Gene Expression Analysis using Bayesian Networks for Breast Cancer Prognosis

Hypothesis testing using Bayesian networks has been proven time and again to be very useful for various applications. One of these application areas is gene expression analysis. Gene expression analysis using Bayesian networks is widely researched topic since the early '90s. Gene expression can be used for prognosis of various diseases including cancer. This paper proposes modeling gene expression data using Bayesian networks for breast cancer prognosis with the help of DNA microarray data. Gene expression data has been used to build a Bayesian Network to study gene regulation in tumor samples. The model has been built using Grow-Shrink algorithm, Hill Climbing algorithm and Incremental Association Markov Blanket algorithm. The Markov blanket of the outcome of the Bayesian network can assist with breast cancer prognosis as well help in deciding the right therapy for patients