The purpose of this study was to identify genes and pathways for osteoarthritis (OA) diagnosis and therapy. We downloaded the gene expression profile of OA from Gene Expression Omnibus (GEO) database including 10 early OA, 9 late OA, and 5 normal control samples. Next, we screened differentially expressed genes (DEGs) between early- and late-stage OA samples comparing with healthy control samples. Then, the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) software was used to construct protein-protein interaction (PPI) network, which was to predict the proteins that may interact with DEGs. The Gene Ontology (GO)-enrichment method was used to analyze the function of genes in the PPI networks. Meanwhile network module analysis was performed using Cytoscape. A total of 24 and 29 DEGs were identified for the early and late OA, respectively.