Identification of osteoarthritis-characteristic genes and immunological micro-environment features by bioinformatics and machine learning

Abstract

Abstract Background Osteoarthritis (OA) is a mechanistically complex chronic joint disease which will reduce the life quality of middle-aged and elderly people as well as increase the socioeconomic burden. Currently, the pathophysiology of OA is not entirely clear. The purpose of this study was to investigate the genes, functional pathways, as well as characteristics of immune infiltration, that are involved in the genesis and progression of osteoarthritis. Methods The GEO database was used to obtain gene expression profiles. R software was used for the screening of differentially expressed genes (DEGs) and enrichment analysis of these genes. OA characteristic genes were screened by WGCNA and the Lasso algorithm. Using ssGSEA, we evaluated the infiltration levels of immune cells in cartilage, followed by correlation analysis between immune cells and OA characteristic genes. Results We identified 80 DEGs in total. Results of the functional enrichment indicated that these DEGs were associated with chondrocyte metabolism, apoptosis, and inflammation. Three OA characteristic genes were identified by WGCNA analysis and the lasso algorithm, and then their expression levels were verified by the verification set. Finally, immune cells infiltration analysis revealed that T cells and B cells were mainly associated with OA. In addition, Tspan2, HtrA1 showed correlation with some of the infiltrating immune cells. Conclusions The results of a comprehensive bioinformatic analysis showed OA is associated with a number of characteristic genes, functional pathways, immune cell infiltration processes. Characteristic genes and functional pathways identified in this study can be used as biomarkers to guide drug treatment and molecular-level research on OA.