Identifying influential vertices in boolean networks through dynamical voter rank

摘要

Boolean network model has been applied to describe a series of biological systems. The stability of attractors of certain type of Boolean Networks is considered as one of the key directions to investigate the properties of Boolean network model. Due to the vast heterogeneity in topological and dynamical properties among different vertices, a small fraction of vertices could make a great influence on the dynamics. In this paper, we propose a dynamical voter rank algorithm to identify the influential vertices regarding the stability. In this algorithm, the voting score takes into account not only the topological properties, but also the dynamical properties of the vertices. The dynamical voter rank algorithm is observed to be more efficient than high degree adaptive, eigenvector centrality and Google PageRank algorithms in cases of both real and classical Boolean network model simulation. Our work provides an efficient method to identify the important vertices in Boolean networks, which may help to locate certain kinds of virulence genes.