Abstract:
DNA hybridization arrays measure the expression levels for thousands of genes. These measurements provide us with a “snapshot” of transcription levels in the cell. A major challenge in computational biology is to identify the gene-protein, gene-gene, and protein-protein interactions using such measurements, as well as some biological features of cellular systems. In our study we aimed at building up our framework on the use of Bayesian networks. A Bayesian network is a graph-based model of joint multivariate probability distributions that captures properties of conditional independence between variables. Such models are deemed attractive for their ability to describe complex stochastic processes. They also provide a clear methodology for learning from observations, even for noisy ones. However, Bayesian Networks work only for stationary data, require prior information in model selection, and applies to acyclic directed graphs. Dynamic Bayesian network (DBN) is an improved model to overcome the cyclicity and stationary limitations.|Keywords : Gene Regulatory Networks, Structure Learning, Bayesian Networks, Dynamic BN, Gene Expression Profiles.
