\name{plotCorGraph}
\alias{plotCorGraph}
\title{Plots graph using}
\usage{
plotCorGraph(dataGraph, edgeWeight = "cor", nodeAttrs,
nodeShape, nodeCol = "labAn", nodeName = "indComp",
col, shape, title = "", reciproCol = "reciprocal",
tkplot = FALSE, ...)
}
\arguments{
\item{dataGraph}{A data.frame containing the graph
description. It must have two columns \code{n1} and
\code{n2}, each row denoting that there is an edge from
n1 to n2. Node labels in columns \code{n1} and \code{n2}
of \code{dataGraph} must correspond to node IDs in column
\code{id} of \code{nodeAttrs}.}
\item{edgeWeight}{The column of dataGraph used to weight
edges.}
\item{nodeAttrs}{A data.frame with node description, see
function \code{nodeAttrs}.}
\item{nodeShape}{Denotes the column of \code{nodeAttrs}
used to attribute the node shapes.}
\item{nodeCol}{Denotes the column of \code{nodeAttrs}
used to color the nodes in the graph.}
\item{nodeName}{Denotes the column of \code{nodeAttrs}
used as labels for the nodes in the graph.}
\item{col}{A vector of colors, for the nodes, indexed by
the unique elements of \code{nodeCol} column from
\code{nodeAttrs}. If missing, colors will be
automatically attributed.}
\item{shape}{A vector of shapes indexed by the unique
elements of column \code{nodeShape} from
\code{nodeAttrs}. If missing, shapes will be
automatically attributed.}
\item{title}{Title for the plot}
\item{reciproCol}{Denotes the column of \code{dataGraph}
containing \code{TRUE} if the row defines a reciprocal
node, else \code{FALSE}. See
\code{\link{annotReciprocal}}.}
\item{tkplot}{If TRUE, performs interactive plot with
function \code{tkplot}, else uses \code{plot.igraph}.}
\item{\dots}{Additional parameters as required by
\code{tkplot}.}
}
\value{
A list consisting of \describe{ \item{dataGraph}{a
data.frame defining the correlation graph}
\item{nodeAttrs}{a data.frame describing the node of the
graph} \item{graph}{the graph as an object of class
\code{igraph}} \item{graphid}{the id of the graph plotted
using \code{tkplot}} }
}
\description{
This function plots the correlation graph in an
interactive device using function \code{tkplot}.
}
\details{
You have to slighly move the nodes to see cliques because
strongly related nodes are often superimposed. The
\code{edgeWeight} column is used to weight the edges
within the fruchterman.reingold layout available in the
package \code{igraph}.
The argument \code{nodeCol} typically denotes the column
containing the names of the datasets. Colors are
automatically attributed to the nodes using palette Set3
of package \code{RColorBrewer}. The corresponding colors
can be directly specified in the 'col' argument. In that
case, 'col' must be a vector of colors indexed by the
unique elements contained in \code{nodeCol} column (e.g
dataset ids).
As for colors, one can define the column of
\code{nodeAttrs} that is used to define the node shapes.
The corresponding shapes can be directly specified in the
\code{shape} argument. In that case, \code{shape} must be
one of \code{c("circle","square", " vcsquare",
"rectangle", "crectangle", "vrectangle")} and must be
indexed by the unique elements of \code{nodeShape}
column.
Unfortunately, shapes can't be taken into account when
tkplot is TRUE (interactive plot).
If \code{reciproCol} is not missing, it is used to color
the edges, either in grey if the edge is not reciprocal
or in black if the edge is reciprocal.
}
\examples{
dat1 <- data.frame(matrix(rnorm(10000),ncol=10,nrow=1000))
rownames(dat1) <- paste("g", 1:1000, sep="")
colnames(dat1) <- paste("s", 1:10, sep="")
dat2 <- data.frame(matrix(rnorm(10000),ncol=10,nrow=1000))
rownames(dat2) <- paste("g", 1:1000, sep="")
colnames(dat2) <- paste("s", 1:10, sep="")
## run ICA
resJade1 <- runICA(X=dat1, nbComp=3, method = "JADE")
resJade2 <- runICA(X=dat2, nbComp=3, method = "JADE")
## build params
params <- buildMineICAParams(resPath="toy/")
## build IcaSet object
icaSettoy1 <- buildIcaSet(params=params, A=data.frame(resJade1$A), S=data.frame(resJade1$S),
dat=dat1, alreadyAnnot=TRUE)$icaSet
icaSettoy2 <- buildIcaSet(params=params, A=data.frame(resJade2$A), S=data.frame(resJade2$S),
dat=dat2, alreadyAnnot=TRUE)$icaSet
icaSets <- list(icaSettoy1, icaSettoy2)
resCompareAn <- compareAn(icaSets=list(icaSettoy1,icaSettoy2), labAn=c("toy1","toy2"),
type.corr="pearson", level="genes", cutoff_zval=0)
## Build a graph where edges correspond to maximal correlation value (useVal="cor"),
dataGraph <- compareAn2graphfile(listPairCor=resCompareAn, useMax=TRUE, useVal="cor", file="myGraph.txt")
## construction of the data.frame with the node description
nbComp <- rep(3,2) #each IcaSet contains 3 components
nbAn <- 2 # we are comparing 2 IcaSets
# labels of components created as comp*i*
labComp <- foreach(icaSet=icaSets, nb=nbComp, an=1:nbAn) \%do\% {
paste(rep("comp",sum(nb)),1:nbComp(icaSet),sep = "")}
# creation of the data.frame with the node description
nodeDescr <- nodeAttrs(nbAn = nbAn, nbComp = nbComp, labComp = labComp,
labAn = c("toy1","toy2"), file = "nodeInfo.txt")
## Plot correlation graph, slightly move the attached nodes to make the cliques visible
## use tkplot=TRUE to have an interactive graph
res <- plotCorGraph(title = "Compare toy 1 and 2", dataGraph = dataGraph, nodeName = "indComp", tkplot = FALSE,
nodeAttrs = nodeDescr, edgeWeight = "cor", nodeShape = "labAn", reciproCol = "reciprocal")
\dontrun{
## load two microarray datasets
library(breastCancerMAINZ)
library(breastCancerVDX)
data(mainz)
data(vdx)
## Define a function used to build two examples of IcaSet objects
treat <- function(es, annot="hgu133a.db") {
es <- selectFeatures_IQR(es,10000)
exprs(es) <- t(apply(exprs(es),1,scale,scale=FALSE))
colnames(exprs(es)) <- sampleNames(es)
resJade <- runICA(X=exprs(es), nbComp=10, method = "JADE", maxit=10000)
resBuild <- buildIcaSet(params=buildMineICAParams(), A=data.frame(resJade$A), S=data.frame(resJade$S),
dat=exprs(es), pData=pData(es), refSamples=character(0),
annotation=annot, typeID= typeIDmainz,
chipManu = "affymetrix", mart=mart)
icaSet <- resBuild$icaSet
}
## Build the two IcaSet objects
icaSetMainz <- treat(mainz)
icaSetVdx <- treat(vdx)
icaSets <- list(icaSetMainz, icaSetVdx)
labAn <- c("Mainz", "Vdx")
## correlations between gene projections of each pair of IcaSet
resCompareAn <- compareAn(icaSets = icaSets, level = "genes", type.corr= "pearson",
labAn = labAn, cutoff_zval=0)
## construction of the correlation graph using previous output
dataGraph <- compareAn2graphfile(listPairCor=resCompareAn, useMax=TRUE, file="corGraph.txt")
## construction of the data.frame with the node description
nbComp <- rep(10,2) #each IcaSet contains 10 components
nbAn <- 2 # we are comparing 2 IcaSets
# labels of components created as comp*i*
labComp <- foreach(icaSet=icaSets, nb=nbComp, an=1:nbAn) \%do\% {
paste(rep("comp",sum(nb)),1:nbComp(icaSet),sep = "")}
# creation of the data.frame with the node description
nodeDescr <- nodeAttrs(nbAn = nbAn, nbComp = nbComp, labComp = labComp,
labAn = labAn, file = "nodeInfo.txt")
## Plot correlation graph, slightly move the attached nodes to make the cliques visible
res <- plotCorGraph(title = "Compare two ICA decomsitions obtained on \\n two
microarray-based data of breast tumors", dataGraph = dataGraph, nodeName = "indComp",
nodeAttrs = nodeDescr, edgeWeight = "cor", nodeShape = "labAn", reciproCol = "reciprocal")
}
}
\author{
Anne Biton
}
\seealso{
\code{\link{compareAn}}, \code{\link{nodeAttrs}},
\code{\link{compareAn2graphfile}},
\code{\link{runCompareIcaSets}}
}
