Boolean network inference and control using metaheuristic algorithms

This thesis focuses on the topic of gene regulatory network inference and control based on the Boolean network model. Though investigated for decades, inferring gene regulatory networks from the experimental data remains a challenging problem in computational systems biology. The major goal of the network inference is to discover and model the interactions between the genes and the regulatory rules that govern their dynamical changes. The control of gene regulatory networks is also an outstanding challenge in computational systems biology. One of the central tasks is to achieve an understanding of the network dynamics that guiding a cellular state to a desired state with control targets. This thesis addresses these problems based on the Boolean network model. The main contributions of this thesis include: • Propose an And/Or tree ensemble algorithm for inferring gene regulatory networks from short and noisy data • Propose a genetic algorithm with marker-based encoding for inferring gene regulatory networks • Propose a genetic algorithm with a coding of directed ordered tree for controlling gene regulatory network