Application of IgA < + > plasma cell marker in HNSCC prognosis evaluation

Abstract

The invention relates to the technical field of prognosis evaluation of cancers, in particular to application of an IgA < + > plasma cell marker in prognosis evaluation of HNSCC. The expression level of the IgA + plasma cell marker can reflect the prognosis condition of the HNSCC, high expression of S100A10, MANF, SPCS3, SSR4, ALG5, KDELR1, CYBA or SEC61G is in positive correlation with poor prognosis of the HNSCC (poor prognosis), and high expression of JCHAIN, GNG7, SSR3, CD38, IRF4 or FAM30A is in negative correlation with poor prognosis of the HNSCC (good prognosis). Therefore, an HNSCC prognosis risk prediction model is established, and the problem that HNSCC prognosis evaluation is inaccurate or low in efficiency is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of cancer prognosis assessment, and in particular to IgA + Application of plasma cell markers in prognostic evaluation of HNSCC. Background Art

[0002] Head and neck squamous cell carcinoma (HNSCC) is a malignant tumor that originates from epithelial cells of the oral cavity, pharynx, and larynx. It is aggressive and heterogeneous and is therefore associated with a poor overall prognosis. Despite advances in surgery, radiotherapy, chemotherapy, targeted therapy, and immunotherapy for HNSCC, the 5-year survival rate of HNSCC remains low. Given the poor prognosis and the complexity of treatment decisions, it is critical to achieve accurate prognostic prediction. This not only helps to accurately stratify patients, but also supports the optimization of clinical decisions and assists patients in making informed choices about their end-of-life care. Therefore, in order to improve patient prognosis, it is crucial to improve patient stratification, and identifying effective biomarkers is a key strategy to achieve this goal. Summary of the invention

[0003] Based on this, the present invention provides IgA + The application of plasma cell markers in the prognosis assessment of HNSCC solves at least one problem in the prior art, such as the problem of inaccurate or low efficiency in the prognosis assessment of HNSCC.

[0004] In a first aspect, the present invention provides IgA + Application of plasma cell markers in the prognosis assessment of HNSCC, among which IgA + The plasma cell markers include at least one of S100A10, JCHAIN, GNG7, MANF, SSR3, CD38, SPCS3, IRF4, SSR4, ALG5, FAM30A, KDELR1, CYBA, and SEC61G.

[0005] IgA + The expression levels of plasma cell markers can reflect the prognosis of HNSCC. Among them, high expression of S100A10, MANF, SPCS3, SSR4, ALG5, KDELR1, CYBA or SEC61G is positively correlated with poor prognosis of HNSCC (poor prognosis), and high expression of JCHAIN, GNG7, SSR3, CD38, IRF4 or FAM30A is negatively correlated with poor prognosis of HNSCC (good prognosis).

[0006] In some embodiments, IgA + Plasma cell markers include S100A10, MANF, SPCS3, SSR4, ALG5, KDELR1, CYBA, and SEC61G.

[0007] In some embodiments, IgA + Plasma cell markers include JCHAIN, GNG7, SSR3, CD38, IRF4, and FAM30A.

[0008] In some embodiments, IgA + Plasma cell markers include S100A10, JCHAIN, GNG7, MANF, SSR3, CD38, SPCS3, IRF4, SSR4, ALG5, FAM30A, KDELR1, CYBA, and SEC61G. + Plasma cell markers can greatly improve the accuracy of HNSCC prognosis assessment.

[0009] In a second aspect, the present invention provides a method for evaluating the prognosis of HNSCC, comprising the following steps: Obtaining single-cell data of HNSCC and identifying IgA + Plasma cells and IgA extraction + Plasma cell markers; Plasma cell-related module gene sets were identified using WGCNA and CIBERSORT algorithms; Obtain a bulk RNA-seq sequencing dataset with prognostic data and divide it into a training set and a validation set; Based on the training set, Lasso regression algorithm and Cox regression algorithm were used to analyze IgA + Plasma cell markers and plasma cell-related module gene sets were used for feature screening to establish a HNSCC prognostic risk prediction model; The HNSCC prognostic risk prediction model is verified using the validation set, the parameters of the HNSCC prognostic risk prediction model are updated, and the HNSCC prognostic risk prediction model with optimized parameters is obtained; Based on the bulk RNA-seq sequencing data of HNSCC tissues, the HNSCC prognostic risk score was calculated using the HNSCC prognostic risk prediction model with optimized parameters; Among them, the bulk RNA-seq sequencing data of HNSCC tissues include expression data of at least one of S100A10, JCHAIN, GNG7, MANF, SSR3, CD38, SPCS3, IRF4, SSR4, ALG5, FAM30A, KDELR1, CYBA, and SEC61G.

[0010] Based on the bulk RNA-seq sequencing data of HNSCC tissues from patients, the HNSCC prognosis evaluation can be performed through the HNSCC prognosis risk prediction model after parameter optimization, which can improve the efficiency of HNSCC prognosis evaluation. Among them, the use of Lasso regression algorithm and Cox regression algorithm for feature screening can improve the accuracy and stability of HNSCC prognosis evaluation.

[0011] In some embodiments, the bulk RNA-seq sequencing data of HNSCC tissue includes expression data of S100A10, JCHAIN, GNG7, MANF, SSR3, CD38, SPCS3, IRF4, SSR4, ALG5, FAM30A, KDELR1, CYBA, and SEC61G.

[0012] In some embodiments, IgA with prognostic data + Plasma cell bulk RNA-seq sequencing data sets include the HNSCC cohort data set of the TCGA database, the GSE65858 data set of the GEO database, the GSE41613 data set of the GEO database, the GSE27020 data set of the GEO database, and the GSE117973 data set of the GEO database.

[0013] In some embodiments, the HNSCC prognostic risk score is calculated as: Riskscore=S100A10 expression× 0.31828 + JCHAIN ​​expression× (-0.97118) + GNG7 expression× (-0.45969) + MANF expression× 0.35197 + SSR3 expression× (-0.07333) + CD38 expression× (-0.53677) + SPCS3 expression× 1.04363 + IRF4 expression× (-0.79329) + SSR4 expression× 0.73511 + ALG5 expression× 0.84510 + FAM30A expression× (-0.06392) + KDELR1 expression× 0.63198 + CYBA expression× 1.68725 + SEC61G expression× 0.57559; Among them, Riskscore represents the HNSCC prognostic risk score.

[0014] In a third aspect, the present invention provides an electronic device, comprising a processor and a memory; wherein the memory stores program instructions, and when the program instructions are executed by the processor, the above-mentioned HNSCC prognosis assessment method is implemented.

[0015] In a fourth aspect, the present invention provides a computer-readable storage medium storing program instructions, which, when executed by a processor, implements the above-mentioned HNSCC prognosis assessment method.

[0016] Due to the adoption of the above technical solution, the embodiments of the present invention have at least the following beneficial effects: (1) Screening of IgA associated with HNSCC prognosis + Plasma cell marker genes reveal the key molecular mechanisms related to prognosis in HNSCC, thus providing a scientific basis for personalized treatment of patients; (2) By combining multiple bioinformatics and machine learning algorithms, an efficient and stable HNSCC prognostic risk prediction model was successfully constructed, which improved the efficiency, accuracy and stability of HNSCC prognostic assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 The results of cell population annotation in the examples of the present invention are shown.

[0018] Figure 2 The differences in the proportions of each cell population in the primary tumor group, recurrent tumor group, and metastasis group in the examples of the present invention are shown, where * indicates P≤0.05, ** indicates P≤0.01, *** indicates P≤0.001, and ns indicates P>0.05.

[0019] Figure 3 The spatial distribution of plasma cells and cancer cells in the embodiment of the present invention is shown.

[0020] Figure 4 The correlation of the spatial distribution of various types of cells in the embodiments of the present invention is shown.

[0021] Figure 5 The results of plasma cell subset annotation in the examples of the present invention are shown.

[0022] Figure 6 The figure shows the difference in the proportion of prognosis-related cells in the two types of plasma cell subsets in the embodiment of the present invention, where **** indicates P≤0.0001.

[0023] Figure 7 The correlation between each gene module and plasma cell infiltration in the embodiment of the present invention is shown.

[0024] Figure 8 The C index of each modeling scheme in the embodiment of the present invention is shown.

[0025] Fig. 9 The C-index comparison results of various models in the embodiments of the present invention on the TCGA and GSE65858 datasets are shown.

[0026] Fig.10 The C-index comparison results of various models in the embodiments of the present invention on the GSE117973, GSE41613 and GSE27020 datasets are shown.

[0027] Fig.11 The survival curves of the high-risk group and the low-risk group in each data set in the embodiment of the present invention are shown.

[0028] Fig.12 The ROC curves of various data sets in the embodiments of the present invention are shown.

[0029] Fig.13 This is a nomogram for predicting disease risk based on each biomarker in the model in an embodiment of the present invention.

[0030] Fig.14 The differences in risk scores for different tumor stages in the examples of the present invention are shown.

[0031] Fig.15 It is shown that the tumor-related pathways in the examples of the present invention are upregulated in the high-risk group.

[0032] Fig.16 The differences in matrix and immune infiltration between high-risk and low-risk groups in the examples of the present invention are shown.

[0033] Fig.17 The difference in immunophenotypic scores between high-risk and low-risk groups in the examples of the present invention is shown.

[0034] Fig.18 The differences in tumor immune dysfunction and rejection scores between high-risk and low-risk groups in the examples of the present invention are shown.

[0035] Fig.19 The correlation between the risk score and tumor immune dysfunction and rejection score in the embodiment of the present invention is shown.

[0036] Fig. 20 The difference in treatment response between high-risk and low-risk groups in the embodiment of the present invention is shown.

[0037] Fig.21 The correlation between the risk score and the sensitivity of anti-tumor drugs in the embodiment of the present invention is shown. DETAILED DESCRIPTION

[0038] The following is a clear and complete description of the concept of the present invention and the technical effects produced, so as to fully explain the purpose, scheme and effect of the present invention.

[0039] In one embodiment of the present invention, the GSE234933 dataset was downloaded from the public database Gene Expression Omnibus (GEO, https: / / www.ncbi.nlm.nih.gov / geo / ) to obtain HNSCC single-cell RNA sequencing data. The R language Seurat package was used for preliminary quality control, and the RNA sequencing data of the following low-quality cells were eliminated: cells with less than 500 or more than 5000 detected genes, and cells with a mitochondrial gene expression ratio greater than 20%, and finally 231,883 cells were retained. The first 40 principal components were selected for principal component analysis, and then UMAP (Uniform ManifoldApproximation and Projection) was used for visual dimensionality reduction. Cell clusters were identified by the FindClusters function of the R language Seurat package, where the resolution parameter was set to 0.8. The marker genes of various types of cells were obtained through the Cell Taxonomy database (https: / / ngdc.cncb.ac.cn / celltaxonomy / ), and the clustered cell populations were annotated, such as Figure 1 Compare the changes in the proportion of each cell among multiple groups, such as Figure 2 As shown in Figure 2, it can be found that the proportion of plasma cells changes significantly in the primary tumor group, recurrent tumor group, and metastasis group. Then, the GSE200310 spatial transcriptome dataset of the GEO database was deconvoluted using the R language CARD package and single-cell RNA sequencing data to obtain the distribution relationship between plasma cells and cancer cells, as shown in Figure 2. Figure 3 and Figure 4 As shown, it can be found that there is a negative correlation between the two.

[0040] Extract RNA sequencing data of plasma cell subsets, reuse R language Seurat package for dimensionality reduction, cluster analysis and annotate into IgA + Plasma cells, IgG + Plasma cells and other plasma cells, such as Figure 5 The Scissors algorithm was then used to integrate HNSCC data from the TCGA database and combined with the survival information in the bulk RNA-seq data to identify plasma cell subsets associated with HNSCC prognosis. + The prognostic protective cells in plasma cells are significantly higher than IgG + Plasma cells, and the prognostic risk cells are only from IgG + Plasma cells, such as Figure 6 shown.

[0041] Use the Findmarker function of the Seurat package to obtain IgA +Marker genes of plasma cells in the single-cell dimension. Then, the bulk RNA-seq data from the TCGA database were analyzed using the R language CIbersort algorithm and the WGCNA package to obtain plasma cell-related gene modules (black modules) in the bulk dimension, such as Figure 7 As shown. Data sets with prognostic data (HNSCC cohort data set from TCGA database and GSE65858, GSE41613, GSE27020, GSE117973 data sets from GEO database) were collected to analyze IgA + Plasma cell-related gene set (i.e., IgA + Plasma cell markers and plasma cell-related module genes) were used for modeling and internal and external validation of the model. The HNSCC cohort dataset from the TCGA database was randomly divided into a training set and an internal validation set in a 1:1 ratio, and all datasets from the GEO database were used as external validation sets. IgA + Plasma cell-related gene sets were included in the modeling comparison of multiple groups of machine learning in multiple data sets. Enet, StepCox, SVM, CoxBoost, SuperPC, plsRcox, RSF, GBM, Lasso, Ridge and other algorithms were used for feature screening and model building. The best modeling scheme (Lasso+cox regression) was selected considering the C index and interpretability, such as Figure 8 Finally, the Lasso regression algorithm was combined with the Cox regression algorithm to screen features and establish a HNSCC prognosis risk prediction model, and the C index was compared with the existing HNSCC prognosis prediction model published on Pubmed, as shown in the figure. Fig. 9 and Fig.10 As shown, the HNSCC prognosis risk prediction model (My-model) in the embodiment of the present invention shows a stronger stable prediction ability. Among them, the final risk score formula of the HNSCC prognosis risk prediction model in the embodiment of the present invention is: Riskscore= S100A10 × 0.31828 + JCHAIN ​​× (-0.97118) + GNG7 × (-0.45969) + MANF × 0.35197 + SSR3 × (-0.07333) + CD38 × (-0.53677) + SPCS3× 1.04363 + IRF4 × (-0.79329) + SSR4 × 0.73511 + ALG5 × 0.84510 + FAM30A × (-0.06392) + KDELR1 × 0.63198 + CYBA × 1.68725 + SEC61G × 0.57559.

[0042] Table 1. Existing model author names and PMIDs of corresponding source articles for comparison Based on the risk scores obtained by the model, the survminer package was used to find the optimal cutoff value and divide each data set into high-risk and low-risk groups. After modeling, the Kaplan-Meier survival analysis curve was used to divide the high-risk group and the low-risk group (such as Fig.11 The time-dependent ROC curve was used to evaluate its predictive performance, as shown in Fig.12 As shown in Figure 1, all are greater than 0.6 and most are greater than 0.65, indicating that it has stable accuracy. The embodiment of the present invention plots the prediction results into a nomogram, such as Fig.13 shown.

[0043] Functional validation of the model showed that the risk scores of various common cancer stages were different, indicating that they were correlated with the risk scores to a certain extent, such as Fig.14 The R language GSEA package was used to analyze the pathway activity between high-risk and low-risk groups, as shown in Fig.15 As shown in Figure 2, it can be found that tumor-related pathways (oxidative phosphorylation, epithelial-mesenchymal transition) are upregulated in the high-risk group. The ESTIMATE package in R language was used for cell infiltration analysis, as shown in Figure 2 Fig.16 As shown, significant differences in matrix and immune cell infiltration were found between the high-risk and low-risk groups.

[0044] Finally, the relationship between risk score and treatment effect was tested by using immunotherapy data from the TCIA database, tumor immune dysfunction and rejection scores and immunotherapy responses predicted by the TIDE algorithm, and the half-maximal inhibitory concentrations of common anti-head and neck cancer drugs predicted by the OncoPredict algorithm, such as Figure 17-Figure 21 As shown, significant differences and correlations were found between the high-risk and low-risk groups.

[0045] It will be appreciated by those skilled in the art that all or part of the processes in the above method may be accomplished by computer program instructions, which may be stored in a computer-readable storage medium, and which, when executed, may include the processes of the above method. The computer-readable storage medium may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory.

[0046] The above is only a preferred embodiment of the present invention. The present invention is not limited to the above implementation. As long as the technical effect of the present invention is achieved by the same or equivalent means, it should belong to the protection scope of the present invention. Within the protection scope of the present invention, its technical scheme and / or implementation method can have various modifications and changes.

Claims

1. IgA + The application of plasma cell markers in the prognosis assessment of HNSCC is characterized by: IgA + The plasma cell markers include at least one of S100A10, JCHAIN, GNG7, MANF, SSR3, CD38, SPCS3, IRF4, SSR4, ALG5, FAM30A, KDELR1, CYBA, and SEC61G.

2. The use according to claim 1, characterized in that: IgA + Plasma cell markers include S100A10, MANF, SPCS3, SSR4, ALG5, KDELR1, CYBA, and SEC61G.

3. The use according to claim 1, characterized in that: IgA + Plasma cell markers include JCHAIN, GNG7, SSR3, CD38, IRF4, and FAM30A.

4. The use according to claim 1, characterized in that: IgA + Plasma cell markers include S100A10, JCHAIN, GNG7, MANF, SSR3, CD38, SPCS3, IRF4, SSR4, ALG5, FAM30A, KDELR1, CYBA, and SEC61G.

5. A method for evaluating the prognosis of HNSCC, characterized in that: The following steps are involved: Obtaining single-cell data of HNSCC and identifying IgA + Plasma cells and IgA extraction + Plasma cell markers; Plasma cell-related module gene sets were identified using WGCNA and CIBERSORT algorithms; Obtain a bulk RNA-seq sequencing dataset with prognostic data and divide it into a training set and a validation set; Based on the training set, Lasso regression algorithm and Cox regression algorithm were used to analyze IgA + Plasma cell markers and plasma cell-related module gene sets were used for feature screening to establish a HNSCC prognostic risk prediction model; The HNSCC prognostic risk prediction model is verified using the validation set, the parameters of the HNSCC prognostic risk prediction model are updated, and the HNSCC prognostic risk prediction model with optimized parameters is obtained; Based on the bulk RNA-seq sequencing data of HNSCC tissues, the HNSCC prognostic risk score was calculated using the HNSCC prognostic risk prediction model with optimized parameters; Among them, the bulk RNA-seq sequencing data of HNSCC tissues include expression data of at least one of S100A10, JCHAIN, GNG7, MANF, SSR3, CD38, SPCS3, IRF4, SSR4, ALG5, FAM30A, KDELR1, CYBA, and SEC61G.

6. The method according to claim 5, characterized in that The bulk RNA-seq sequencing data of HNSCC tissues include the expression data of S100A10, JCHAIN, GNG7, MANF, SSR3, CD38, SPCS3, IRF4, SSR4, ALG5, FAM30A, KDELR1, CYBA and SEC61G.

7. The method according to claim 5, characterized in that Bulk RNA-seq sequencing data sets with prognostic data include the HNSCC cohort data set of the TCGA database, the GSE65858 data set of the GEO database, the GSE41613 data set of the GEO database, the GSE27020 data set of the GEO database, and the GSE117973 data set of the GEO database.

8. The method according to claim 5, characterized in that The calculation formula for the HNSCC prognostic risk score is: Riskscore=S100A10 expression× 0.31828 + JCHAIN ​​expression× (-0.97118) + GNG7 expression× (-0.45969) + MANF expression× 0.35197 + SSR3 expression× (-0.07333) + CD38 expression× (-0.53677) + SPCS3 expression× 1.04363 + IRF4 expression× (-0.79329) + SSR4 expression× 0.73511 + ALG5 expression× 0.84510 + FAM30A expression× (-0.06392) + KDELR1 expression× 0.63198 + CYBA expression× 1.68725 + SEC61G expression× 0.57559; Among them, Riskscore represents the HNSCC prognostic risk score.

9. An electronic device, characterized in that: The method comprises a processor and a memory; wherein the memory stores program instructions, and when the program instructions are executed by the processor, the HNSCC prognosis assessment method according to any one of claims 5 to 8 is implemented.

10. A computer-readable storage medium, characterized in that: Program instructions are stored, and when the program instructions are executed by a processor, the HNSCC prognosis assessment method described in any one of claims 5 to 8 is implemented.

Images