Reconstruction of gene regulatory network of colon cancer using information theoretic approach

TL;DR

This paper reconstructs colon cancer gene regulatory networks from microarray data using mutual information, revealing complex interactions and validating results with biological databases.

Contribution

It introduces a novel application of mutual information for colon cancer gene network reconstruction with topological analysis and validation.

Findings

Large number of gene interactions identified in colon cancer

Constructed networks with 30 to 500 interactions

Validated inferred networks with biological databases

Abstract

Reconstruction of gene regulatory networks or 'reverse-engineering' is a process of identifying gene interaction networks from experimental microarray gene expression profile through computation techniques. In this paper, we tried to reconstruct cancer-specific gene regulatory network using information theoretic approach - mutual information. The considered microarray data consists of large number of genes with 20 samples - 12 samples from colon cancer patient and 8 from normal cell. The data has been preprocessed and normalized. A t-test statistics has been applied to filter differentially expressed genes. The interaction between filtered genes has been computed using mutual information and ten different networks has been constructed with varying number of interactions ranging from 30 to 500. We performed the topological analysis of the reconstructed network, revealing a large number of interactions in colon cancer. Finally, validation of the inferred results has been done with available biological databases and literature.