Inference of Gene Regulatory Networks for Prostate Cancer using Bayesian Networks with Feedback and Feed Forward Loops

Abstract

The solution to any problem depends on the depth of our understanding of it. Cancer is a disease that is being investigated at multiple levels and from multiple perspectives to understand the details of its origins and expansions in order to be able to figure a cure for it. We can now computationally analyze the biological data produced by genome analysis techniques like genomics, proteomics, and transcriptomics. DNA micro array technology has made available large gene expression datasets for entire genomes. It has been clinically observed that inside a human cell, activity of a gene often turns on or turns off one or more other genes. Such relationships in the co-regulation of genes is captured by gene regulatory network models which are computationally constructed from gene expression datasets. It has been observed that healthy and diseased states of a human cell show different regulatory interrelations between genes. In this paper, we have proposed to use a stochastic approach called Bayesian Networks with Feedback and Feedforward loops for inference of inter dependence in the regulation of genes in case of Prostate Cancer. It was observed that 4 of the networks revealed by the proposed approach matched the ones observed in clinical studies.