Identification of potential biomarkers and analysis of prognostic values in head and neck squamous cell carcinoma by bioinformatics analysis.

The purpose of this study was to find disease-associated genes and potential mechanisms in head and neck squamous cell carcinoma (HNSCC) with deoxyribonucleic acid microarrays. The gene expression profiles of GSE6791 were downloaded from the Gene Expression Omnibus database. Differentially expressed genes (DEGs) were obtained with packages in R language and STRING constructed protein-protein interaction (PPI) network of the DEGs with combined score >0.8. Subsequently, module analysis of the PPI network was performed by Molecular Complex Detection plugin and functions and pathways of the hub gene in subnetwork were studied. Finally, overall survival analysis of hub genes was verified in TCGA HNSCC cohort. A total of 811 DEGs were obtained, which were mainly enriched in the terms related to extracellular matrix (ECM)-receptor interaction, ECM structural constituent, and ECM organization. A PPI network was constructed, consisting of 401 nodes and 1,254 edges and 15 hub genes with high degrees in the network. High expression of 4 genes of the 15 genes was associated with poor OS of patients in HNSCC, including , , , and . Two significant modules were detected from the PPI network, and the enriched functions and pathways included proteasome, ECM organization, and ECM-receptor interaction. In conclusion, we propose that , , , and may be further explored as potential biomarkers to aid HNSCC diagnosis and treatment.