Identification of the candidate genes of diagnosing rheumatoid arthritis using the single-cell sequencing technology and T cell subclusters analysis of patients with rheumatoid arthritis.

Abstract

This study aims to analyze the heterogeneity among different cell types in peripheral blood mononuclear cells (PBMC) in rheumatoid arthritis (RA) patients and to analyze T cell subsets to obtain key genes that may lead to RA. The sequencing data of 10,483 cells were obtained from the GEO data platform. The data were filtered and normalized initially and, then, principal component analysis (PCA) and t-Distributed Stochastic Neighbor Embedding (TSNE) cluster analysis were performed using the Seurat package in R language to group the cells, thereby obtaining the T cells. The T cells were subjected to subcluster analysis. The differentially expressed genes (DEGs) in T cell subclusters were obtained, and the hub genes were determined by Gene Ontology (GO) functional enrichment analysis, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis and protein-protein interaction (PPI) network construction. Finally, the hub genes were validated using other datasets in the GEO data platform. The PBMC of RA patients were mainly divided into T cells, natural killer (NK) cells, B cells, and monocyte cells. The number of T cells was 4,483, which were further divided into seven clusters. The pseudotime trajectory analysis showed that the differentiation of T cells developed from cluster 0 and cluster 1 to cluster 5 and cluster 6. Through GO, KEGG and PPI analysis, the hub genes were identified. After validation by external data sets, nine genes were identified as candidate genes highly associated with the occurrence of RA, including CD8A, CCL5, GZMB, NKG7, PRF1, GZMH, CCR7, GZMK, and GZMA. Based on single-cell sequencing analysis, we identified nine candidate genes for diagnosing RA, and validated their diagnostic value for RA patients. Our findings may provide new sights for the diagnosis and treatment of RA.