Development a prognostic risk model based on B cell-related immune genes in Ovarian Cancer by integrative analysis of single-cell and bulk RNA sequencing data

Background: Ovarian cancer (OV) is the most serious malignancy of the female reproductive system and is generally diagnosed at an advanced stage with peritoneal metastasis. The aim of this research was to construct a tumor-infiltrating B (TIL-B) lymphocyte-associated immune-related gene profile for prognostic assessment of ovarian cancer patients. Methods: . The single cell RNA-seq data of metastatic ovarian cancer in this study were obtained by GEO data, and bulk RNA-seq data was obtained from TCGA database. Identification of immune-related marker genes in infiltrating B lymphocytes in tumor tissues of patients with metastatic ovarian cancer by scRNA-seq data analysis. Subsequently, based on bulk RNA-seq data and clinical follow-up data, univariate Cox analysis was used to identify the prognostic targets associated with tumor infiltrating B lymphocytes, and least absolute shrinkage and selection operator (LASSO) regression and multivariate Cox regression were used to construct prognostic risk models. The Kaplan-Merier survival curve and the ROC curve were used to test the prognostic performance of the model. EPIC, MCPcounter and ssGSEA software were used to predict the proportion of tumor-infiltrating lymphocytes in bulk RNA-seq expression profile samples. Multivariate Cox regression analysis was used to screen independent prognostic factors and construct linear plots to predict and assess 3 and 5-year survival of ovarian cancer patients. Results: . Single cell data from metastatic ovarian cancer tissue were clustered into 19 subgroups. After known cell type annotation, these cell subpopulations were annotated as six known cell types. The proportion of B cells was contrary to the clinical stage of the patient's tumor. Difference analysis identified 88 immune-related genes specifically expressed by B cells. Univariate Cox regression analysis, the LASSO regression analysis and multivariate Cox regression analysis were used to identify the independent prognostic factors associated with tumor invasion B cell immunity, ISG20 and SLAMF7 . Based on the risk model constructed by ISG20 and SLAMF7 , the AUC values of the 3-year and 5-year survival in the training set were 0.619 and 0.736, the AUC values of the test set were 0.694 and 0.758, and the AUC values of the external validation set were 0.6 and 0.61. The proportion of CD8+T cells, B cells, cytotoxic lymphocytes and aDC cells in the low-risk group was higher than in the high-risk group. The prognostic model has better independence and has good prognostic evaluation effect combined with clinical characteristics (p=0.013). At the same time, we also construct a nomogram based on the prediction model. Conclusions: . Our study identified immune-related prognostic marker genes ISG20 and SLAMF7 in TIL-B cells by analyzing single-cell and bulk RNA-seq data of metastatic ovarian cancer, and established and verified the prognostic model as an independent prognostic indicator of ovarian cancer. It can provide potential therapeutic targets for immunotherapy and chemotherapy in patients with advanced cancer and predict patients' therapeutic response.