Identification and validation of biomarkers related to lipid metabolism in osteoarthritis based on machine learning algorithms.

ABSTRACT

BACKGROUND: Osteoarthritis and lipid metabolism are strongly associated, although the precise targets and regulatory mechanisms are unknown. METHODS: Osteoarthritis gene expression profiles were acquired from the GEO database, while lipid metabolism-related genes (LMRGs) were sourced from the MigSB database. An intersection was conducted between these datasets to extract gene expression for subsequent differential analysis. Following this, functional analyses were performed on the differentially expressed genes (DEGs). Subsequently, machine learning was applied to identify hub genes associated with lipid metabolism in osteoarthritis. Immune-infiltration analysis was performed using CIBERSORT, and external datasets were employed to validate the expression of these hub genes. RESULTS: Nine DEGs associated with lipid metabolism in osteoarthritis were identified. UGCG and ESYT1, which are hub genes involved in lipid metabolism in osteoarthritis, were identified through the utilization of three machine learning algorithms. Analysis of the validation dataset revealed downregulation of UGCG in the experimental group compared to the normal group and upregulation of ESYT1 in the experimental group compared to the normal group. CONCLUSIONS: UGCG and ESYT1 were considered as hub LMRGs in the development of osteoarthritis, which were regarded as candidate diagnostic markers. The effects are worth expected in the early diagnosis and treatment of osteoarthritis.