overlap: Overlap of ISA biclusters

In eisa: Expression data analysis via the Iterative Signature Algorithm

Description

Plots a network, where each node is a module and modules that overlap are closer to each other.

Usage

overlap (modules, algorithm = c("mds", "fr", "drl"), edge.limit = 0.5)
overlapPlot (graph, xsize = 400, ysize = 400, vertex.size = 20,
     vertex.size2 = 10, ...)

Arguments

modules	An ISAModules object.
algorithm	The algorithm to use for placing the vertices, a character scalar. See details below.
edge.limit	Numeric constant between zero and one, only edges between modules that have a Pearson correlation higher than edge.limit will be drawn.
graph	An igraph object, as returned by overlap.
xsize	The width of the plot in pixels, only used to calculate the return value, it does not influence the plot itself.
ysize	The height of the plot in pixels, only used to calculate the return value, it does not influence the plot itself.
vertex.size	The width of the vertices on the plot.
vertex.size2	The height of the vertices on the plot.
...	Additional arguments, these are passed to the plot.igraph function from the igraph package.

Details

An ISAModules object may potentially contain many modules that overlap. These functions visualize the overlapping relationships of a set of modules.

overlap creates an igraph graph with additional information on how to plot this graph in a way that nodes representing overlapping modules are close to each other.

overlapPlot takes such a graph and plots it.

overlap can use various algorithms, depending on the algorithm argument. If it is ‘mds’, then multi-dimensional scaling is used, by calling the isaMDS function in the MASS package. If it is ‘fr’, then the Fruchterman-Reingold algorithm is used, through the layout.fruchterman.reingold function of the igraph package. If it is ‘drl’, then the DrL graph layout algorithm is used, see the layout.drl function in the igraph0 package.

Value

overlap returns an igraph graph.

overlapPlot returns the coordinates of the vertices in a two-column matrix, invisibly.

Author(s)

Gabor Csardi [e-mail]

References

Bergmann S, Ihmels J, Barkai N: Iterative signature algorithm for the analysis of large-scale gene expression data Phys Rev E Stat Nonlin Soft Matter Phys. 2003 Mar;67(3 Pt 1):031902. Epub 2003 Mar 11.

Examples

data(ALLModulesSmall)
G <- overlap(ALLModulesSmall, algorithm="drl", edge.limit=0.3)
if (interactive()) {
  overlapPlot(G)
}

eisa documentation built on Nov. 8, 2020, 6:47 p.m.