0115] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the system embodiments described above are only schematic. For example, the division of units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0116] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed over multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0117] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0118] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0119] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for identifying rare macrophage subsets and disease markers in idiopathic pulmonary fibrosis, characterized in that: The following steps are involved: step S10: obtaining a public dataset of human lung tissue, and preprocessing the dataset; Step S20: performing data fusion on the preprocessed data set and removing batch effects; Step S30: clustering the fused data set, marking the cell type of each group of cells, and screening macrophages; Step S40: Identify rare macrophage subtypes and their differentially expressed genes in the macrophages; wherein: In step S10, a public dataset of human lung tissue is obtained and the dataset is preprocessed, which specifically includes the following steps: Step S11: selecting 4 data sets from the GEO database, namely GSE136831, GSE135893, GSE128033 and GSE122960, all of which extracted human lung tissue and separated single cells for sequencing; Step S12: performing data cleaning on each of the above data sets, wherein the data cleaning includes removing genes that are not expressed in any cells, removing cells that express less than 200 genes, and removing cells that express more than 25% of mitochondrial-related genes; Step S13: performing data preprocessing on the cleaned data set, the preprocessing including Log Normalization data standardization, data scaling, removing cell cycle signals, and selecting 8000 highly variable genes for each data set for subsequent data integration; In step S20, the step of performing data fusion on the preprocessed data set and removing batch effects specifically includes the following steps: Step S21: finding projections of the plurality of data sets by a canonical correlation analysis algorithm so as to maximize the correlation between all data sets; Step S22: using a dynamic time planning adjustment algorithm to determine the best mapping between the data sets; In step S30, the step of clustering the fused data set, marking the cell type of each group of cells, and screening macrophages specifically includes the following steps: clustering the fused data set, marking the cell type of each group of cells, and screening macrophages using a clustering analysis algorithm based on community detection, the steps are as follows: Step S31: using the data after dimensionality reduction as input data and cells as nodes, the Euclidean distance between each node is calculated by the KNNG algorithm to determine the K-nearest neighbors of each node; Step S32: Use the following formula to calculate the neighborhood overlap between each node and its K nearest neighbor nodes to construct a shared nearest neighbor matrix, where A and B represent the number of "neighbor" nodes of the two nodes): J(A,B)=(|A∩B|) / (|A∪B|)=(|A∩B|) / (|A|+|B|-|A∩B|); Step S33: The Louvain algorithm is used to scan nodes and their neighboring nodes, calculate the modularity of each node and measure the module benefit after adding the module, select the neighboring node with the maximum benefit to add to the module, repeat the iteration, and finally form a node cluster; Step S34: using the Wilcoxon rank sum test to perform differential analysis, and judging the cell type and screening macrophages based on the genes differentially upregulated in each group of cells; In step S40, the step of identifying rare macrophage subtypes and differentially expressed genes in the macrophages specifically includes the following steps: Three rare macrophage subsets consisting of IPF samples were identified and recorded as Cluster 0, Cluster 1, and Cluster 2. The three groups of cells all expressed IPF markers, namely SPP1, CCL2, FABP4, and CHIT1, among which: SPP1 is a gene encoding osteopontin, which can significantly promote the migration and proliferation of fibroblasts and epithelial cells, and is used as a disease marker for IPF; CCL2 is a chemokine 2 that can recruit mononuclear macrophages and promote fibrosis through various mechanisms involving inflammation, angiogenesis, and myofibroblast accumulation, and is used as an IPF disease marker; FABP4 encodes fatty acid binding protein, which is a cytoplasmic fatty acid chaperone protein expressed in adipocytes and myeloid cells, and can promote ATP polarization of macrophages to M1 type, and participates in the occurrence of IPF by promoting the activation of M1 macrophages; CHIT1 encodes chitintriosidase, which has a profibrotic effect and is significantly expressed in the lungs of IPF patients.

2. The method for identifying rare macrophage subsets and disease markers in idiopathic pulmonary fibrosis according to claim 1, characterized in that: In step S21, the step of finding projections of the plurality of data sets by a canonical correlation analysis algorithm so as to maximize the correlation between all data sets specifically includes the following steps: Step S211: performing singular value decomposition on the highly variable gene expression matrix of each data set to obtain an initial canonical correlation vector CCV; Step S212: Then, the typical correlation vector W corresponding to each data set is updated by performing singular value decomposition on each data set and other data sets until the difference ratio of W before and after the update is greater than a given threshold.

3. The method for identifying rare macrophage subsets and disease markers in idiopathic pulmonary fibrosis according to claim 2, characterized in that: In step S22, the step of using a dynamic time planning adjustment algorithm to determine the optimal mapping between the data sets specifically includes the following steps: In a single-cell dataset, the dynamic time planning adjustment algorithm is used to calculate a warp path between the datasets to minimize the distance between the datasets, wherein W(w1, w2…wk) is a warp matrix, each vector of which corresponds to a point in the warp path, which maps an element in dataset X to dataset Y and minimizes the distance between them; the typical correlation vector is mapped to the warp matrix to align the two datasets into a low-dimensional space; For more than two datasets, the dataset with the largest number of cells is used as the reference dataset; the other datasets are aligned with the reference dataset respectively, and finally the typical correlation vector of each dataset is normalized to the common calibration space defined by the reference dataset.

4. The method for identifying rare macrophage subsets and disease markers in idiopathic pulmonary fibrosis according to claim 3, characterized in that: In step S33, Louvain algorithm is used to scan nodes and their neighboring nodes, calculate the modularity of each node and measure the module benefit after adding the module, select the neighboring node with the maximum benefit to add the module, repeat the iteration, and finally form a node cluster, which specifically includes the following steps: The Louvain algorithm is used to scan the node and its neighboring nodes, and the modularity of each node is calculated using the following formula, where m is the sum of the weights of the edges in the graph, i and j represent two nodes, Aij represents the weight between the two nodes, which is calculated by SNN, ki and kj represent the sum of the weights of all edges of nodes i and j, ci and cj represent the group to which nodes i and j belong, and δ is the Kronecker delta function): In order to maximize the Q value, the Louvain algorithm calculates the modularity gain and iterates repeatedly, as shown in the following formula, where Σin is the sum of the internal node weights of the community that node i enters, Σtot is the sum of the weights of the edges connecting node i and all communities in the graph, and ki,in is the sum of the weights of the edges between node i and all nodes in the community that i wants to move into:

5. A system for identifying rare macrophage subpopulations and disease markers in idiopathic pulmonary fibrosis using the method for identifying rare macrophage subpopulations and disease markers in idiopathic pulmonary fibrosis according to claim 1, characterized in that: include: A data set acquisition unit: acquiring a public data set of human lung tissue and preprocessing the data set; Preprocessing unit: performing data fusion on the preprocessed data set and removing batch effects; Screening unit: clustering the fused data set, marking the cell type of each group of cells, and screening macrophages; Identification unit: identifying rare macrophage subtypes and their differentially expressed genes in the macrophages.

6. A storage medium, characterized in that: The storage medium stores a program file capable of implementing the method for identifying rare macrophage subpopulations and disease markers in idiopathic pulmonary fibrosis as described in any one of claims 1 to 4.

7. A processor, characterized in that: The processor is used to run a program, wherein the program, when running, executes the method for identifying rare macrophage subpopulations and disease markers in idiopathic pulmonary fibrosis as described in any one of claims 1 to 4.

Images