Comprehensive analysis of anoikis-related genes in diagnosis osteoarthritis: based on machine learning and single-cell RNA sequencing data

Abstract

AbstractOsteoarthritis (OA) is a degenerative disease closely associated with Anoikis. The objective of this work was to discover novel transcriptome-based anoikis-related biomarkers and pathways for OA progression.The microarray datasets GSE114007 and GSE89408 were downloaded using the Gene Expression Omnibus (GEO) database. A collection of genes linked to anoikis has been collected from the GeneCards database. The intersection genes of the differential anoikis-related genes (DEARGs) were identified using a Venn diagram. Infiltration analyses were used to identify and study the differentially expressed genes (DEGs). Anoikis clustering was used to identify the DEGs. By using gene clustering, two OA subgroups were formed using the DEGs. GSE152805 was used to analyse OA cartilage on a single cell level. 10 DEARGs were identified by lasso analysis, and two Anoikis subtypes were constructed. MEgreen module was found in disease WGCNA analysis, and MEturquoise module was most significant in gene clusters WGCNA. The XGB, SVM, RF, and GLM models identified five hub genes (CDH2, SHCBP1, SCG2, C10orf10, P FKFB3), and the diagnostic model built using these five genes performed well in the training and validation cohorts. analysing single-cell RNA sequencing data from GSE152805, including 25,852 cells of 6 OA cartilage.

Keywords

• Osteoarthritis
• anoikis
• immunity
• machine learning
• single-cell analysis