Toward Intracellular Delivery and Drug Discovery: Stochastic Logic Networks as Efficient Computational Models for Gene Regulatory Networks

Abstract

Biological functions are regulated through the interactions among genes, proteins and other molecules in a cell. Among various approaches to modeling gene regulatory networks (GRNs), Boolean networks (BNs) and its probabilistic extension, probabilistic Boolean networks (PBNs), have been effective means; in particular, PBNs consider molecular and genetic noise, so they provide significant insights into the understanding of the dynamics of GRNs. The applications of PBNs, however, are hindered by the complexities involved in the computation of the state transition matrix and steady-state distribution of a PBN. This chapter discusses stochastic logic networks as computationally efficient gene network models. Initially, stochastic Boolean networks (SBNs) are presented as a novel implementation of PBNs. SBNs are based on the notions of stochastic logic and stochastic computation. To further exploit the simplicity of logical models, a multiple-valued network employs gene states that are not limited to binary values, thus providing a finer granularity in the modeling of GRNs. Subsequently, stochastic multiple-valued networks (SMNs) are presented for modeling the effects of noise and gene perturbation in a GRN. These novel logical models provide accurate and efficient simulations of probabilistic Boolean and multiple-valued networks (PBNs and PMNs). The analysis of a p53–Mdm2 network and a WNT5A network shows that the stochastic logic networks are efficient in evaluating the network dynamics and steady state distribution of gene networks under random gene perturbation. These techniques are potentially useful in the investigation of intracellular delivery and drug discovery.