Review on statistical methods for gene network reconstruction using expression data

Wang, Y. X. Rachel; Huang, Haiyan

doi:10.1016/j.jtbi.2014.03.040

Cited by 173 publications

(126 citation statements)

References 129 publications

Supporting

Mentioning

122

Contrasting

Unclassified

Order By: Relevance

“…This is known as the "large p and small n" problem and remains a central challenge in the eld of statistics (Kuismin & Sillanpää, 2016). Although these kinds of data structures are common in elds such as genomics (Y. R. Wang & Huang, 2014) and quantitative nance (Ledoit & Wolf, 2004a, 2004b, they are the exception in psychology. Nonetheless, in psychology, 1 penalized maximum likelihood has emerged as the default estimation method (Epskamp & Fried, 2016).…”

Section: Gaussian Graphical Modelmentioning

confidence: 99%

Back to the Basics: Rethinking Partial Correlation Network Methodology

Williams¹,

Rast²

2018

Preprint

View full text Add to dashboard Cite

Gaussian graphical models are an increasingly popular technique in psychology to characterize relationships among observed variables. These relationships are represented as covariances in the precision matrix. Standardizing this covariance matrix and reversing the sign yields corresponding partial correlations that imply pairwise dependencies in which the e ects of all other variables have been controlled for. In order to estimate the precision matrix, the graphical lasso (glasso) has emerged as the default estimation method, which uses 1 -based regularization. Glasso was developed and optimized for high dimensional settings where the number of variables (p) exceeds the number of observations (n) which are uncommon in psychological applications. Here we propose to go "back to the basics", wherein the precision matrix is rst estimated with non-regularized maximum likelihood and then Fisher Z-transformed con dence intervals are used to determine non-zero relationships. We rst show the exact correspondence between the con dence level and speci city, which is due to 1 -speci city denoting the false positive rate (i.e., α). With simulations in low-dimensional settings (p n), we then demonstrate superior performance compared to glasso for determining conditional relationships, in addition to frequentist risk measured with various loss functions. Further, our results indicate that glasso is inconsistent for the purpose of model selection, whereas the proposed method converged on the true model with a probability that approached 100%. We end by discussing implications for estimating Gaussian graphical models in psychology.

show abstract

Section: Gaussian Graphical Modelmentioning

confidence: 99%

Back to the Basics: Rethinking Partial Correlation Network Methodology

Williams¹,

Rast²

2018

Preprint

View full text Add to dashboard Cite

show abstract

“…In the case of a negative regulatory interaction, the expression profile of Y is expected to be highly similar to the inverted expression profile of X shifted forward in time by m time units ( Figure 1C). Hence, by analyzing statistical linkage between gene expression profiles, it should be possible in principle to infer regulatory relationships, and it has already been successfully done in diverse biological systems (De Smet and Marchal, 2010;Wang and Huang, 2014).…”

Section: Introductionmentioning

confidence: 99%

“…Many GRN inference algorithms have been developed over the last decade, based on different formalisms and having different specificities and abilities (De Smet and Marchal, 2010;Marbach et al, 2010;Wang and Huang, 2014). In their comparative study as part of the DREAM3 challenge, Marbach et al (2010) pointed out important limitations of the GRN inference approach.…”

Section: Introductionmentioning

confidence: 99%

Inference of the Arabidopsis Lateral Root Gene Regulatory Network Suggests a Bifurcation Mechanism That Defines Primordia Flanking and Central Zones

et al. 2015

View full text Add to dashboard Cite

A large number of genes involved in lateral root (LR) organogenesis have been identified over the last decade using forward and reverse genetic approaches in Arabidopsis thaliana. Nevertheless, how these genes interact to form a LR regulatory network largely remains to be elucidated. In this study, we developed a time-delay correlation algorithm (TDCor) to infer the gene regulatory network (GRN) controlling LR primordium initiation and patterning in Arabidopsis from a time-series transcriptomic data set. The predicted network topology links the very early-activated genes involved in LR initiation to later expressed cell identity markers through a multistep genetic cascade exhibiting both positive and negative feedback loops. The predictions were tested for the key transcriptional regulator AUXIN RESPONSE FACTOR7 node, and over 70% of its targets were validated experimentally. Intriguingly, the predicted GRN revealed a mutual inhibition between the ARF7 and ARF5 modules that would control an early bifurcation between two cell fates. Analyses of the expression pattern of ARF7 and ARF5 targets suggest that this patterning mechanism controls flanking and central zone specification in Arabidopsis LR primordia.

show abstract

“…Associations between two genes are routinely described by their correlation to each other (43). In terms of spatial relationships, positive gene interactions exhibit spatial overlap whereas repressive gene interactions exhibit spatial exclusivity.…”

Section: Pp-based Correlation Network Inference Leads To Accurate De mentioning

confidence: 99%

Stability-driven nonnegative matrix factorization to interpret spatial gene expression and build local gene networks

Joseph

Hammonds

et al. 2016

Proc. Natl. Acad. Sci. U.S.A.

125

107

View full text Add to dashboard Cite

Spatial gene expression patterns enable the detection of local covariability and are extremely useful for identifying local gene interactions during normal development. The abundance of spatial expression data in recent years has led to the modeling and analysis of regulatory networks. The inherent complexity of such data makes it a challenge to extract biological information. We developed staNMF, a method that combines a scalable implementation of nonnegative matrix factorization (NMF) with a new stability-driven model selection criterion. When applied to a set of Drosophila early embryonic spatial gene expression images, one of the largest datasets of its kind, staNMF identified 21 principal patterns (PP). Providing a compact yet biologically interpretable representation of Drosophila expression patterns, PP are comparable to a fate map generated experimentally by laser ablation and show exceptional promise as a data-driven alternative to manual annotations. Our analysis mapped genes to cell-fate programs and assigned putative biological roles to uncharacterized genes. Finally, we used the PP to generate local transcription factor regulatory networks. Spatially local correlation networks were constructed for six PP that span along the embryonic anterior–posterior axis. Using a two-tail 5% cutoff on correlation, we reproduced 10 of the 11 links in the well-studied gap gene network. The performance of PP with the Drosophila data suggests that staNMF provides informative decompositions and constitutes a useful computational lens through which to extract biological insight from complex and often noisy gene expression data.

show abstract

Review on statistical methods for gene network reconstruction using expression data

Cited by 173 publications

References 129 publications

Back to the Basics: Rethinking Partial Correlation Network Methodology

Back to the Basics: Rethinking Partial Correlation Network Methodology

Inference of the Arabidopsis Lateral Root Gene Regulatory Network Suggests a Bifurcation Mechanism That Defines Primordia Flanking and Central Zones

Stability-driven nonnegative matrix factorization to interpret spatial gene expression and build local gene networks

Contact Info

Product

Resources

About