Bioinformatics Analysis Identifies Key Genes in Recurrent Implantation Failure Based on Immune Infiltration

Recurrent implantation failure (RIF) is a thorny problem often encountered in the field of assisted reproduction. Existing evidences suggest that immune dysregulation may be involved in the pathogenesis of RIF. The purpose of this study is to explore immune-related genes contributing to RIF through data mining. The endometrial expression profiles of 24 RIF and 24 controls were obtained from the GEO database. The immune infiltration in bulk tissue was estimated by single sample gene set enrichment analysis (ssGSEA) method based on marker gene sets for immune cells generated from endometrial single-cell RNA sequencing data. The results showed that the infiltration levels of B cells and regulatory T cells (Tregs) were significantly reduced in the RIF group. Four hub genes (GJA1, PRKAG2, CPT1A, and ICA1) were identified by integrated analysis of weighted gene co-expression network analysis (WGCNA), random forest and LASSO regression. Moreover, these hub genes were significantly correlated with certain immune-related factors, especially CXCL12, CEACAM1, and XCR1. Single-gene GSEA indicated that the pathways associated with hub genes included the regulation of cell cycle, the process of epithelial-mesenchymal transition and transplant rejection, etc. A predictive model for RIF was constructed based on hub genes and performed well in the training dataset and the other two external datasets. Thus, this study identified immune-related key genes in RIF and provided new biomarkers for early diagnosis.