Reconstruction of gene networks using Bayesian learning and manipulation experiments

Pournara, Iosifina; Wernisch, Lorenz

doi:10.1093/bioinformatics/bth337

Cited by 70 publications

(44 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Reverse engineering gene networks from microarray time course data is a well studied problem in the literature. Techniques have been proposed to reverse engineer gene networks based on Bayesian networks [9,10,1], undirected Gaussian graphical models [11,12], ordinary differential equations [13], partial correlations [2], and others. Comparisons of different methods used for reverse engineering gene networks have been performed [14,15].…”

Section: Related Workmentioning

confidence: 99%

Inferring Gene Interaction Networks from ISH Images via Kernelized Graphical Models

Puniyani¹,

Xing²

2012

Computer Vision – ECCV 2012

View full text Add to dashboard Cite

Abstract. New bio-technologies are being developed that allow highthroughput imaging of gene expressions, where each image captures the spatial gene expression pattern of a single gene in the tissue of interest. This paper addresses the problem of automatically inferring a gene interaction network from such images. We propose a novel kernel-based graphical model learning algorithm, that is both convex and consistent. The algorithm uses multi-instance kernels to compute similarity between the expression patterns of different genes, and then minimizes the L1 regularized Bregman divergence to estimate a sparse gene interaction network. We apply our algorithm on a large, publicly available data set of gene expression images of Drosophila embryos, where our algorithm makes novel and interesting predictions.

show abstract

Section: Related Workmentioning

confidence: 99%

Inferring Gene Interaction Networks from ISH Images via Kernelized Graphical Models

Puniyani¹,

Xing²

2012

Computer Vision – ECCV 2012

View full text Add to dashboard Cite

show abstract

“…In most applications, this is a reasonable assumption. For example, the assumption of having a unique steady state distribution is made implicitly in practically every biological application where, e.g., static transcriptional regulatory network models, such as BNs, are learned from steady state gene expression or protein level data (Pe'er et al 2001;Hartemink et al 2002;Imoto et al 2003;Dobra et al 2004;Pournara and Wernisch 2004;Wille et al 2004;Sachs et al 2005;Schäfer and Strimmer 2005;Werhli et al 2006). For a more complete list of previous work, see (Markowetz 2007).…”

Section: Steady State Analysis Of Dbnsmentioning

confidence: 99%

“…DBNs and their non-temporal versions, i.e., static Bayesian networks (BN), have successfully been used in different modeling problems, such as in speech recognition, target tracking and identification, genetics, probabilistic expert systems, and medical diagnostic systems (see, e.g., Cowell et al 1999, and the references therein). Recently, BNs and DBNs have also been intensively studied in the context of modeling genomic regulation, see, e.g., (Hartemink et al 2001(Hartemink et al , 2002Husmeier 2003;Imoto et al 2003;Friedman 2004;Pournara and Wernisch 2004;Sachs et al 2005;Bernard and Hartemink 2005;Werhli et al 2006;Lähdesmäki et al 2006).…”

Section: Introductionmentioning

confidence: 99%

Learning the structure of dynamic Bayesian networks from time series and steady state measurements

Lähdesmäki

Shmulevich

2008

Mach Learn

View full text Add to dashboard Cite

Dynamic Bayesian networks (DBN) are a class of graphical models that has become a standard tool for modeling various stochastic time-varying phenomena. In many applications, the primary goal is to infer the network structure from measurement data. Several efficient learning methods have been introduced for the inference of DBNs from time series measurements. Sometimes, however, it is either impossible or impractical to collect time series data, in which case, a common practice is to model the non-time series observations using static Bayesian networks (BN). Such an approach is obviously sub-optimal if the goal is to gain insight into the underlying dynamical model. Here, we introduce Bayesian methods for the inference of DBNs from steady state measurements. We also consider learning the structure of DBNs from a combination of time series and steady state measurements. We introduce two different methods: one that is based on an approximation and another one that provides exact computation. Simulation results demonstrate that dynamic network structures can be learned to an extent from steady state measurements alone and that inference from a combination of steady state and time series data has the potential to improve learning performance relative to the inference from time series data alone. Keywords

show abstract

“…These models have been used to describe dynamical evolution of expression timecourses, and involve a range of strategies for fitting weight coefficients, ranging from direct least squares (D'haeseleer et al, 1999(D'haeseleer et al, , van Someren et al, 2000 to imposing constraints such as sparseness (Yeung et al, 2002). With the growth in gene expression measurement technologies, interest in method development for the extraction of networks from gene expression data continues to increase (Pournara and Wernisch, 2004, Yu et al, 2004, Bickel, 2005, Schäfer and Strimmer, 2005.…”

Section: Introductionmentioning

confidence: 99%

Reverse Engineering Galactose Regulation in Yeast through Model Selection

Thórsson¹,

Hörnquist²,

Siegel³

et al. 2005

Statistical Applications in Genetics and Molecular Biology

View full text Add to dashboard Cite

We examine the application of statistical model selection methods to reverse-engineering the control of galactose utilization in yeast from DNA microarray experiment data. In these experiments, relationships among gene expression values are revealed through modifications of galactose sugar level and genetic perturbations through knockouts. For each gene variable, we select predictors using a variety of methods, taking into account the variance in each measurement. These methods include maximization of log-likelihood with Cp, AIC, and BIC penalties, bootstrap and cross-validation error estimation, and coefficient shrinkage via the Lasso.

show abstract

Reconstruction of gene networks using Bayesian learning and manipulation experiments

Cited by 70 publications

References 9 publications

Inferring Gene Interaction Networks from ISH Images via Kernelized Graphical Models

Inferring Gene Interaction Networks from ISH Images via Kernelized Graphical Models

Learning the structure of dynamic Bayesian networks from time series and steady state measurements

Reverse Engineering Galactose Regulation in Yeast through Model Selection

Contact Info

Product

Resources

About