The following recent applied papers demonstrate that WGCNA (and screening based on module eigengene based connectivity) leads to more significant findings than those of a standard differential expression network analysis:
The utility of the network edge orienting (NEO) software for suggesting causal directions is demonstrated in Farber et al (2009). Another application (Plaisier et al 2009) shows how WGCNA and NEO facilitate a systems genetic analysis based on human adipose microarray samples and genetic marker data.
January 2, 2009
A bioinformatics paper by Peter Langfelder and Steve Horvath describes a new R package (called WGCNA) which presents a collection of new and improved R functions for carrying out different aspects of gene coexpression network analysis. The paper can be found here (LINK TO PAPER) and the R package and tutorials can be found here (LINK TO WEBPAGE)
October 21, 2008
A recent publication by Oldham et al (2008) LINK TO PAPER highlights the value of WGCNA for annotating genes with regard to coexpression module membership. This is the first comprehensive analysis of gene coexpression relationships in human cerebral cortex, caudate nucleus and cerebellum. The results demonstrate that the transcriptomes of human brain regions are robustly organized into modules of coexpressed genes that reflect the underlying cellular composition of brain tissue. This makes use of fuzzy module membership measures which are highly related to intramodular connectivity (Dong and Horvath 2008). The fuzzy module membership measures (and intramodular connectivity) can be used i) to determine whether a gene is close to one or more modules, ii) to determine whether a module is preserved across data, iii) and to find differentially connected genes. The paper also demonstrates the use of of dynamic tree cutting for module detection and the use of eigengene networks to describe relationships between coexpression modules.
August 15, 2008
We have published a rather theoretical paper on the relationship between co-expression networks and standard statistical microarray data analysis methods: Geometric Interpretation of Gene Co-Expression Network Analysis LINK TO PAPER. In this article, we show how coexpression network language affects our understanding of biology. For example, there are geometric reasons why highly connected hub genes in important coexpression modules tend to be important, and why hub genes in one module cannot be hubs in another distinct module. We provide a short dictionary for translating between microarray data analysis language and network theory language to facilitate communication between the two fields. We describe several examples that illustrate how the two data analysis fields can inform each other. LINK TO WEBPAGE
April 15, 2008
We have published an article on orienting (also known as directing) the edges of a quantitative trait or gene co-expression network if genetic marker data (SNPs) are available. See the reference Aten JE, Fuller TF, Lusis AJ, Horvath S (2008) Using genetic markers to orient the edges in quantitative trait networks: the NEO software. BMC Systems Biology 2008, 2:34 LINK TO PAPER. Accompanying R software can be found here LINK TO WEBPAGE
March 24, 2008
We are releasing a user-friendly, menu driven stand-alone software (called WGCNA), which can be used to identify significant modules and pathways. The software does *not* require knowledge of R software commands but it requires that R be installed. LINK TO WEBPAGE
November 21, 2007
Two new methods papers have come out. The first article describes the dynamic tree cut algorithm that is used to cut branches of a hierarchical cluster tree (Langfelder P, Zhang B, Horvath S 2007) LINK TO WEBPAGE. Different from the default method that uses the same cutting height for each branch, this algorithm automatically chooses a height cut-off by taking the cluster shape into account.
The second paper describes how to use module eigengenes to construct a network between modules (Langfelder Horvath 2007) LINK TO WEBPAGE.