Background: Colon cancer (CC) is a heterogeneous disease influenced by complex gene networks. As such, the relationship between networks and CC should be elucidated to obtain further insights into tumour biology.
Results: Weighted gene co-expression network analysis, a powerful technique used to extract co-expressed gene networks from mRNA expressions, was conducted to identify 11 co-regulated modules in a discovery dataset with 461 patients. A transcriptional module enriched in cell cycle processes was correlated with the recurrence-free survival of the CC patients in the discovery (HR = 0.59; 95% CI = 0.42-0.81) and validation (HR = 0.51; 95% CI = 0.25-1.05) datasets. The prognostic potential of the hub gene Centromere Protein-A (CENPA) was also identified and the upregulation of this gene was associated with good survival. Another cell cycle phase-related gene module was correlated with the survival of the patients with a KRAS mutation CC subtype. The downregulation of several genes, including those found in this co-expression module, such as cyclin-dependent kinase 1 (CDK1), was associated with poor survival.
Conclusion: Network-based approaches may facilitate the discovery of biomarkers for the prognosis of a subset of patients with stage II or III CC, these approaches may also help direct personalised therapies.
Keywords: Biomarker; Colon cancer; Gene expression profiling; Systems biology; WGCNA.