Co-expression clustering

This analysis is provided through the R-language integration into Perseus and therefore requires R as well as the WGCNA package to be installed. Visit WGCNA¹ page for more information and installation instructions.

More information about co-expression clustering can be found at the following resources:

• official WGCNA tutorials.
• Network analysis with WGCNA
• WGCNA - RNA-seq

1 Description

The co-expression network is created using the defined correlation function. The determined power is applied to the network (see Soft-threshold for more info). Topological overlap distance is used to create the hierarchical clustering dendrogram. The co-expression modules are determined using the dynamic tree-cut method. For each module, a module eigengene is reported, with its name corresponding to the color of the cluster.

2 Output

• Hierarchical clustering heatmap with a dendrogram and automatic cluster assignments.
• Matrix of module eigengenes that represent a cluster. See Correlate for identifying clusters that correlate with clinical/phenotype data.

References

1.

Langfelder, P. & Horvath, S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics 9, (2008).