Identification of A Gene Set Associated with Colorectal Cancer in Microarray Data Using The Entropy Method

Objective: We sought to apply Shannon's entropy to determine colorectal cancer genes in a microarray dataset.

Materials and methods: In the retrospective study, 36 samples were analysed, 18 colorectal carcinoma and 18 paired normal tissue samples. After identification of the gene fold-changes, we used the entropy theory to identify an effective gene set. These genes were subsequently categorised into homogenous clusters.

Results: We assessed 36 tissue samples. The entropy theory was used to select a set of 29 genes from 3128 genes that had fold-changes greater than one, which provided the most information on colorectal cancer. This study shows that all genes fall into a cluster, except for the R08183 gene.

Conclusion: This study has identified several genes associated with colon cancer using the entropy method, which were not detected by custom methods. Therefore, we suggest that the entropy theory should be used to identify genes associated with cancers in a microarray dataset.