Controlling Large Boolean Networks with Temporary and Permanent Perturbations

Abstract

A salient objective of studying gene regulatory networks (GRNs) is to identify potential target genes whose perturbations would lead to effective treatment of diseases. In this paper, we develop two control methods for GRNs in the context of asynchronous Boolean networks. Our methods compute a minimal subset of nodes of a given Boolean network, such that temporary or permanent perturbations of these nodes drive the network from an initial state to a target steady state. The main advantages of our methods include: (1) temporary and permanent perturbations can be feasibly conducted with techniques for genetic modifications in biological experiments; and (2) the minimality of the identified control sets can reduce the cost of experiments to a great extent. We apply our methods to several real-life biological networks in silico to show their efficiency in terms of computation time and their efficacy with respect to the number of nodes to be perturbed.

Sequential Temporary and Permanent Control of Boolean Networks

“…Temporary control applies perturbations for sufficient time and then is released, while permanent control maintains the perturbations for all the following time steps. Benefited from the extended intervention effects, temporary and permanent controls can potentially reduce the number of perturbations, which makes experiments easier to carry out and less costly [26].…”

“…We have developed efficient methods to tackle the minimal OI, OT and OP control [21,22,26], as well as ASI control [11,12]. Considering the advantages of sequential control and temporary and permanent perturbations, in this paper, we shall develop methods to solve the AST and ASP control problems based on the methods for the minimal OT and OP control.…”

“…Attractor-based sequential control is more practical than the general sequential control [12], where any state can play the role of intermediate states. Moreover, temporary and permanent controls have proved their potential in reducing the number of perturbations [26]. In this work, we continue to develop efficient methods to solve the AST and ASP control problems.…”

“…Instantaneous perturbations are applied instantaneously; temporary perturbations are applied for sufficient time steps and then released; permanent perturbations are applied for all the following steps. So far, we have developed methods for the minimal one-step instantaneous, temporary and permanent control (OI, OT and OP) [21,22,26] and the attractor-based sequential instantaneous control (ASI) [11]. In this work, we focus on the attractor-based sequential temporary and permanent control (AST and ASP).…”

“…In this work, we continue to develop efficient methods to solve the AST and ASP control problems. We apply our methods for AST and ASP to several biological networks to show their ability in finding new control paths with fewer perturbations compared to our previous methods [21,22,26,11]. We believe our new methods can provide a better understanding of the mechanismof-action of interventions and improve the efficiency of translating identified reprogramming paths into practical applications.…”