Weighted gene co-expression network analysis for hub genes in colorectal cancer

BackgroundThis study is designed to explore hub genes participating in colorectal cancer (CRC) development through weighted gene co-expression network analysis (WGCNA).MethodsExpression profiles of CRC and normal samples were retrieved from the Gene Expression Omnibus (GEO) and the Cancer Genome Atlas (TCGA), and were subjected to WGCNA to filter differentially expressed genes with significant association with CRC. Functional enrichment analysis and protein-protein interaction (PPI) analysis were carried out to filter the candidate genes, further and survival analysis was performed for the candidate genes to obtain potential regulatory hub genes in CRC. Expression analysis was conducted for the candidate genes and a multifactor model was established.ResultsAfter differential analysis and WGCNA, 289 candidate genes were filtered from the GEO and TCGA. Further functional enrichment analysis demonstrated possible regulatory pathways and functions. PPI analysis filtered 15 hub genes and survival analysis indicated a significant correlation of CLCA1, CLCA4, and CPT1A with prognosis of patients with CRC. The multifactor Cox risk model established based on the three genes revealed that if the three genes were a gene set, they had well predictive capacity for the prognosis of patients with CRC.ConclusionsCLCA1, CLCA4, and CPT1A express at low levels in CRC and function as core anti-tumor genes. As a gene set, they can predict prognosis well.