In silico analysis identifies genes common between five primary gastrointestinal cancer sites with potential clinical applications

Abstract

Background Previous studies have investigated differential gene expression in gastrointestinal (GI) epithelial cancers by microarray. The aim of the present study was to use data from the Oncomine database to identify genes that share a similar differential expression in two or more primary GI cancer sites.

Methods Five thousand of the most differentially expressed genes in epithelial cancers (compared to normal tissue) arising in the pancreas, liver, stomach, esophagus or colorectum were identified (1,000 per primary site) from Oncomine. Using Venn diagrams, genes common to two or more primary GI sites were identified. Functional and pathway analysis was performed on genes that were similarly expressed in ≥3 of the five areas of the GI tract.

Results Forty six studies comprising 5,876 samples were included. Overall, 90.6% genes were unique to the respective primary sites, 7.4% shared between two GI primary sites, 1.8% between three and 0.2% between four GI primary sites. Pancreatic and hepatocellular cancers (HCC) shared most number of upregulated genes (N=66) while HCC and gastric cancer shared most downregulated genes (N=59). Genes encoding enzymes comprised the most commonly shared genes between GI primary sites (30.4% of upregulated and 63.2% of downregulated genes). Those genes that were shared between three or more GI primary sites also showed significant differential expression in the same direction in other non-GI cancers.

Conclusion The present study has identified several genes that show similar differential expression in cancers arising from two or more sites in the GI tract. These genes can be potentially useful as novel therapeutic targets.

Keywords Oncomine, pathway, network, database, cancer, gastrointestinal, therapy

Ann Gastroenterol 2014; 27 (3): 231-236