| plotNW {GSCA} | R Documentation |
This function is a custom wrapper of plot.graph implemented in
Rgraphviz. It draws a network from a given correlation matrix. Nodes
can be colored by DE information, and edges can be colored based on
the correlation magnitude and direction.
plotNW(genes = NULL, cormatrix, node.de = NULL, ncolor = "red", ecolor = NULL, ewidth = 1, gtype = "neato", ...)
genes |
Node names to be displayed in the network plot. If
NULL (default), rownames(cormatrix) is used. |
cormatrix |
A symmetric correlation matrix to draw a network from. |
node.de |
A vector of 0's and 1's. The "1" nodes are colored in
ncolor. |
ncolor |
Node color for selected genes specified by node.de. |
ecolor |
Edge color palette. The default is bluered(201)
implemented by gplots, which sets c("blue", "white", "red")
for the correlation of c(-1, 0, 1). |
ewidth |
Edge width. Default is set to 1. |
gtype |
Graph type (layout) to be passed to
plot.graph. One of dot, neato, twopi,
circo, and fdp. The default is neato. |
... |
The rest of arguments are passed to plot.graph. |
YounJeong Choi
Choi and Kendziorski (2009)
data(LungCancer3)
GS <- LungCancer3$info$GSdef
GSdesc <- LungCancer3$info$Name
setid <- "GO:0019216"
gid <- GS[[setid]]
ss <- c("SERPINA3", "SOD1", "SCAP", "NPC2", "ADIPOQ", "PRKAA1", "AGT",
"PPARA", "BMP6", "BRCA1")
plotNW(genes = ss, cormatrix = cor(t(LungCancer3$data$Michigan[gid, 87:96]), use = "pairwise.complete.obs"), node.de = c(rep(1, 5), rep(0, 5)), ncolor = "yellow", ecolor = bluered(201), ewidth = 5, gtype = "circo")