Analysis and validation of diagnostic biomarkers and immune cell infiltration characteristics in osteoarthritis by integrating bioinformatics and machine learning.

Objective: Exosomes as important carriers of intercellular communication have frequently appeared in recent studies related to osteoarthritis (OA), while the specific mechanism of exosome action in osteoarthritis remains unclear. The aim of this study was to identify potential exosome-related biomarkers in osteoarthritis, to explore the role and mechanism of exosome-related genes in articular cartilage.

Methods: The data on exosome related genes and normal and OA cartilage genes were obtained through online databases. The potential mechanisms of these genes were revealed by multiple gene enrichment analysis algorithms. Machine learning methods were utilized to identify exosome-related differential genes (ERDEGs) with highly correlated OA features (Hub OA-ERDEGs). In addition, we created a nomogram to assess the ability of Hub OA-ERDEGs to diagnose OA. Single-sample gene set enrichment analysis (ssGSEA) was used to observe the infiltration characteristics of immune cells in OA and their relationship with Hub OA-ERDEGs.

Results: The results of screening Hub OA-ERDEGs using machine learning algorithms show that: TOLLIP, ALB, HP, RHOBTB3, GSTM2, S100A8 and AKR1B1 were significantly up-regulated or down-regulated in OA samples and verified by qRT- PCR for validation. Using the ssGSEA algorithm, we discovered that 8 types of immune cell infiltration and 5 types of immune cell activation.