A novel single-cell tcr sequencing computational method

Abstract

The application belongs to the technical field of single cell sequencing, and particularly relates to a single cell TCR sequencing calculation method. The application firstly sets a sliding rectangle in a plane rectangular coordinate system, then performs difference analysis on multi-group data in the rectangle, and finally integrates the difference analysis results. A sliding window with a suitable width is created according to different data sets, and the narrower the sliding window, the more accurate the analysis. The proportions of different cells in each subgroup are subjected to difference analysis, and the difference results are recorded. The sliding window continues to slide and continuously records, and finally a difference analysis matrix is generated. The difference analysis by the sliding window method can effectively avoid the bias caused by the number of cells, so that more accurate analysis results are obtained. The application has very good analysis effect on small data sets.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of single-cell sequencing, and particularly relates to a single-cell TCR sequencing calculation method. BACKGROUND

[0002] Single-cell sequencing technology is a major breakthrough in the field of biology in recent years, which can conduct research on genome, transcriptome, epigenetics and other aspects at the level of single cell. The rapid development of single-cell sequencing technology promotes the in-depth understanding of cell diversity, cell function and cell interaction.

[0003] Single-cell transcriptome sequencing (scRNA-seq) is the most common technology in single-cell sequencing technology. scRNA-seq extracts RNA from single cells, and conducts reverse transcription and sequencing to obtain the gene expression profile of single cells. scRNA-seq technology can quantitatively analyze the expression level of all transcripts in cells at the level of single cell, thereby revealing the functional state and differentiation state of cells. Single-cell TCR sequencing (scTCR-seq) is another important application of single-cell sequencing technology. scTCR-seq extracts TCR genes from single T cells, and conducts amplification and sequencing to obtain the TCR receptor sequence of single T cells. scTCR-seq technology can analyze the TCR diversity of T cells at the level of single cell, and study the recognition and response of T cells to antigens.

[0004] The application of single-cell sequencing technology in T cell clonal expansion research mainly includes the following aspects:

[0005] Kinetics of T cell clonal expansion. Single-cell sequencing technology can be used to analyze the starting time, expansion speed and cell types of T cell clonal expansion. For example, researchers used scRNA-seq technology to analyze the kinetics of T cell clonal expansion in the human immune system, and found that T cell clones rapidly expanded within a few days after infection, and gradually subsided within a few weeks; Immune regulation mechanism of T cell clonal expansion. Single-cell sequencing technology can be used to analyze immune regulation factors such as cytokines, chemokines and growth factors during T cell clonal expansion. For example, researchers used scTCR-seq technology to analyze cytokine expression during T cell clonal expansion, and found that cytokines IL-17A and IL-21 played an important role in T cell clonal expansion; Tumor immunotherapy research of T cell clonal expansion. Single-cell sequencing technology can be used to analyze the role of T cell clonal expansion in tumor immunotherapy. For example, researchers used scTCR-seq technology to analyze the characteristics of T cell clonal expansion after CAR-T cell therapy, and found that T cell clones highly expanded in tumor tissues after CAR-T cell therapy, and produced strong killing effect on tumor cells.

[0006] The single-cell sequencing technology has the following advantages in the study of T cell clonal expansion: it can be studied at the single-cell level. Traditional immunology research is usually carried out at the batch level, which is difficult to reflect the true situation of T cell clonal expansion. The single-cell sequencing technology can be studied at the single-cell level, thereby obtaining more accurate and comprehensive results; it can analyze the dynamic process of T cell clonal expansion. Traditional immunology research is usually static, and it is difficult to reflect the dynamic process of T cell clonal expansion. The single-cell sequencing technology can analyze the dynamic process of T cell clonal expansion, thereby revealing the rules of T cell clonal expansion; it can analyze the immune regulation mechanism of T cell clonal expansion. Traditional immunology research is usually single-factor analysis, and it is difficult to comprehensively analyze the immune regulation mechanism of T cell clonal expansion. The single-cell sequencing technology can analyze multiple factors in the process of T cell clonal expansion, thereby revealing the immune regulation mechanism of T cell clonal expansion.

[0007] With the continuous development of single-cell sequencing technology, single-cell sequencing technology will play an increasingly important role in the study of T cell clonal expansion, but due to the relative abundance of TCR itself in different cell numbers, it is impossible to directly prove that the difference is caused by clonal expansion or by the bias caused by cell number. Therefore, a new analysis method is needed to balance the difference caused by cell number, so that the analysis of single-cell TCR level is more accurate. SUMMARY

[0008] The purpose of the present application is to provide a single-cell TCR sequencing calculation method to balance the bias caused by different cell numbers, so that the analysis result is more accurate.

[0009] The single-cell TCR sequencing calculation method provided by the present application is a new analysis method based on the concept of "sliding window" (Sliding Window). Specifically, in a plane rectangular coordinate system, a sliding rectangle is set, then the multi-group data inside the rectangle is analyzed, and finally the difference analysis results are integrated. The specific steps are as follows:

[0010] (1) The gene expression profile of the single-cell data set is numerically standardized to generate a single-cell expression matrix with behavior genes and columns as cell barcodes, which facilitates subsequent data processing. Then, principal component analysis (PCA) is used to reduce the dimensionality of the data. Specifically, first, the data is centered, i.e. the mean of each gene is set to 0. Then, the covariance of the de-centered gene expression data is calculated. According to the size of the covariance, the contribution of the gene to the difference between cells is represented, thereby realizing the dimensionality reduction of the data;

[0011] (2) Since different cell types have different gene expressions, according to the covariance obtained after the above dimension reduction, the top 2000 genes with the largest contribution to the difference between cells are selected as the high variable genes for distinguishing cell types, cells with similar high variable genes are clustered, and the clustered cells are embedded in the UMAP (Uniform Manifold Approximation and Projection) coordinate system. In the UMAP coordinate system, each data point represents a cell, and the closer the cells are on the coordinate axis, the closer the gene expression levels of the cells are, which can be considered as similar cell groups;

[0012] (3) According to the number of cells and the required analysis accuracy, set the width of the sliding window. The smaller the width, the more accurate the difference analysis result, but the computational load increases. In actual operation, a balance point between accuracy and computational load needs to be determined;

[0013] (4) Starting from the origin of the coordinate axis, calculate the proportion of TCR in the corresponding group from cells in different groups within the sliding window. The difference calculation formula is:

[0014]

[0015] Where i and j represent different groups; sum i represents the number of TCR in the i group within the sliding window, and ∑ i represents the data of all TCR in the i group, and the proportion of TCR in the i group to the total TCR in the sliding window can be obtained. The same is true for the j group;

[0016] (5) Calculate the proportion difference between the i group and the j group within the above sliding window, and record the difference value in the behavior difference proportion column in the difference matrix of the cells;

[0017] (6) Then move the sliding window to the next position and repeat steps (4)-(5) until the sliding window moves outside the coordinate system (or there are no cells in the sliding window). Then, add the difference matrix to the metadata according to the barcode of the cells, and the sliding window analysis is complete.

[0018] Since TCR reflects the activation state of T cells, the difference in the proportion of TCR in different cell groups can be used to infer the state of T cells, which can reflect the immune status of the body. For single-cell data sets, fewer TCR species generally means better T cell activation, but the number of species is greatly affected by sequencing depth and cell number. In large-scale data sets, the proportion of TCR in different groups in a certain cell group is usually directly calculated, but for small data sets, such calculation does not take into account the impact of sequencing depth. Therefore, sliding window analysis is a more accurate and necessary analysis method.

[0019] The method has the following advantages:

[0020] (1) The concept of sliding window is first applied to single cell data, which has the advantages of easy understanding, simple operation and fast analysis speed;

[0021] (2) Single cell data sets are expensive, and most researchers cannot obtain a large amount of single cell data. The method can not only analyze large-scale single cell data sets, but also has better analysis effect on smaller data sets. BRIEF DESCRIPTION OF DRAWINGS

[0022] Figure 1 is a schematic diagram of sliding window analysis, which shows the working principle and basic concept of sliding window.

[0023] Figure 2 is a flow chart of sliding window analysis.

[0024] Figure 3 is a coordinate diagram of the example data after PCA dimension reduction.

[0025] Figure 4 is a UMAP coordinate diagram after clustering the PCA reduced data.

[0026] Figure 5 is the starting sliding position of the starting site of the sliding window in the example data. The black rectangular part in the figure is a sliding window position diagram.

[0027] Figure 6 is a part of the TCR difference matrix generated after sliding window analysis.

[0028] Figure 7 is the final result of sliding window analysis in the example data. DETAILED DESCRIPTION

[0029] The application will be further described by specific examples combined with the drawings.

[0030] Step one: Seurat is a classic package in single cell transcriptome analysis. First, import the single cell data set into R through the R language package Seurat, and at the same time, import the single cell TCR data according to the one-to-one correspondence of cell barcode into R as metadata for subsequent analysis.

[0031] Step two: process the transcriptome data using the NormalizeData command in Seurat to obtain the Log2 normalized expression matrix, where the rows are genes and the columns are cell barcodes, so that it becomes linearly distributed.

[0032] Step three: The log processed data is filtered for the first time. First, the PercentageFeatureSet function is used to calculate the proportion of mitochondrial genes in each cell. Cells containing more than 10% mitochondrial genes are considered dead cells, and the expression of dead cells is very different from that of living cells, which cannot truly reflect cell information, so they need to be removed to prevent affecting subsequent analysis.

[0033] Step four: The second filtering is performed. Cells expressing less than 200 genes or more than 6000 genes in a cell are considered abnormal cells. These cells mainly come from system errors in the sequencing process, so they need to be excluded.

[0034] Step five: After data filtering, RunPCA command is used to reduce the dimension of the data, as shown in Figure 3 , which shows the distribution of the data after reducing to two dimensions. The reduced data is further clustered and projected into the UMAP coordinate system, and the cells are labeled according to gene expression, as shown in Figure 4 , which shows the clustering of different cells.

[0035] Step six: The UMAP coordinate matrix is extracted, and the sliding window width is set to 3, as shown in Figure 5 , the TCR information of all cells within the left and right boundaries of the sliding window is extracted according to the coordinate information.

[0036] Step seven: Calculate the proportion of A group cell TCR in all A group cells in the sliding window as Diff A , calculate the proportion of B group cell TCR in all B group cells as Diff B , and record Diff AB .

[0037] Step eight: The sliding window moves to the next position, and step five is calculated, and so on, until the sliding window exceeds the UMAP cell margin.

[0038] Step nine: The difference matrix is generated by counting all Diff data, as shown in Figure 6 , and then the difference matrix is imported into the metadata of the original data to analyze the difference in TCR proportion of different cell types in different groups.

[0039] Figure 7 is the final result of the sliding window analysis in the example data. As can be seen from Figure 7 , the number of cells in this single cell data set is 1500, which is relatively small, and cannot be analyzed using conventional methods. After using the sliding window, it can be seen that the TCR in A group is mainly in effector CD8 +The higher number of T cells in group A means that the T cell activation is stronger, which can only be concluded by the sliding window analysis.

Claims

1. A single-cell TCR sequencing computational method, characterized in that, In the Cartesian coordinate system, a sliding rectangle is set, and then the multi-group data inside the rectangle is analyzed, and finally the difference analysis results are integrated. The specific steps are as follows: (1) The gene expression profile of the single cell data set is numerically standardized to generate the behavior gene, and the single cell expression matrix is set as the cell barcode, which is convenient for subsequent data processing. Then, principal component analysis is used to reduce the dimension of the data. Specifically, first, the data is centralized, that is, the mean value of each gene is set to 0, and then the covariance of the decentered gene expression data is calculated. According to the size of the covariance, the contribution degree of the gene to the difference between cells is represented to realize the dimension reduction of the data; (2) Because different cell types have different gene expressions, according to the covariance obtained after dimension reduction, the top 2000 genes with the largest contribution to the difference between cells are selected as the high variable genes for distinguishing cell types. Cells with similar high variable genes are clustered, and the clustered cells are embedded in the UMAP coordinate system. In the UMAP coordinate system, each data point represents a cell, and the closer the cells are on the coordinate axis, the closer the gene expression levels of the cells are, and they are considered to be similar cell groups; (3) According to the number of cells and the required analysis accuracy, the width of the sliding window is set; (4) Starting from the origin of the coordinate axis, the proportion of TCR in the corresponding group is calculated for cells from different groups in the sliding window. The difference calculation formula is: where i and j represent different groups; sum i represents the number of TCRs in group i within the sliding window, ∑ i represents the data of all TCRs in group i, and further obtains the proportion of TCRs of group i in the sliding window to the total TCRs of the cells, and the same is true for group j; (5) The proportion difference of i group and j group in the above sliding window is calculated, and the difference value is recorded in the behavior difference proportion and the difference matrix of the cells; (6) Then the sliding window moves to the next position, and the steps of (4)-(5) are repeated until the sliding window moves to the outside of the coordinate system, or there is no cell in the sliding window. Then, according to the cell barcode, the difference matrix is added to the metadata, and the sliding window analysis is completed.

Images