High-order dynamic Bayesian Network learning with hidden common causes for causal gene regulatory network

Abstract

BackgroundInferring gene regulatory network (GRN) has been an important topic in Bioinformatics. Many computational methods infer the GRN from high-throughput expression data. Due to the presence of time delays in the regulatory relationships, High-Order Dynamic Bayesian Network (HO-DBN) is a good model of GRN. However, previous GRN inference methods assume causal sufficiency, i.e. no unobserved common cause. This assumption is convenient but unrealistic, because it is possible that relevant factors have not even been conceived of and therefore un-measured. Therefore an inference method that also handles hidden common cause(s) is highly desirable. Also, previous methods for discovering hidden common causes either do not handle multi-step time delays or restrict that the parents of hidden common causes are not observed genes.ResultsWe have developed a discrete HO-DBN learning algorithm that can infer also hidden common cause(s) from discrete time series expression data, with some assumptions on the conditional distribution, but is less restrictive than previous methods. We assume that each hidden variable has only observed variables as children and parents, with at least two children and possibly no parents. We also make the simplifying assumption that children of hidden variable(s) are not linked to each other. Moreover, our proposed algorithm can also utilize multiple short time series (not necessarily of the same length), as long time series are difficult to obtain.ConclusionsWe have performed extensive experiments using synthetic data on GRNs of size up to 100, with up to 10 hidden nodes. Experiment results show that our proposed algorithm can recover the causal GRNs adequately given the incomplete data. Using the limited real expression data and small subnetworks of the YEASTRACT network, we have also demonstrated the potential of our algorithm on real data, though more time series expression data is needed.

Analysis of high-resolution 3D intrachromosomal interactions aided by Bayesian network modeling

“…While BNs are generally well established in several biomedical research areas (genomics, expression data, metabolomics, etc.) ( 12 , 32 – 34 ), they have not been used for chromatin interaction analysis. Therefore, a brief introduction is in order.…”

A novel method of gene regulatory network structure inference from gene knock-out expression data

“…Model-based methods usually formulate a computational model of the system and further learn the parameters of such a model. Typical computational models include Boolean network [28][29][30][31] , Bayesian network [17,[32][33][34][35] , and differential equation models [15,[36][37][38][39][40] . The Boolean network model is the simplest network model, which is implemented through Boolean variables and Boolean logic.…”

New Algorithm and Software (BNOmics) for Inferring and Visualizing Bayesian Networks from Heterogeneous Big Biological and Genetic Data

“…BN modeling, and network implementation in general, has been extensively used in genetics, bioinformatics, and computational biology since the turn of the century (Friedman et al, 2000; Pe'er, 2005; Rodin et al, 2005; Djebbari and Quackenbush, 2008; Rodin et al, 2012; Liu et al, 2014; Lo et al, 2015; Tasaki et al, 2015; Li et al, 2016, to name but a few). A detailed treatment of BN methodology, while outside of the scope of this communication, can be found in Pearl (1988), Pearl (2000), Heckerman (1995), and Chickering et al (2004).…”