De-redundancy in a random boolean network using knockout
Junxiu Liu, Jufang Dai, Min Su, Shunsheng Zhang, Yifan Hua, Yanhu Wang, Haiping Shu · Proceedings of the Genetic and Evolutionary Computation Conference Companion · 2022
Random Boolean Network (RBN) is one of the regulatory networks and the state value of its node is a binary variable. The network usually contains complex interaction and redundancy such as redundant nodes and connections. However, redundant nodes and connections often lead to increased computational and analytical complexities. To address these problems, the node knockout method inspired by the gene regulatory networks is applied in this work to reduce the number of the redundant nodes in the RBN. For the purpose of verifying the performance of the method, three control task experiments with different dynamic characteristics are implemented. The results show that the fitness distribution of the RBN after node knockout is similar to the original RBN but the number of nodes is reduced by > 50% compared to the original RBN. This proposed method significantly reduces the computational complexity while maintaining the task completion. Thus this paper provides an alternative solution for the de-redundancy of the RBNs.