How to infer gene networks from expression profiles

Abstract

Correction to: Molecular Systems Biology 3:78. doi:; Published online 13 February 2007

A GPU-based solution for fast calculation of the betweenness centrality in large weighted networks

“…We also downloaded a large synthetic network with 2 20 vertices and more than 44 million edges. These networks are publicly available on the Internet and have been analyzed extensively in previous studies (Rossi & Ahmed, 2015;Bansal et al, 2007;Palla et al, 2008;Barabási & Albert, 1999;Leskovec & Krevl, 2014;De Domenico et al, 2013;Leskovec, Adamic & Huberman, 2007;Bader et al, 2012Bader et al, , 2014. The details of these networks are listed in Table 1.…”

“…For networks with low average degrees, such as ca-MathSciNet-dir, rt-higgs and mt-higgs, applying the warp-centric method with the real WARP_SIZE (32) is always inefficient because the nodes' degrees are always smaller than WARP_SIZE. Using a smaller virtual WARP_SIZE enables better performance on (Rossi & Ahmed, 2015;Bansal et al, 2007) 14,340 9,041,364 7,230 1,261.00 Human gene regulatory network bio-mouse-gene (Rossi & Ahmed, 2015;Bansal et al, 2007) 45, 101 14,506,196 8,033 643.28 Mouse gene regulatory network ca-MathSciNet-dir (Rossi & Ahmed, 2015;Palla et al, 2008) Table 2, and we will also further demonstrate this later.…”

Computational and Systems Biology

“…Several reverseengineering methods have been developed to extract transcription hypotheses from microarray data. [1][2][3] They model the data in transcription regulatory networks that describe gene expression at a systems level as a function of regulatory inputs specified by interactions between regulatory proteins and DNA. The basic assumption is that regulatory proteins are themselves regulated by transcription, so that their expression profiles provide information about their activity level.…”

Statistical Methods for Inference of Genetic Networks and Regulatory Modules

“…Large-scale microarray gene expression data, promotor sequences data and genome-wide chromatin immunoprecipitation (ChIP-chip) data provide the possibility of learning gene regulation and constructing the gene regulatory networks and pathways or cellular networks (Ideker et al, 2001;Friedman, 2004;Das et al, 2006). In a recent review, Bansal et al (2007) summarized the methods for inferring genetic networks into two broad classes: those based in the "physical interaction" approach that aims at identifying interactions among transcription factors and their target genes and those based on the "influence interaction" approach that aims to relate the expression of a gene to the expression of the other genes in the cell, rather than relating it to the sequence motif found in its promotor. In this paper, we review some recently developed statistical methods for several problems related to inferences of genetic network and regulatory modules, including both "physical interaction" networks using diverse data sets and "influence interaction" networks using gene expression data alone.…”

“…The goal of such probabilistic modeling is to investigate the patterns of association in order to generate biological insights plausibly related to underlying biological and regulatory pathways. It is important to note that the interaction between two genes in a gene network defined by such graphical models does not necessarily imply a physical interaction, but can refer to an indirect regulation via proteins, metabolites and ncRNA that have been measured directly and therefore its interpretation depends on the model formulations (Bansal et al, 2007). In this paper, we will present some details on Gaussian graphical models and methods for estimating the graphical structure in the high-dimensional settings.…”