Methods, devices and uses for analyzing the status of immune activation molecules associated with antigen presentation by graft cells

Abstract

The application discloses a method, device and application for analyzing the state of graft cell antigen presentation related immune activation molecules. The method comprises the following steps: obtaining graft transcriptome data and normalizing; determining an antigen presentation transformation genome composed of major histocompatibility complex class I molecule genes, antigen processing and transport genes, immunoproteasome genes, transcriptional activator genes and interferon response genes; standardizing the genome in a reference cell population and averaging to obtain an antigen presentation transformation score; dividing the activation or silence state according to a threshold; calculating a rejection related molecular state index according to the proportion of activated cells and outputting. The application converts the qualitative antigen presentation sample transformation into a continuous score and discrete typing which can be compared across samples by using a fixed genome combination, reference population standardization, etc. The output of the application is an ex vivo molecular state parameter, which does not constitute a clinical diagnosis, prognosis prediction or treatment decision.

Description

Technical Field

[0001] This application relates to the field of transplantation immunoassay technology, and in particular to a method, apparatus and application for analyzing the molecular state of immune activation associated with antigen presentation in transplanted cells. Background Technology

[0002] Islet transplantation is an effective treatment for type 3 diabetes, but the grafts often gradually lose function due to immune rejection. Clinically, the assessment of the degree of immune activation after transplantation mainly relies on functional indicators such as blood glucose and C-peptide, as well as histological observation of graft biopsy. The former is lagging and cannot reflect immune activation at the molecular level, while the latter is an invasive procedure and difficult to perform continuously.

[0003] Transcriptome-based analysis can reflect graft immune status at the molecular level. However, current transcriptome-based graft immune assessments are mostly qualitative, focusing on listing differentially expressed genes or enriching pathways. They lack fixed gene combinations, clear scoring rules, and reproducible genotyping thresholds, making it difficult to directly compare results between different samples and batches, and also making it difficult to standardize and implement them in devices and kits.

[0004] Therefore, there is a need for a technical solution that, based on fixed gene combinations and clear scoring rules, can transform the molecular state of graft immune activation into reproducible continuous scoring and discrete typing. Summary of the Invention

[0005] The purpose of this application is to provide a new method, apparatus, and application for analyzing the molecular state of immune activation associated with graft cell antigen presentation.

[0006] To achieve the above objectives, this application adopts the following technical solution:

[0007] This application discloses a method for analyzing the molecular state of immune activation related to antigen presentation in graft cells, comprising the following steps:

[0008] The normalized expression matrix analysis steps include acquiring single-cell transcriptome expression data based on graft tissue, performing quality control and log normalization on the data, and obtaining the normalized expression matrix.

[0009] The steps to identify the antigen-presenting transformation genome include determining the antigen-presenting transformation genome from the normalized expression matrix. The antigen-presenting transformation genome includes major histocompatibility complex class I molecular genes, antigen processing and transport genes, immunoproteasome genes, transcription activator genes, and interferon response genes.

[0010] The antigen presentation conversion scoring step includes: for each gene in the antigen presentation conversion genome, subtracting the mean and standard deviation of the gene's expression value in the reference cell population from the expression value of the target cell population to obtain a standardized score; and calculating the arithmetic mean of the standardized scores of each gene in the antigen presentation conversion genome of the same cell with equal weights to obtain the antigen presentation conversion score of the cell; the target cell population is one or more cell types specified in the graft; the reference cell population is a set of cells corresponding to the target cell type in the reference dataset, and the reference dataset includes a first immune status graft group and a second immune status graft group, which are combinations of allogeneic grafts and syngeneic grafts, or combinations of rejected grafts and non-rejected grafts, or combinations of active rejected grafts and tolerant grafts;

[0011] The molecular state typing determination step includes classifying each cell in the target cell population into an antigen-presenting activated state or a quiescent state based on a pre-determined threshold on a reference dataset independent of the data to be typed, thereby obtaining the molecular state typing.

[0012] The steps for calculating the rejection-related molecular state index include dividing the number of cells in the antigen-presenting activated state by the total number of cells in the target cell population that have passed quality control to obtain the proportion of antigen-presenting activated cells, which serves as the rejection-related molecular state index of the graft.

[0013] The output steps include outputting molecular state typing and rejection-related molecular state indices as quantitative results of the molecular state of graft cell antigen presentation-related immune activation.

[0014] It should be noted that the method of this application, through a fixed antigen-presenting transformation genome combination, antigen-presenting transformation score, and a pre-determined threshold, transforms qualitative antigen-presenting transformation into a reproducible, cross-sample comparable continuous score and discrete genotyping, enabling direct comparison of results from different samples and batches. In some implementations of this application, the method is used to analyze the molecular status of islet cell antigen presentation-related immune activation. The area under the curve (AUC) for the score to distinguish between allogeneic and syngeneic islet cells reaches 0.896; the AUC for migration validation of this genome in an independent mouse kidney transplant rejection cohort after freezing reaches 0.969.

[0015] It should also be noted that the method in this application was initially developed based on single-cell transcriptome expression data from mouse islet transplantation; furthermore, it can be extended to graft analysis in other species such as humans via homology mapping; the validation examples in this application cover real clinical samples from human kidney transplant biopsies, and further validation of human islets can be implemented in the future after accumulating more publicly available human islet transplantation data. Furthermore, research has found that this method is not only applicable to the analysis of cell antigen presentation-related immune activation molecular states in islet transplantation, but also to the analysis of other graft cells, such as endothelial cells, myeloid cells, and mesenchymal cells.

[0016] It should also be noted that the output steps of this application yield cellular and molecular state parameters reflected in in vitro transcriptomic data, which themselves do not constitute a diagnosis, prognostic prediction, or treatment decision regarding individual graft rejection, tolerance, graft function, or clinical outcome; this application's method does not include a step of determining whether a subject has experienced graft rejection based on the quantitative results. The output of this application can be used for standardized assessment of graft immune activation status, research purposes, or non-clinical diagnostic purposes such as drug screening.

[0017] In this application, the combination of the first immune status graft group and the second immune status graft group includes, but is not limited to: (i) Allogeneic grafts and syngeneic grafts; (ii) Rejected grafts and non-rejected grafts; (iii) Active rejection grafts and tolerant grafts.

[0018] In this application, "first" and "second" are used only to distinguish two different immune status groups and do not imply their superiority, priority, or any biological order. It is understood that other pairwise distinguishable graft combinations with different immune statuses are also applicable to this application.

[0019] In one implementation of this application, when the target is a mouse, the major histocompatibility complex class I molecular genes include B2m, H2-K1, H2-D1, H2-Q7, and H2-T23; antigen processing and transport genes include Tap1, Tap2, Tapbp, and Tapbpl; immunoproteasome genes include Psmb8, Psmb9, and Psmb10; transcription activator genes include Nlrc5 and Irf1; and interferon response genes include Stat1, Gbp2, Isg15, and Ifit3. That is, the antigen presentation and transformation genome comprises a total of 18 genes.

[0020] When the subject of the test is a human, the antigen-presenting transformation genome includes HLA-A, HLA-B, HLA-C, HLA-E, HLA-F, B2M, TAP1, TAP2, TAPBP, TAPBPL, PSMB8, PSMB9, PSMB10, NLRC5, IRF1, STAT1, GBP1, GBP2, ISG15, and IFIT3.

[0021] It should be noted that the above 18 genes are mouse major histocompatibility complex class I molecular genes, antigen processing and transport genes, immunoproteasome genes, transcription activator genes, and interferon response genes. When the test subject is another species, these genes need to be mapped to the genes of the corresponding species according to homology. For example, when mapping the above antigen-presenting transformation genomes to human genes according to homology, the antigen-presenting transformation genomes include HLA-A, HLA-B, HLA-C, HLA-E, HLA-F, B2M, TAP1, TAP2, TAPBP, TAPBPL, PSMB8, PSMB9, PSMB10, NLRC5, IRF1, STAT1, GBP1, GBP2, ISG15, and IFIT3.

[0022] In this application, homology mapping is based on homology relationships in authoritative databases, including but not limited to publicly available resources such as HomoloGene, Ensembl Compara, and NCBI Orthologs maintained by the National Institutes of Health. When applied across species, the corresponding genes for each module are determined one by one according to one-to-one or one-to-many homology relationships, without subjective selection. For example, when mapping the 18 genes in this application to human genes according to homology, it includes 20 genes: HLA-A, HLA-B, HLA-C, HLA-E, HLA-F, B2M, TAP1, TAP2, TAPBP, TAPBPL, PSMB8, PSMB9, PSMB10, NLRC5, IRF1, STAT1, GBP1, GBP2, ISG15, and IFIT3.

[0023] In one implementation of this application, the antigen presentation conversion score is calculated according to the following formula: for cell i and gene j in the target cell population, the standardized score is the expression value of gene j in cell i minus the mean expression value of gene j in the reference cell population, divided by the standard deviation of gene j expression in the reference cell population; the antigen presentation conversion score is the arithmetic mean of the standardized scores of all genes in the antigen presentation conversion genome of cell i.

[0024] In one implementation of this application, the predetermined threshold for the molecular state typing judgment step is the score value that maximizes the Youden index for distinguishing between the two types of cells, the first immune state graft group and the second immune state graft group, on the reference dataset.

[0025] In one implementation of this application, the target cell population is one or more of pancreatic islet cells, endothelial cells, myeloid cells, and interstitial cells, but does not include lymphocytes.

[0026] This application also discloses a device for analyzing the molecular state of immune activation related to antigen presentation in graft cells, including a normalized expression matrix analysis module, a module for confirming antigen presentation transformation genome, an antigen presentation transformation scoring module, a molecular state typing judgment module, a rejection-related molecular state index calculation module, and an output module.

[0027] The normalized expression matrix analysis module includes a method for acquiring single-cell transcriptome expression data based on graft tissue, performing quality control and log normalization on the data, and obtaining a normalized expression matrix.

[0028] The antigen-presenting transformation genome module was identified, including the antigen-presenting transformation genome for identification from the normalized expression matrix. The antigen-presenting transformation genome includes major histocompatibility complex class I molecular genes, antigen processing and transport genes, immunoproteasome genes, transcription activator genes, and interferon response genes.

[0029] The antigen presentation conversion scoring module includes a standardized score for each gene in the antigen presentation conversion genome. This score is calculated by subtracting the mean and standard deviation of the gene's expression value in the target cell population from the mean and then dividing by the standard deviation. The standardized scores of all genes in the same cell within the antigen presentation conversion genome are then arithmetically averaged with equal weights to obtain the antigen presentation conversion score for that cell. The target cell population is one or more specified cell types within the graft. The reference cell population is a set of cells in a reference dataset corresponding to the target cell type. The reference dataset includes a first immune state graft group and a second immune state graft group, which can be a combination of allogeneic and syngeneic grafts, a combination of rejected and non-rejected grafts, or a combination of active rejected grafts and tolerant grafts.

[0030] The molecular state typing module includes a threshold determined in advance on a reference dataset independent of the data to be typed, which is used to classify each cell in the target cell population into an antigen-presenting activated state or a quiescent state to obtain the molecular state typing.

[0031] The rejection-related molecular state index calculation module includes a method for dividing the number of cells in the antigen-presenting activated state by the total number of cells in the target cell population that have passed quality control to obtain the proportion of antigen-presenting activated cells, which serves as the rejection-related molecular state index of the graft.

[0032] The output module includes molecular state typing and rejection-related molecular state indices, which serve as quantitative results of the molecular state of graft cell antigen presentation-related immune activation.

[0033] It should be noted that the apparatus for analyzing the molecular state of graft-cell antigen presentation-related immune activation in this application is actually implemented by each module separately as described in this application. Therefore, the specific limitations of each module can be found in the method for analyzing the molecular state of graft-cell antigen presentation-related immune activation in this application. For example, the specific genes of the antigen presentation transformation genome, the calculation method of the antigen presentation transformation score, the determination method of the threshold, the target cell population, etc., can all be found in the method for analyzing the molecular state of graft-cell antigen presentation-related immune activation in this application, and will not be elaborated here.

[0034] This application also discloses an apparatus for analyzing the molecular state of graft cell antigen presentation-related immune activation, the apparatus comprising a memory and a processor; wherein the memory is used to store a program; and the processor is used to execute the program stored in the memory to implement the method of this application for analyzing the molecular state of graft cell antigen presentation-related immune activation.

[0035] This application also discloses a computer-readable storage medium including a program that can be executed by a processor to implement the method of this application for analyzing the molecular state of immune activation related to graft cell antigen presentation.

[0036] This application also discloses a detection kit for quantifying and genotyping graft rejection-related molecular states, comprising reagents for detecting the expression levels of each gene in the antigen-presenting transformed genome in the method for analyzing the molecular state of graft cell antigen presentation-related immune activation of this application; the reagents include one or more of specific amplification primers, hybridization probes, or hybridization chips, for in vitro specific amplification, hybridization capture, or chip hybridization detection of the mRNA or cDNA of each gene in the test sample.

[0037] In one implementation of this application, the kit further includes an instruction manual, which records the rules for calculating the antigen presentation conversion score and a predetermined threshold.

[0038] This application also provides the application of antigen-presenting transforming genomes in the preparation of detection kits for quantifying and genotyping molecular states related to graft rejection. The antigen-presenting transforming genome includes major histocompatibility complex (MHC) class I genes, antigen processing and transport genes, immunoproteasome genes, transcription activator genes, and interferon response genes. When the detection subject is a mouse, the MHC class I genes include B2m, H2-K1, H2-D1, H2-Q7, and H2-T23; the antigen processing and transport genes include Tap1, Tap2, Tapbp, and Tapbpl; and the immunoproteasome genes... Because it includes Psmb8, Psmb9, and Psmb10, transcription activator genes include Nlrc5 and Irf1, and interferon response genes include Stat1, Gbp2, Isg15, and Ifit3; when the subject is human, the antigen-presenting transformation genome includes HLA-A, HLA-B, HLA-C, HLA-E, HLA-F, B2M, TAP1, TAP2, TAPBP, TAPBPL, PSMB8, PSMB9, PSMB10, NLRC5, IRF1, STAT1, GBP1, GBP2, ISG15, and IFIT3.

[0039] It should be noted that this application creatively uses the combination of the above genes to analyze the molecular state of immune activation related to graft cell antigen presentation. Combined with the method and device of this application, the molecular state of antigen presentation transformation related to pancreatic islet graft rejection can be transformed into a reproducible and cross-sample comparable continuous score and discrete subtyping, and a quantitative characterization at the graft level can be provided.

[0040] Due to the adoption of the above technical solutions, the beneficial effects of this application are as follows:

[0041] The method for analyzing the molecular state of antigen presentation-related immune activation in graft cells in this application transforms qualitative antigen presentation transformation into reproducible, cross-sample comparable continuous scores and discrete genotyping by using fixed antigen presentation transformation genome combinations, antigen presentation transformation scores, and predetermined thresholds, so that results from different samples and batches can be directly compared. Attached Figure Description

[0042] Figure 1 This is a flowchart of the method for analyzing the molecular state of immune activation related to graft cell antigen presentation in the embodiments of this application;

[0043] Figure 2 This is a structural block diagram of the device for analyzing the molecular state of immune activation related to graft cell antigen presentation in the embodiments of this application;

[0044] Figure 3This is a UMAP dimensionality reduction map of the single-cell transcriptome atlas in the embodiments of this application; wherein, the four maps A, B, C, and D are colored according to 5 principal cell types, 10 cell subtypes, transplantation group (allogeneic / homogeneous) and 23 Seurat clusters, respectively;

[0045] Figure 4 These are marker gene expression point maps in the embodiments of this application. Map A shows the expression of classic marker genes in 5 principal cell types and Map B shows the expression of classic marker genes in 10 cell subtypes. The size of the dot represents the proportion of cells expressing the gene, and the color intensity represents the average normalized expression level.

[0046] Figure 5 This is a heatmap of the expression of 18 genes of the antigen-presenting transformation genome in pancreatic islet cells of allogeneic and syngeneic grafts in the embodiments of this application, arranged according to the functional modules of MHC-I molecules, antigen processing and transport, immunoproteasome, transcription activator and interferon response;

[0047] Figure 6 This is a distribution density map of pancreatic islet cell antigen presentation conversion score (APTS) in allogeneic and syngeneic grafts in the embodiments of this application. The dashed line represents the Youden threshold of 0.063, and the caption indicates the activation ratio of the two groups.

[0048] Figure 7 This application example shows the ROC curve and Youden threshold determination diagram (cell level) for distinguishing between allogeneic and syngeneic pancreatic islet cells by antigen presentation conversion score, with the area under the curve and its 95% confidence interval, threshold, sensitivity, specificity and Youden index marked.

[0049] Figure 8 In this embodiment of the application, after the 18-gene antigen-presented transformed genome obtained from GSE198865 was frozen, the ROC curve (cell level) of the rejected and immature graft cells was distinguished on an independent mouse kidney transplant rejection single-cell cohort (GSE157292), showing the area under the curve and its 95% confidence interval for cross-organ migration.

[0050] Figure 9 The following are bar charts showing the discrimination of antigen-presenting transformed genomes after freezing in this application embodiment: Figure A shows the discrimination between allogeneic and syngeneic grafts by each principal cell population in the internal islet transplantation cohort (GSE198865), and Figure B shows the area under the curve of the discrimination between rejected and immature grafts by each coarsely subdivided cell population in the independent external kidney transplantation cohort (GSE157292). In both cohorts, the score is discriminative in multiple parenchymal and stromal cell populations, rather than just islets or a single cell type, showing that this method is not limited to islet cells and that this characteristic also holds true in external datasets.

[0051] Figure 10This embodiment of the application maps the antigen-presenting transformed genome to human genes according to homology, and then distinguishes between rejecting and non-rejecting cells on the independent human kidney transplant biopsy dataset GSE145927; Figure A shows the ROC curve at the entire cell level, and Figure B shows the area under the curve for each cell type; it shows that the fixed genome, after being mapped to human homology, still has migration characterization ability in real clinical human transplant rejection, and is cell type dependent;

[0052] Figure 11 This embodiment of the application maps antigen-presenting transformed genomes to human genes according to homology, and then distinguishes between rejecting and non-rejecting cells on a second independent human kidney transplant biopsy dataset, GSE228300; Figure A shows the ROC curves at the entire cell level, and Figure B shows the area under the curve for each cell type; and Figure 10 Together, they demonstrate the migration characterization capability of this fixed genome in two independent human cohorts;

[0053] Figure 12 This is a UMAP dimension reduction plot of the external mouse kidney transplantation cohort GSE157292 in the embodiments of this application; Figure A is a coloring plot of immune T / NK cells, myeloid cells, B cells, endothelial cells, epithelial / renal tubules, and interstitial cells according to coarse cell types; Figure B is a coloring plot of transplantation status: immature, rejection, and tolerance.

[0054] Figure 13 These are UMAP dimensionality reduction maps of two external human kidney transplant cohorts in this application embodiment; Figures A and B are color maps of GSE145927 grouped by cell type and rejection, respectively; Figures C and D are color maps of GSE228300 grouped by cell type and rejection, respectively. Detailed Implementation

[0055] To address the challenge of transforming the molecular state of antigenic translocation associated with islet graft rejection into a reproducible, cross-sample comparable continuous scoring and discrete typing using fixed gene combinations and well-defined scoring and typing rules, and to provide a quantitative characterization at the graft level.

[0056] This application creatively proposes the following solution: Transcriptome expression data of pancreatic islet graft samples are obtained and subjected to quality control and logarithmic normalization to obtain a normalized expression matrix; an antigen-presenting transformation genome is identified, composed of major histocompatibility complex class I molecular genes, antigen processing and transport genes, immunoproteasome genes, transcription activator genes, and interferon response genes; each gene within this genome is standardized within the target cell population using the mean and standard deviation of a reference cell population, and the average of the standardized scores of each gene within this genome for the same cell is obtained to obtain the antigen-presenting transformation score of the islet cell; islet cells are classified into antigen-presenting activated or quiescent states using a preset threshold to obtain molecular state typing; a rejection-related molecular state index is calculated based on the proportion of activated islet cells; and the molecular state typing and rejection-related molecular state index are output.

[0057] In some implementations, this application uses a method to analyze the molecular state of immune activation associated with graft cell antigen presentation, such as... Figure 1 As shown, the process includes step 11, normalized expression matrix analysis, step 12, confirmation of antigen-presenting transformation genome, step 13, antigen-presenting transformation scoring, step 14, molecular state typing, step 15, calculation of rejection-related molecular state index, and output step 16.

[0058] Step 11 of the normalized expression matrix analysis includes acquiring single-cell transcriptome expression data based on graft tissue, performing quality control and logarithmic normalization to obtain a normalized expression matrix. For example, in some implementations, taking islet cell antigen presentation conversion scoring and typing as an example, specifically using single-cell transcriptome data from mouse islet syngeneic and allogeneic grafts on day 7 post-transplantation, the dataset contains 18,130 cells, including 6,948 syngeneic grafts and 11,182 allogeneic grafts. Clustering yields 23 clusters, 5 principal cell types, and 10 cell subtypes. Annotated islet cells comprise 663 cells, including 501 syngeneic grafts and 162 allogeneic grafts. The expression data undergoes quality control and logarithmic normalization to obtain the normalized expression matrix.

[0059] Step 12, identifying the antigen-presenting conversion genome, involves determining the antigen-presenting conversion genome from the normalized expression matrix. The antigen-presenting conversion genome includes major histocompatibility complex (MHC) class I molecular genes, antigen processing and transport genes, immunoproteasome genes, transcription activator genes, and interferon response genes. For example, in some implementations, the antigen-presenting conversion genome includes major histocompatibility complex (MHC) class I molecular genes B2m, H2-K1, H2-D1, H2-Q7, and H2-T23; antigen processing and transport genes Tap1, Tap2, Tapbp, and Tapbpl; immunoproteasome genes Psmb8, Psmb9, and Psmb10; transcription activator genes Nlrc5 and Irf1; and interferon response genes Stat1, Gbp2, Isg15, and Ifit3, totaling 18 genes. The aforementioned 18 genes have been reported in the art to be associated with antigen presentation or transplant rejection; this application solidifies them into a specific combination of 18 genes and combines them with a reference cell population-standardized explicit equal-weighted scoring formula and a freezing threshold for the quantification and typing of the molecular state of graft cells.

[0060] Step 13 of the antigen presentation conversion scoring includes: for each gene in the antigen presentation conversion genome, subtracting the mean and standard deviation of the gene's expression value in the reference cell population from the expression value of the gene in each cell of the target cell population and dividing by the standard deviation to obtain a standardized score; and calculating the arithmetic mean of the standardized scores of each gene in the antigen presentation conversion genome of the same cell with equal weights to obtain the antigen presentation conversion score of the cell. The target cell population is one or more cell types specified within the graft; the reference cell population is the set of cells corresponding to the target cell type in the reference dataset. The reference dataset includes a first immune state graft group and a second immune state graft group, which are combinations of allogeneic and syngeneic grafts, or combinations of rejected and non-rejected grafts, or combinations of active rejected grafts and tolerant grafts. For example, in some implementations, each gene is standardized in the reference cell population according to Formula 1, and the average of the standardized scores of each gene in the same cell is calculated to obtain the antigen presentation conversion score of the islet cell. The median score for pancreatic islet cells in allogeneic grafts was 0.552, while that for syngeneic grafts was -0.276. At the cellular level, the area under the curve for the score to distinguish between the two types was 0.896 (95% confidence interval of bootstrap method: 0.860 to 0.927).

[0061] Formula 1:

[0062]

[0063] In Formula 1, xij represents the log-normalized expression value of gene j in islet cell i, μj and σj represent the mean and standard deviation of gene j's expression value in the reference cell population (in Example 1, the reference cell population is a combined population of 663 islet cells from two groups), zij represents the intra-population standardized score of gene j in islet cell i, G represents the antigen-presenting transformation genome, |G| represents the number of genes in the genome, and Si represents the antigen-presenting transformation score of islet cell i. This formula first standardizes each gene within the reference cell population to eliminate differences in dimensions and expression amplitude between genes, and then calculates the arithmetic mean of the standardized scores of each gene within the genome of the same cell, thereby obtaining a continuous score that reflects the antigen-presenting transformation intensity of the cell and can be compared between different samples and batches.

[0064] Step 14, molecular state typing, includes classifying cells in the target cell population into antigen-presenting active or quiescent states using a pre-determined threshold on a reference dataset independent of the data to be typed, thus obtaining the molecular state typing. For example, in some implementations, a threshold of 0.063, which maximizes the Youden index for distinguishing allogeneic from homologous cells, is used. Islet cells with scores not lower than the threshold are classified as antigen-presenting active, while those without are classified as quiescent. At this threshold, the cell-level typing accuracy is 0.870.

[0065] Step 15, which calculates the rejection-related molecular state index, includes dividing the number of cells in an antigen-presenting activated state by the total number of cells in the target cell population that have passed quality control, to obtain the proportion of antigen-presenting activated cells, which is used as the rejection-related molecular state index of the graft. For example, in some implementations, the proportion of activated islet cells to islet cells is statistically analyzed, with 83.3% for allogeneic grafts and 11.8% for syngeneic grafts, and this proportion is used as the rejection-related molecular state index.

[0066] Step 16 outputs molecular state typing and rejection-related molecular state indices as quantitative results of the molecular state of graft cell antigen presentation-related immune activation. For example, in some implementations, the molecular state typing of each islet cell and the rejection-related molecular state index of the graft are output.

[0067] Those skilled in the art will understand that all or part of the functions of the methods described above can be implemented in hardware or by computer programs. When all or part of the functions in the above embodiments are implemented by computer programs, the program can be stored in a computer-readable storage medium, which may include: read-only memory, random access memory, disk, optical disk, hard disk, etc., and the program is executed by a computer to achieve the above functions. For example, the program can be stored in the memory of a device, and when the program in the memory is executed by the processor, all or part of the above functions can be achieved. Alternatively, when all or part of the functions in the above embodiments are implemented by computer programs, the program can also be stored in a server, another computer, disk, optical disk, flash drive, or portable hard drive, etc., and can be downloaded or copied to the memory of a local device, or the system of the local device can be updated. When the program in the memory is executed by the processor, all or part of the functions in the above embodiments can be achieved.

[0068] Therefore, based on the method for analyzing the molecular state of graft cell antigen presentation-related immune activation in this application, this application proposes a device for analyzing the molecular state of graft cell antigen presentation-related immune activation, such as... Figure 2 As shown, it includes a normalized expression matrix analysis module 21, a genome confirmation module 22, an antigen presentation and transformation scoring module 23, a molecular state typing judgment module 24, a rejection-related molecular state index calculation module 25, and an output module 26.

[0069] The module 21 for normalized expression matrix analysis includes: acquiring single-cell transcriptome expression data from graft tissue, performing quality control and log-normalization to obtain a normalized expression matrix; identifying the antigen-presenting transformation genome module 22, which identifies the antigen-presenting transformation genome from the normalized expression matrix, including major histocompatibility complex class I genes, antigen processing and transport genes, immunoproteasome genes, transcription activator genes, and interferon response genes; and an antigen-presenting transformation scoring module 23, which calculates a standardized score for each gene in the antigen-presenting transformation genome by subtracting the mean and standard deviation of the gene's expression value in the reference cell population from the mean and dividing by the standard deviation, and then calculating the arithmetic mean of the standardized scores of each gene in the antigen-presenting transformation genome of the same cell with equal weights to obtain the antigen-presenting transformation score of that cell. The target cell population is one or more cell types specified within the graft, and the reference cell population is... The reference dataset contains a collection of cells corresponding to the target cell type. The reference dataset includes a first immune state graft group and a second immune state graft group. The first immune state graft group and the second immune state graft group are combinations of allogeneic grafts and syngeneic grafts, or combinations of rejected grafts and non-rejected grafts, or combinations of active rejected grafts and tolerant grafts. The molecular state typing module 24 includes a threshold determined in advance on the reference dataset, independent of the data to be typed, to classify each cell in the target cell population into an antigen-presenting activated state or a quiescent state, thus obtaining a molecular state typing. The rejection-related molecular state index calculation module 25 includes a method to divide the number of cells in the antigen-presenting activated state by the total number of cells in the target cell population that have passed quality control, thus obtaining the proportion of antigen-presenting activated cells, which serves as the rejection-related molecular state index of the graft. The output module 26 includes a method to output the molecular state typing and the rejection-related molecular state index as the quantitative result of the antigen-presenting-related immune activation molecular state of the graft cells.

[0070] Another implementation of this application provides a computer device for analyzing the molecular state of antigen presentation-related immune activation in graft cells. The device includes a memory and a processor. The memory includes a program for storing data. The processor includes a program for executing the program stored in the memory to implement the following methods: a normalized expression matrix analysis step, including acquiring single-cell transcriptome expression data based on graft tissue, performing quality control and log-normalization on the data to obtain a normalized expression matrix; an antigen-presenting transformation genome identification step, including acquiring a pre-fixed antigen-presenting transformation genome from the normalized expression matrix, the antigen-presenting transformation genome including major histocompatibility complex class I molecular genes, antigen processing and transport genes, immunoproteasome genes, transcription activator genes, and interferon response genes; and an antigen-presenting transformation scoring step, including for each gene in the antigen-presenting transformation genome, subtracting the mean and then dividing the expression value of the gene in each cell of the target cell population from the mean and standard deviation of the gene's expression value in a reference cell population to obtain a standardized score, and calculating the arithmetic mean of the standardized scores of each gene in the antigen-presenting transformation genome of the same cell with equal weights to obtain the cell's score. The process includes: an antigen presentation conversion score; a target cell population (the specified cell type or type within the graft); a reference cell population (the set of cells corresponding to the target cell type in a reference dataset); a reference dataset containing a first immune status graft group and a second immune status graft group (the first and second immune status graft groups being combinations of allogeneic and syngeneic grafts, or combinations of rejected and non-rejected grafts, or combinations of active rejected grafts and tolerant grafts); a molecular status typing step, which uses a pre-determined threshold on a reference dataset independent of the data to be typed to classify cells in the target cell population into antigen-presenting active or quiescent states, thus obtaining a molecular status typing; a rejection-related molecular status index calculation step, which divides the number of cells in the antigen-presenting active state by the total number of cells in the target cell population that have passed quality control, to obtain the proportion of antigen-presenting active cells, which serves as the rejection-related molecular status index of the graft; and an output step, which outputs the molecular status typing and the rejection-related molecular status index as a quantitative result of the antigen-presenting-related immune activation molecular status of the graft cells. This quantitative result is used for graft immune status assessment for non-diagnostic purposes.

[0071] Another implementation of this application also provides a computer-readable storage medium, which includes a program executable by a processor to implement the following method: a normalized expression matrix analysis step, including acquiring single-cell transcriptome expression data based on graft tissue, performing quality control and log normalization on the data to obtain a normalized expression matrix; an antigen-presenting transformation genome identification step, including identifying the antigen-presenting transformation genome from the normalized expression matrix, the antigen-presenting transformation genome including major histocompatibility complex class I molecular genes, antigen processing and transport genes, immunoproteasome genes, transcription activator genes, and interferon response genes; an antigen-presenting transformation scoring step, including for each gene in the antigen-presenting transformation genome, subtracting the mean and then dividing the expression value of the gene in each cell of the target cell population from the mean and standard deviation of the gene's expression value in the reference cell population to obtain a standardized score, and calculating the arithmetic mean of the standardized scores of each gene in the antigen-presenting transformation genome of the same cell with equal weights to obtain the antigen-presenting transformation score of the cell; the target cell The target cell population is defined as one or more cell types within the graft. The reference cell population is a collection of cells corresponding to the target cell type in a reference dataset. The reference dataset includes a first immune state graft group and a second immune state graft group, which can be a combination of allogeneic and syngeneic grafts, a combination of rejected and non-rejected grafts, or a combination of active rejected grafts and tolerant grafts. The molecular state typing step includes classifying each cell in the target cell population into an antigen-presenting activated state or a quiescent state using a pre-determined threshold on the reference dataset independent of the data to be typed, thus obtaining the molecular state typing. The rejection-related molecular state index calculation step includes dividing the number of cells in the antigen-presenting activated state by the total number of cells in the target cell population that have passed quality control, thus obtaining the proportion of antigen-presenting activated cells, which serves as the rejection-related molecular state index of the graft. The output step includes outputting the molecular state typing and the rejection-related molecular state index as the quantitative result of the antigen-presenting-related immune activation molecular state of the graft cells.

[0072] Based on the method for analyzing the molecular state of antigen presentation-related immune activation in graft cells in this application, this application also proposes a detection kit for quantifying and genotyping the molecular state of graft rejection. The kit includes reagents for detecting the expression levels of various genes in the antigen-presenting transformed genome in the method for analyzing the molecular state of antigen presentation-related immune activation in graft cells in this application. The reagents include specific amplification primers, hybridization probes, or hybridization chips. The kit further includes an instruction manual, which records the rules for calculating the antigen presentation transformation score and the predetermined threshold.

[0073] The method for analyzing the molecular state of immune activation related to graft cell antigen presentation in this application has the following advantages compared with the prior art:

[0074] (1) In some implementation methods, qualitative antigens are transformed into reproducible continuous scores and discrete typing by using fixed gene combinations, explicit scoring formulas and Youden thresholds, so that the results of different samples and different batches can be directly compared.

[0075] (2) In some implementations, in a validation involving 663 islet cells, at the cellular level, the area under the curve for the antigen presentation conversion score to distinguish islet cells from allogeneic and syngeneic grafts was 0.896, with a 95% confidence interval of 0.860 to 0.927. At the Youden threshold of 0.063, the sensitivity was 0.833, the specificity was 0.882, and the typing accuracy was 0.870. The proportion of islet cells in the antigen presentation activation state was 83.3% in allogeneic grafts and 11.8% in syngeneic grafts.

[0076] (3) In some implementations, the above-mentioned fixed genome was frozen and applied to an independent mouse kidney transplant rejection single-cell cohort (GSE157292). The area under the curve that distinguishes between rejected and immature transplant cells was 0.969, with a 95% confidence interval of 0.966 to 0.972, indicating that the genome has cross-organ mobility in the molecular direction of transplant rejection-related antigen presentation activation.

[0077] (4) In some implementations, the score can distinguish between allogeneic and syngeneic grafts in all parenchymal and stromal cell groups of the graft, including islets, endothelial cells, myeloid cells, and stromal cells, with an area under the curve between 0.79 and 0.90. However, it has no distinguishing ability in infiltrative lymphocytes, with an area under the curve of only 0.499. This indicates that it specifically reflects the antigen presentation and transformation of the graft target cells rather than general immune activation, supporting the unconventional technical effect of genome fixation.

[0078] (5) In some implementations, the fixed genome is further used via homology mapping for an independent human kidney transplant biopsy dataset GSE145927, which contains real clinical rejection. The area under the curve for distinguishing between clinically rejected and non-rejected cells is 0.715, and it reaches 0.949, 0.921 and 0.844 in intercalated cells, macrophages and endothelial cells, respectively, indicating that the molecular state characterization ability of the genome can be transferred across species to real clinical transplant rejection in humans.

[0079] (6) In some implementations, the fixed genome is homologously mapped to two independent human kidney transplant biopsy datasets GSE145927 and GSE228300, which contain real clinical rejection. The area under the curve for distinguishing between clinical rejection and non-rejection cells is 0.715 and 0.777, respectively, indicating that the molecular state representation ability of the genome can be transferred across species to real clinical transplant rejection in humans. It is consistent with the comparison with mouse rejection and tolerated transplants, indicating that the score represents the antigen-presenting activation state rather than clinical rejection subtyping.

[0080] The inventiveness of this application is mainly reflected in the following aspects: (1) The antigen presentation transformation genome consists of 18 genes from five functional modules: major histocompatibility complex class I molecules, antigen processing and transport, immunoproteasome, transcription activator and interferon response. As a whole, it is pre-fixed, and no gene addition, deletion or re-screening is performed on all samples to be classified and external validation data. This is fundamentally different from the conventional signature scoring method that screens genes or fits gene weights based on the data labels to be evaluated; (2) The antigen presentation transformation score of this application adopts the explicit equal weight arithmetic mean formula of reference cell population standardization, without introducing any bias weights; the standardization parameters are pre-calculated and fixed on the reference dataset, and the test samples are standardized only with these fixed parameters, thereby achieving the comparability of the score between different samples, different batches, different organs and different species. This is a feature that existing signature scoring methods do not have; (3) The threshold of this application is pre-determined and frozen only on the reference dataset according to the principle of maximizing the Yoden index. The test samples do not participate in the threshold selection, thereby ensuring the reproducibility of the classification results; (4) Unconventional techniques Surgical efficacy: In infiltrative lymphocytes, the area under the curve (AUC) of this application's score was only 0.499, lacking discriminative power, while in graft parenchyma and stromal cells (islets, endothelium, myeloid, and stroma), the AUC was between 0.79 and 0.90. This contrast indicates that the fixed genome-specific response used in this application reflects the molecular process of antigen transfer in graft target cells, rather than the general immune activation or pan-IFN response within the graft. This is an unconventional technical effect that is difficult to predict with signature scores composed of arbitrary sets of immune-related genes. 5) Cross-organ and cross-species transferability: On the independent mouse kidney transplant rejection cohort GSE157292, the area under the curve (AUC) of the score proposed in this application reached 0.969; in two independent human kidney transplant biopsy datasets GSE145927 and GSE228300, the AUCs were 0.715 and 0.777, respectively. These results indicate that the score proposed in this application has cross-organ and cross-species transferability in characterizing the molecular direction of transplant rejection-related antigen presentation activation. The aforementioned unconventional technical effects together constitute the core of the inventiveness of this application. Although there are reports in the prior art of single or a few MHC-I molecules or interferon-stimulated genes being associated with transplant rejection, the method of pre-fixing a total of 18 genes from five functional modules as a whole, using equal-weighted scoring with fixed parameters from a reference cell population, and combining it with a frozen threshold to achieve comparable continuous scoring and discrete typing across samples, batches, organs, and species has not been disclosed in any prior art, nor does it present any technical inspiration in the prior art.The inventive aspect of this application lies in establishing a standardized quantitative system based on antigen presentation transformation state. This system transforms the continuous transformation process of graft cells from a quiescent state to an antigen presentation activation state into reproducible molecular state parameters through functional modular gene sets, reference cell population normalized scores, and independent threshold typing.

[0081] The present application will be further described in detail below through specific embodiments. The following embodiments are only for further illustration of the present application and should not be construed as limiting the present application.

[0082] Unless otherwise specified, the technical means used in the embodiments are conventional means well known to those skilled in the art.

[0083] Example 1: Islet Cell Antigen Presentation Transformation Scoring and Typing

[0084] The steps for normalized expression matrix analysis include: acquiring single-cell transcriptome expression data based on graft tissue, performing quality control and log normalization on the data, and obtaining the normalized expression matrix.

[0085] This example uses single-cell transcriptome data from mouse islet syngeneic and allogeneic transplants on day 7 post-transplantation. For mouse islet syngeneic transplantation: 6-8 week old SPF-grade male mice were used as both donor and recipient. In the syngeneic group, both donor and recipient were C57BL / 6 inbred mice, while in the allogeneic group, the donor was BALB / c inbred mice and the recipient was still C57BL / 6 mice. Both donor and recipient mice were purchased from the Guangdong Provincial Medical Laboratory Animal Center. The recipient was housed in a barrier environment at the Central Laboratory of Shenzhen Longhua District Central Hospital prior to surgery. The diabetes model was induced by a single intraperitoneal injection of streptozotocin (STZ, Sigma-Aldrich) prepared in 0.5 mol / L citrate buffer at 250 mg / kg, with a confirmed diagnosis of diabetes based on a non-fasting blood glucose level >16.8 mmol / L for 5 consecutive days. The transplantation surgery used a subcapsular renal transplantation method, and the recipient was routinely housed post-operatively. Allogeneic transplantation of mouse islets: 500 allogeneic BALB / c islets were implanted subcapsularly into streptozotocin-induced diabetic C57BL / 6 recipient mice (n=5); simultaneously, 500 syngeneic C57BL / 6 islets were implanted into syngeneic diabetic C57BL / 6 recipient mice as controls (n=5). Mice were sacrificed on day 7 post-surgery, and the renal capsule tissue attached to the transplant was harvested. The tissue was digested in RPMI 1640 at 37°C for approximately 10 minutes with 0.01% (w / v) LiberaseTH combined with 100 U / mL DNase I. The tissue was then filtered through a 40 μm cell sieve, washed with 1×PBS containing 2 mmol / L EDTA and 0.5% BSA, centrifuged at 200×g for 5 minutes, and resuspended in RPMI 1640 ± 10%. FBS (preserved on ice) showed >85% viability as determined by trypan blue (TC20 automated cell counter, Bio-Rad); single-cell library construction was performed using the 10× Genomics Chromium Single Cell 3' Reagent Kits (Single Cell B Chip v3, PN-1000153) platform at Tianjin Novogene, using an Illumina NovaSeq 6000 sequencer, with a target sequencing depth of at least 50,000 raw reads per cell. Surgery and postoperative observation were performed as described above; mice were sacrificed and tissue samples were collected on day 7 postoperatively.

[0086] The study retained 18,130 cells, including 6,948 syngeneic grafts and 11,182 allogeneic grafts. Clustering yielded 23 clusters, 5 principal cell types, and 10 cell subtypes. Single-cell atlases and target cell identification were performed as follows: Figure 3 As shown, the expression of classic marker genes for each chief cell type is as follows: Figure 4As shown; 663 pancreatic islet cells were annotated, including 501 syngeneic grafts and 162 allogeneic grafts. Expression data were quality-controlled to obtain a normalized expression matrix. Figure 3 It is used to demonstrate the basis of data acquisition and target cell identification, and is not used as a basis for determining molecular state typing. Figure 4 This is used to illustrate the identification criteria for each cell type and subtype. Quality control was performed sequentially using Seurat R software version 3.1.5: cells with nFeature_RNA < 200 or > 4500 were removed; cells with percentage.mt > 15% were removed; only the unique gene expressed in at least three cells was retained; and global scaling and log-normalization were performed using the LogNormalize method (scale.factor = 10000). After normalization, the top 2000 hypervariable genes were selected for principal component analysis, with a principal component count of 75. After dimensionality reduction using UMAP, FindClusters clustering was performed with a resolution of 0.6.

[0087] Confirmation of antigen-presenting transformation genome steps: In this embodiment, 18 genes are pre-fixed to form an antigen-presenting transformation genome, which includes major histocompatibility complex class I molecular genes, antigen processing and transport genes, immunoproteasome genes, transcription activator genes, and interferon response genes.

[0088] The antigen-presenting transformation genome was determined based on the following criteria: (a) Major histocompatibility complex class I molecular module: Based on the core role of mouse MHC class I molecules in cell surface antigen peptide presentation, five genes were identified from B2m and classical and non-classical MHC class I molecular genes: B2m, H2-K1, H2-D1, H2-Q7, and H2-T23; (b) Antigen processing and transport module: Based on the TAP complex and its cofactors necessary for the loading and transport of antigen peptides in the endoplasmic reticulum lumen, four genes were identified: Tap1, Tap2, Tapbp, and Tapbpl. (c) Immunoplasminosome module: Based on the role of alternative immunoproteasomes in the shaping of MHC-I antigen peptide libraries, three genes, Psmb8, Psmb9, and Psmb10, were identified; (d) Transcription activator module: Based on the core transcriptional regulatory roles of NLRC5 and IRF1 in the MHC-I pathway and interferon response, two genes, Nlrc5 and Irf1, were identified; (e) Interferon response module: Based on representative ISGs that are synergistically upregulated with antigen presentation under interferon stimulation, four genes, Stat1, Gbp2, Isg15, and Ifit3, were identified. In this embodiment, the above five modules, totaling 18 genes, are pre-fixed as the entire antigen-presenting transformation genome, and no further gene additions, deletions, or re-screening are performed on any samples to be genotyped or external validation data.

[0089] The expression heatmap of the 18 genes in allogeneic and syngeneic pancreatic islet cells in this example is shown below. Figure 5 As shown. The above 18 genes have been reported in the art to be associated with antigen presentation or transplant rejection; this application solidifies them into a specific combination of 18 genes and combines them with a reference cell population-standardized explicit equal-weighted scoring formula and a freezing threshold for the quantification and typing of the molecular state of graft cells.

[0090] The antigen presentation conversion scoring steps include: for each gene in the antigen presentation conversion genome, subtracting the mean and standard deviation of the gene's expression value in the reference cell population from the expression value of the gene in each cell of the target cell population and dividing by the standard deviation to obtain a standardized score; and calculating the arithmetic mean of the standardized scores of each gene in the antigen presentation conversion genome of the same cell with equal weights to obtain the antigen presentation conversion score of the cell; the target cell population is one or more cell types specified within the graft, and the reference cell population is a set of cells of the corresponding target cell type in the reference dataset. The reference dataset includes a first immune status graft group and a second immune status graft group, which are combinations of allogeneic grafts and syngeneic grafts, or combinations of rejected grafts and non-rejected grafts, or combinations of active rejected grafts and tolerant grafts.

[0091] Specifically, in this example, each gene is standardized within the reference cell population according to Formula 1, and the average of the standardized scores for each gene in the same cell is calculated to obtain the antigen presentation conversion score for that islet cell. The median score for islet cells from allogeneic grafts is 0.552, while that for syngeneic grafts is -0.276. At the cellular level, the area under the curve (AUC) for the score distinguishing the two classes is 0.896, with a 95% confidence interval of 0.860 to 0.927. Its distribution and threshold genotyping are shown below. Figure 6 As shown. It should be noted that this dataset consists of one homologous library and one allogeneic library, without single mouse identification. The statistics above are cellular level comparisons, and the cells within the same graft are not independent biological replicates.

[0092] Formula 1:

[0093]

[0094] In Formula 1, xij represents the log-normalized expression value of gene j in islet cell i, μj and σj represent the mean and standard deviation of gene j's expression value in the reference cell population (in Example 1, the reference cell population is a combined population of 663 islet cells from two groups), zij represents the intra-population standardized score of gene j in islet cell i, G represents the antigen-presenting transformation genome, |G| represents the number of genes in the genome, and Si represents the antigen-presenting transformation score of islet cell i. This formula first standardizes each gene within the reference cell population to eliminate differences in dimensions and expression amplitude between genes, and then calculates the arithmetic mean of the standardized scores of each gene within the genome of the same cell, thereby obtaining a continuous score that reflects the antigen-presenting transformation intensity of the cell and can be compared between different samples and batches.

[0095] Molecular state typing determination steps include classifying each cell in the target cell population into an antigen-presenting activated state or a quiescent state using a pre-determined threshold on a reference dataset independent of the data to be typed, thus obtaining the molecular state typing.

[0096] Specifically, in this example, the threshold value of 0.063, which maximizes the Youden index for distinguishing allogeneic and homologous individuals, is taken. The ROC curve and the Youden operating point are shown below. Figure 7 The Youden index was 0.716, the sensitivity was 0.833, and the specificity was 0.882. Pancreatic islet cells with scores not lower than the threshold were classified as antigen-presenting activated, otherwise they were classified as quiescent. The cell-level typing accuracy was 0.870 at this threshold.

[0097] The steps for calculating the rejection-related molecular state index include dividing the number of cells in the antigen-presenting activated state by the total number of cells in the target cell population that have passed quality control, to obtain the proportion of antigen-presenting activated cells, which serves as the rejection-related molecular state index of the graft.

[0098] Specifically, in this case, the proportion of activated pancreatic islet cells was statistically analyzed, with 83.3% for allogeneic grafts and 11.8% for syngeneic grafts. This proportion was used as an index of rejection-related molecular states.

[0099] Output steps include outputting molecular state typing and rejection-related molecular state indices, which serve as quantitative results of the molecular state of graft cell antigen presentation-related immune activation.

[0100] Specifically, this example outputs the molecular state classification of each pancreatic islet cell and the molecular state index related to graft rejection.

[0101] As an alternative implementation, the threshold for the molecular state typing step can also be taken from a score value that has been pre-calibrated and solidified on a reference dataset; the standardization of the antigen presentation conversion scoring step can also be carried out within a larger cell set containing pancreatic islet cells to enhance cross-sample consistency.

[0102] This method can be implemented using a parting device, such as... Figure 2 As shown, the device includes a normalized expression matrix analysis module 21, an antigen-presenting transformation genome confirmation module 22, an antigen-presenting transformation scoring module 23, a molecular state typing determination module 24, a rejection-related molecular state index calculation module 25, and an output module 26, which sequentially execute the normalized expression matrix analysis step, the antigen-presenting transformation genome confirmation step, the antigen-presenting transformation scoring step, the molecular state typing determination step, the rejection-related molecular state index calculation step, and the output step in this example. This method can also be implemented by a computer program and stored in a computer-readable storage medium for execution by a processor.

[0103] Transcriptome expression data can be single-cell transcriptome sequencing data or batch transcriptome sequencing data obtained by cell type deconvolution to obtain the expression components of pancreatic islet cells and other cell types. The pancreatic islet graft samples to be tested can be from mice, pigs, non-human primates or humans, and when crossing species, the genome is constructed using corresponding homologous genes.

[0104] The expression levels of each gene in the antigen-presenting transformed genome can be detected by a detection kit containing gene-specific amplification primers, hybridization probes, or hybridization chips. Given that this application discloses the specific genes contained in the aforementioned genome, those skilled in the art can design corresponding amplification primers, hybridization probes, or hybridization chips for these genes based on existing technologies to detect the expression levels of each gene; therefore, the detection kit can be developed using conventional methods in the field, thereby enabling the in vitro detection of the molecular characteristics upon which this method relies.

[0105] Example 2: Validation of external migration of frozen genomes in an independent transplant rejection cohort

[0106] To verify the transferability of the antigen-presenting transformed genome, the 18-gene genome identified in Example 1 was frozen without any gene additions, deletions, or reselection and applied to an independent mouse transplant rejection single-cell transcriptome cohort GSE157292 (https: / / www.ncbi.nlm.nih.gov / geo / query / acc.cgi?acc=GSE157292). This cohort included three types of grafts: naive, rejection, and tolerance, with 4963, 14944, and 12506 cells retained after quality control, respectively; all 18 genes were present in this cohort. The single-cell UMAP atlas of this external cohort is shown below. Figure 12 .

[0107] For each cell, an antigen presentation conversion score was calculated within the cohort according to Formula 1 of Example 1. The median score for rejected graft cells was 0.208, and for immature grafts it was -0.935; at the cellular level, the area under the curve (AUC) distinguishing rejected from immature graft cells was 0.969, with a 95% confidence interval of 0.966 to 0.972. Figure 8 As shown, the area under the curve (AUC) for distinguishing between rejected and tolerant graft cells was 0.621, which is lower than the distinguishing factor between rejection and immaturity. This is consistent with the biological expectation of partial immune activation in tolerant grafts, indicating that this genome has relative specificity for active rejection.

[0108] It should be noted that this cohort consists of kidney grafts, which belong to a different tissue than the islet grafts in this application. The above results verify the transferability of the genome to the molecular direction of transplant rejection-related antigen presentation activation, rather than the typing performance of the islet grafts. Since the area under the curve is not sensitive to monotonic changes in the score, this transferability conclusion does not depend on the specific standardization method. The threshold of 0.063 in Example 1 is calibrated within the islet cell population and should not be directly transferred to different tissues or different standardization calibers. When applied across cohorts, the threshold should be recalibrated.

[0109] Example 3: Expansion of the target cell population and cell number

[0110] To illustrate that the method of this application is not limited to pancreatic islet cells and that the number of cells analyzed is not limited to only 663, the antigen-presenting transformation genome determined in Example 1 was frozen, and antigen-presenting transformation scores were calculated in each principal cell population of 18,130 cells in the same dataset to compare the distinguishability between allogeneic and syngeneic graft cells.

[0111] The results are as follows Figure 9 As shown: the area under the curve for distinguishing pancreatic islet cells was 0.896 (663 cells), endothelial cells were 0.818 (549 cells), myeloid cells were 0.797 (4954 cells), stromal cells were 0.790 (4167 cells), and the total area under the curve for all 18130 cells in the graft was 0.747; while lymphocytes were 0.499 and showed no distinguishing value.

[0112] The above results indicate that: First, the method of this application can act on various graft parenchymal and stromal cells, such as islet cells, endothelial cells, myeloid cells, and interstitial cells, as well as all cells within the graft. The calculated number of cells can reach thousands to tens of thousands, rather than being limited to 663 islet cells. Second, the score does not show differentiation in infiltrative lymphocytes, indicating that it reflects the specific molecular process of antigen presentation transformation of graft target cells, rather than the general immune activation within the graft, thus demonstrating the specificity of a fixed genome. Third, islet cells have the highest differentiation due to the most significant antigen presentation transformation, making them the preferred target cell population.

[0113] Furthermore, in an independent external cohort of GSE157292 mouse kidney transplants (https: / / www.ncbi.nlm.nih.gov / geo / query / acc.cgi?acc=GSE157292), after crudely classifying cells into epithelial / tubular, interstitial, endothelial, myeloid, B, and T / NK types based on marker genes, the areas under the curve (AUC) of the frozen antigen-presenting transforming genome in distinguishing between rejected and immature grafts in each type were 0.991 (epithelial / tubular), 0.979 (interstitial cells), 0.956 (B cells), 0.931 (myeloid cells), 0.888 (endothelial cells), and 0.844 (T / NK cells), respectively. Figure 9 As shown in Figure B, it can be seen that the method of this application is not limited to pancreatic islet cells, and its applicability to multiple cell types is also valid in independent external datasets.

[0114] Conclusion and summary:

[0115] Statistical units and data independence: The area under the curve in Examples 1 to 3 is calculated with cells as the observation unit; the internal data comes from one syngeneic graft library and one allogeneic graft library. Cells are not equivalent to independent biological replicates. Therefore, the confidence interval reflects the uncertainty of differentiation at the cell level and does not represent validation at the donor, recipient or clinical sample level.

[0116] Fixed genome and no training weights: The gene set, expression direction and equal weighting method of the antigen presentation transformation genome are predetermined before evaluation, and genes are not screened or fitted with weights based on the category labels of the data to be evaluated; the antigen presentation transformation score assigns the same weight to each gene.

[0117] Threshold freezing and leakage prevention: The threshold is determined only in the reference dataset according to the Yoden index, and the samples to be genotyped do not participate in the threshold selection; the data to be genotyped adopts the same quality control, normalization and standardization process as the reference dataset.

[0118] Cell population effect boundary: In the internal dataset, the area under the curve for pancreatic islets, endothelial cells, myeloid cells and stromal cells are 0.896, 0.818, 0.797 and 0.790, respectively, while the area under the curve for all cells is 0.747 and for lymphocytes is 0.499. Therefore, the distinguishing ability of this method is cell type dependent, and lymphocytes are not considered as the preferred target cell population.

[0119] External migration boundary: The area under the curve for the frozen genome in the external dataset GSE157292 was 0.969 for distinguishing between rejection and immature graft cells and 0.621 for distinguishing between rejection and tolerance. These results support the ability of the genome to migrate and characterize the molecular states of antigen presentation-related immune activation in different transplanted tissues, but do not constitute a validation for clinical diagnosis, tolerance assessment or outcome prediction of islet transplantation.

[0120] Application Boundaries: The output of this application is cellular and molecular state parameters reflected by in vitro transcriptome data, which in themselves do not constitute a diagnosis, prognostic prediction, or treatment decision regarding individual graft rejection, tolerance, graft function, or clinical outcome.

[0121] Reference cell population and standardization criteria: The mean and standard deviation used for standardization of each gene were taken from the reference cell population. In Example 1, this involved 663 pancreatic islet cells from two groups, and the data to be classified were standardized using this reference parameter. Since standardization was performed on the merged reference cell population, and the scores at the graft level or group level were taken from the mean of their respective cell subsets, they were generally not zero.

[0122] Example 4: Validation of migration of rejection in real-world human transplantation

[0123] To examine the migration characterization ability of genomes under real-world clinical rejection scenarios in human donors, an independent human kidney transplant biopsy single-cell dataset, GSE145927 (https: / / www.ncbi.nlm.nih.gov / geo / query / acc.cgi?acc=GSE145927), was used. This dataset contains 81,139 cells from both clinical rejection (rejn) and non-rejection (no_rejn) biopsies. Frozen antigen-presenting transformed genomes were mapped according to homology to human genes HLA-A, HLA-B, HLA-C, HLA-E, HLA-F, B2M, TAP1, TAP2, TAPBP, TAPBPL, PSMB8, PSMB9, PSMB10, NLRC5, IRF1, STAT1, GBP1, GBP2, ISG15, and IFIT3. Antigen presentation transformation scores were calculated within each cell population using Formula 1. The single-cell UMAP atlas of this external cohort is shown below. Figure 13 Figures A and B.

[0124] The results are as follows Figure 10As shown: at the entire cell level, the area under the curve (AUC) for distinguishing clinically rejected from non-rejected biopsy cells was 0.715, with a 95% confidence interval of 0.711 to 0.718. Among cell types, the AUCs for intercalated cells, macrophages, and endothelial cells were 0.949, 0.921, and 0.844, respectively, while the AUC for lymphocytes was 0.677. The AUCs for large renal tubular epithelial cells such as proximal tubules and loops of Henle were lower, ranging from only 0.525 to 0.558.

[0125] The above human-derived results indicate that the fixed genome, after homologous mapping to humans, still retains its ability to migrate and characterize in real clinical rejection of human kidney transplantation, and exhibits cell type dependence consistent with mouse results, showing the strongest performance in antigen-presenting intercalated cells, macrophages, and endothelial cells. It should be noted that this cohort consists of human kidney transplant biopsies, which are from different tissues than islet transplants; the human major histocompatibility complex class I molecules are homologous mappings to those in mice, not identical name correspondences, for example, mapping to HLA-A / B / C / E / F, etc.; the area under the correlation curve is measured in cells; therefore, these results represent evidence of molecular state migration across species and tissues, rather than validation of clinical diagnosis, tolerance assessment, or outcome prediction for islet transplantation.

[0126] Example 5: Migration Verification of the Second Human Source Cohort

[0127] To further examine the migration characterization ability under human-derived, real-world clinical rejection scenarios, another independent human kidney transplant biopsy dataset, GSE228300 (https: / / www.ncbi.nlm.nih.gov / geo / query / acc.cgi?acc=GSE228300), was used. This dataset includes three types of biopsies: T-cell-mediated rejection (TCMR, Banff IIa), borderline alteration, and non-rejection. After mapping the frozen antigen-presenting transformed genome to human genes according to homology, antigen-presenting transformation scores were calculated within each cell population using Formula 1. The single-cell UMAP atlas of this external cohort is shown below. Figure 13 Diagrams C and D are shown. The mapped human genes are the same as in Example 4.

[0128] The results are as follows Figure 11 As shown: at the entire cellular level, the area under the curve (AUC) for distinguishing T cell-mediated rejection from non-rejection biopsy cells was 0.777, with a 95% confidence interval of 0.758 to 0.794, and 0.811 after including borderline changes. Among cell types, the AUCs for collecting ducts, intercalated cells, proximal tubules, loop of Henle / distal tubules, and podocytes were 0.934, 0.814, 0.803, 0.778, and 0.719, respectively, and were strongest in parenchymal cells such as renal tubules and collecting systems.

[0129] In conjunction with Example 4, after the fixed genome was mapped to human homology, it was able to distinguish between clinically rejected and non-rejected biopsy cells in two independent human kidney transplant biopsy cohorts, GSE145927 and GSE228300, with areas under the curve of 0.715 and 0.777, respectively. Consistent with the results in mice, the discrimination was cell type dependent, with the strongest performance in parenchymal cells such as renal tubules, collecting system, and endothelium. This reflects that the score characterizes the antigen-presenting activation state of graft cells, rather than the diagnostic classification of clinical rejection type.

[0130] It should be noted that the human results of Examples 4 and 5 are used to illustrate the cross-species migration characterization ability of antigen-presenting transformed genomes. To apply this method to other species such as humans, it is only necessary to map it to the genes of the corresponding species according to the conventional homology relationship.

[0131] The above description, in conjunction with specific embodiments, provides a further detailed explanation of this application and should not be construed as limiting the specific implementation of this application to these descriptions. Those skilled in the art to which this application pertains can make several simple deductions or substitutions without departing from the concept of this application.

Claims

1. A method for analyzing the molecular state of immune activation related to antigen presentation in graft cells, characterized in that: Includes the following steps, The normalized expression matrix analysis step (11) includes obtaining single-cell transcriptome expression data based on graft tissue, performing quality control and log normalization on the data, and obtaining the normalized expression matrix. The step (12) of confirming the antigen-presenting transformation genome includes identifying the antigen-presenting transformation genome from the normalized expression matrix, wherein the antigen-presenting transformation genome includes major histocompatibility complex class I molecular genes, antigen processing and transport genes, immunoproteasome genes, transcription activator genes and interferon response genes; The antigen presentation conversion scoring step (13) includes, for each gene in the antigen presentation conversion genome, subtracting the mean and standard deviation of the gene's expression value in the reference cell population from the expression value of the gene in each cell of the target cell population and dividing by the standard deviation to obtain a standardized score; and calculating the arithmetic mean of the standardized scores of each gene in the antigen presentation conversion genome of the same cell with equal weights to obtain the antigen presentation conversion score of the cell; the target cell population is one or more cell types specified in the graft; The reference cell population is a collection of cells corresponding to the target cell type in the reference dataset. The reference dataset includes a first immune status graft group and a second immune status graft group. The first immune status graft group and the second immune status graft group are combinations of allogeneic grafts and syngeneic grafts, or combinations of rejected grafts and non-rejected grafts, or combinations of active rejected grafts and tolerant grafts. The molecular state typing determination step (14) includes classifying each cell in the target cell population into an antigen-presenting activated state or a quiescent state using a threshold predetermined on a reference dataset independent of the data to be typed, thereby obtaining a molecular state typing. The rejection-related molecular state index calculation step (15) includes dividing the number of cells in the antigen-presenting activated state by the total number of cells in the target cell population that have passed quality control to obtain the proportion of antigen-presenting activated cells, which is used as the rejection-related molecular state index of the graft. The output step (16) includes outputting the molecular state typing and the rejection-related molecular state index as the quantitative result of the graft cell antigen presentation-related immune activation molecular state.

2. The method according to claim 1, characterized in that: In the step (12) of confirming antigen presentation and transforming the genome, When the test subject is a mouse, the major histocompatibility complex class I molecular genes include B2m, H2-K1, H2-D1, H2-Q7 and H2-T23, the antigen processing and transport genes include Tap1, Tap2, Tapbp and Tapbpl, the immunoproteasome genes include Psmb8, Psmb9 and Psmb10, the transcription activator genes include Nlrc5 and Irf1, and the interferon response genes include Stat1, Gbp2, Isg15 and Ifit3; When the subject of the test is a human, the antigen-presenting transformation genome includes HLA-A, HLA-B, HLA-C, HLA-E, HLA-F, B2M, TAP1, TAP2, TAPBP, TAPBPL, PSMB8, PSMB9, PSMB10, NLRC5, IRF1, STAT1, GBP1, GBP2, ISG15, and IFIT3; Optionally, the target cell population may be one or more of pancreatic islet cells, endothelial cells, myeloid cells, and interstitial cells, but may not include lymphocytes.

3. The method according to claim 1, characterized in that: In the antigen presentation conversion scoring step (13), the antigen presentation conversion score is calculated according to the following formula: for cell i and gene j in the target cell population, the standardized score is the expression value of gene j in cell i minus the mean expression value of gene j in the reference cell population, divided by the standard deviation of gene j expression in the reference cell population; the antigen presentation conversion score is the arithmetic mean of the standardized scores of all genes in cell i within the antigen presentation conversion genome.

4. The method according to any one of claims 1-3, characterized in that: In the molecular state classification judgment step (14), the predetermined threshold is the score value that maximizes the Youden index of the antigen presentation transformation score in distinguishing between the two types of cells, the first immune state graft group and the second immune state graft group, on the reference dataset.

5. An apparatus for analyzing the molecular state of immune activation associated with graft cell antigen presentation, characterized in that: It includes a normalized expression matrix analysis module (21), a genome confirmation module for antigen presentation transformation (22), an antigen presentation transformation scoring module (23), a molecular state typing judgment module (24), a rejection-related molecular state index calculation module (25), and an output module (26). The normalized expression matrix analysis module (21) includes a method for acquiring single-cell transcriptome expression data based on graft tissue, performing quality control and log normalization on the data, and obtaining a normalized expression matrix. The confirmed antigen-presenting transformation genome module (22) includes a method for determining the antigen-presenting transformation genome from the normalized expression matrix, the antigen-presenting transformation genome including major histocompatibility complex class I molecular genes, antigen processing and transport genes, immunoproteasome genes, transcription activator genes and interferon response genes; The antigen presentation conversion scoring module (23) includes a standardized score for each gene in the antigen presentation conversion genome. This score is calculated by subtracting the mean and standard deviation of the gene's expression value in the reference cell population from the expression value of the gene in each cell of the target cell population, and then dividing by the standard deviation. The standardized scores of all genes in the same cell within the antigen presentation conversion genome are then arithmetically averaged with equal weights to obtain the antigen presentation conversion score for that cell. The target cell population is one or more cell types specified within the graft. The reference cell population is a collection of cells corresponding to the target cell type in the reference dataset. The reference dataset includes a first immune status graft group and a second immune status graft group. The first immune status graft group and the second immune status graft group are combinations of allogeneic grafts and syngeneic grafts, or combinations of rejected grafts and non-rejected grafts, or combinations of active rejected grafts and tolerant grafts. The molecular state typing module (24) includes a threshold determined in advance on a reference dataset independent of the data to be typed, which is used to classify each cell in the target cell population into an antigen-presenting activated state or a quiescent state to obtain a molecular state typing. The rejection-related molecular state index calculation module (25) includes a method for dividing the number of cells in the antigen-presenting activated state by the total number of cells in the target cell population that have passed quality control to obtain the proportion of antigen-presenting activated cells, which is used as the rejection-related molecular state index of the graft. The output module (26) includes a function to output the molecular state typing and the rejection-related molecular state index as a quantitative result of the graft cell antigen presentation-related immune activation molecular state.

6. The apparatus according to claim 5, characterized in that: In the confirmed antigen-presenting transformation genome module (22), When the test subject is a mouse, the major histocompatibility complex class I molecular genes include B2m, H2-K1, H2-D1, H2-Q7 and H2-T23, the antigen processing and transport genes include Tap1, Tap2, Tapbp and Tapbpl, the immunoproteasome genes include Psmb8, Psmb9 and Psmb10, the transcription activator genes include Nlrc5 and Irf1, and the interferon response genes include Stat1, Gbp2, Isg15 and Ifit3; When the test subject is a mouse, the major histocompatibility complex class I molecular genes include B2m, H2-K1, H2-D1, H2-Q7 and H2-T23, the antigen processing and transport genes include Tap1, Tap2, Tapbp and Tapbpl, the immunoproteasome genes include Psmb8, Psmb9 and Psmb10, the transcription activator genes include Nlrc5 and Irf1, and the interferon response genes include Stat1, Gbp2, Isg15 and Ifit3; When the subject of the test is a human, the antigen-presenting transformation genome includes HLA-A, HLA-B, HLA-C, HLA-E, HLA-F, B2M, TAP1, TAP2, TAPBP, TAPBPL, PSMB8, PSMB9, PSMB10, NLRC5, IRF1, STAT1, GBP1, GBP2, ISG15, and IFIT3; Optionally, the target cell population is one or more of pancreatic islet cells, endothelial cells, myeloid cells and interstitial cells, but does not include lymphocytes; Optionally, in the antigen presentation conversion scoring module (23), the antigen presentation conversion score is calculated according to the following formula: for cell i and gene j in the target cell population, its standardized score is the expression value of gene j in cell i minus the mean expression value of gene j in the reference cell population, divided by the standard deviation of gene j expression in the reference cell population; the antigen presentation conversion score is the arithmetic mean of the standardized scores of all genes in cell i within the antigen presentation conversion genome; Optionally, in the molecular state typing module (24), the predetermined threshold is the score value that maximizes the Youden index for distinguishing between the two types of cells, the first immune state graft group and the second immune state graft group, on the reference dataset.

7. A computer-readable storage medium, characterized in that: Includes a program that can be executed by a processor to implement the method of any one of claims 1-4.

8. A detection kit for quantifying and typing molecular states related to graft rejection, characterized in that: The method includes reagents for detecting the expression levels of each gene in the antigen-presenting transformed genome as described in any one of claims 1-4; the reagents include one or more of specific amplification primers, hybridization probes, or hybridization chips.

9. The application of antigen-presenting transformative genomics in the preparation of a detection kit for quantifying and genotyping molecular states related to graft rejection, characterized in that: The antigen-presenting transformation genome includes major histocompatibility complex class I molecular genes, antigen processing and transport genes, immunoproteasome genes, transcription activator genes, and interferon response genes; among which... When the test subject is a mouse, the major histocompatibility complex class I molecular genes include B2m, H2-K1, H2-D1, H2-Q7 and H2-T23, the antigen processing and transport genes include Tap1, Tap2, Tapbp and Tapbpl, the immunoproteasome genes include Psmb8, Psmb9 and Psmb10, the transcription activator genes include Nlrc5 and Irf1, and the interferon response genes include Stat1, Gbp2, Isg15 and Ifit3; When the subject of the test is a human, the antigen-presenting transformation genome is determined by homology mapping, including HLA-A, HLA-B, HLA-C, HLA-E, HLA-F, B2M, TAP1, TAP2, TAPBP, TAPBPL, PSMB8, PSMB9, PSMB10, NLRC5, IRF1, STAT1, GBP1, GBP2, ISG15, and IFIT3.

10. The application according to claim 9, characterized in that: The detection kit includes reagents for detecting the expression levels of each gene in the antigen-presenting transformed genome; the reagents include one or more of specific amplification primers, hybridization probes, or hybridization chips.

Images