Gene regulatory network modeling via global optimization of high-order dynamic Bayesian network

Nguyen Xuan Vinh,Madhu Chetty,Ross Coppel,Pramod P Wangikar

doi:10.1186/1471-2105-13-131

Abstract

BackgroundDynamic Bayesian network (DBN) is among the mainstream approaches for modeling various biological networks, including the gene regulatory network (GRN). Most current methods for learning DBN employ either local search such as hill-climbing, or a meta stochastic global optimization framework such as genetic algorithm or simulated annealing, which are only able to locate sub-optimal solutions. Further, current DBN applications have essentially been limited to small sized networks.ResultsTo overcome the above difficulties, we introduce here a deterministic global optimization based DBN approach for reverse engineering genetic networks from time course gene expression data. For such DBN models that consist only of inter time slice arcs, we show that there exists a polynomial time algorithm for learning the globally optimal network structure. The proposed approach, named GlobalMIT+, employs the recently proposed information theoretic scoring metric named mutual information test (MIT). GlobalMIT+ is able to learn high-order time delayed genetic interactions, which are common to most biological systems. Evaluation of the approach using both synthetic and real data sets, including a 733 cyanobacterial gene expression data set, shows significantly improved performance over other techniques.ConclusionsOur studies demonstrate that deterministic global optimization approaches can infer large scale genetic networks.

Highlights

Dynamic Bayesian network (DBN) is among the mainstream approaches for modeling various biological networks, including the gene regulatory network (GRN)
Two critical limitations when applying the traditional static Bayesian networks (BN) paradigm to the GRN domain are: (i) BN does not have a mechanism for exploiting the temporal aspect of timeseries data abundant in this field; and (ii) BN does not allow the modeling of cyclic phenomena, such as feedback loops, which are prevalent in biological systems [15]
In our recent preliminary work [25], we have shown that this result holds true for the Mutual Information Test (MIT), a novel scoring metric recently introduced for learning static BN [26]

Summary

Introduction

Dynamic Bayesian network (DBN) is among the mainstream approaches for modeling various biological networks, including the gene regulatory network (GRN). Two critical limitations when applying the traditional static BN paradigm to the GRN domain are: (i) BN does not have a mechanism for exploiting the temporal aspect of timeseries data (such as time-series microarray data) abundant in this field; and (ii) BN does not allow the modeling of cyclic phenomena, such as feedback loops, which are prevalent in biological systems [15] These limitations motivated the development of the dynamic Bayesian network (DBN) which has received significant interest from the bioinformatics community [15,16,17,18,19,20,21,22]. DBN exploits the temporal aspect of time series data to infer edge directions, and allows the modeling of feedback loops (in the form of time delayed cyclic interactions)

Objectives

Methods

Results