diffFactors: Performs differential analysis of chroGPS-factors maps based...

In chroGPS: chroGPS2: Generation, visualization and differential analysis of epigenome maps

Description

The function uses Procrustes analysis to compare two chroGPS-factors MDS solutions, providing a visual map highlighting differences and a ranked list of Procrustes squared errors between analog replicated factors in both maps.

Usage

diffFactors(m1,m2,name1='mds1',name2='mds2',minPoints=10,poslegend='topleft',plot=TRUE,pointcol.m1='red',pointcol.m2='green',textcol.m1='black',textcol.m2='grey',pch.m1=19,pch.m2=19,cex.m1=1,cex.m2=1,segcol='blue',...)

Arguments

m1	A mds object, containing a chroGPS-factors MDS map.
m2	A mds object, containing a chroGPS-factors MDS map, must be conformable with m1 (same number of elements and in the same order in the MDS points).
name1	Name for m1, by default 'mds1'
name2	Name for m2, by default 'mds2'
minPoints	Minimum number of points in each of the MDS objects.
poslegend	Position for the map legend, passed on to legend function.
plot	Wether to plot the differential map or not (in case we only want to retrieve Procrustes errors).
pointcol.m1	Color vector for points in MDS1.
pointcol.m2	Color vector for points in MDS2.
textcol.m1	Color vector for labels in MDS1. Labels are taken from MDS points rownames.
textcol.m2	Color vector for labels in MDS2. Labels are taken from MDS points rownames.
pch.m1	Pch for MDS1 points.
pch.m2	Pch for MDS2 points.
cex.m1	Cex for MDS1 points.
cex.m2	Cex for MDS2 points.
segcol	Color for segments joining replicates from both MDS.
...	Further arguments passed on to the plot function.

Value

The function returns a data frame with Procrustes errors between paired replicates in both chroGPS-factors MDS maps.

Author(s)

Oscar Reina.

See Also

See functions distGPS, mds for generating chroGPS-factors maps.

Examples

## Not run

data(s2)
data(repliSeq)
library(gplots)

# Modify colors and add some transparency
fnames <- s2names$Factor
s2names$Color[s2names$Color=='grey'] <- 'orange'
fcolors <- paste(col2hex(s2names$Color),'BB',sep='')
bcolors <- paste(col2hex(s2names$Color),'FF',sep='')

# Select time points to compare
m1 <- m.origs[['Early.Mid']]
m2 <- m.origs[['Late']]

## Perform differential Procrustes analysis
pp <- diffFactors(m1,m2)

## Plot both maps before and after adjustment
m3 <- pp$mds3
plot(0,xlim=c(-1,1),ylim=c(-1,1),xlab='',ylab='',xaxt='n',yaxt='n',col='NA')
segments(m1@points[,1],m1@points[,2],m3@points[,1],m3@points[,2],col='red')
par(new=TRUE)
plot(m1,drawlabels=TRUE,labels=s2names$Factor,point.pch=19,point.cex=4,text.cex=0.75,point.col=s2names$Color,main=sprintf('S2@Origins, adjusted (Avgdist-isoMDS)'),font=2,xlim=c(-1,1),ylim=c(-1,1))
par(new=TRUE)
plot(m3,drawlabels=TRUE,labels=s2names$Factor,point.pch=19,point.cex=4,text.cex=0.75,point.col=s2names$Darkcolor,text.col='grey',main='',xaxt='n',yaxt='n',font=2,xlim=c(-1,1),ylim=c(-1,1))

## Plot Procrustes errors
pp <- pp$procrustes
par(las=1,mar=c(4,12,4,4)); barplot(sort(residuals(pp),decr=TRUE),horiz=TRUE,xlim=c(0,max(residuals(pp))+.1),col=heat.colors(length(residuals(pp))),main='Procrustes errors')
hist(residuals(pp),breaks=50)

chroGPS documentation built on Oct. 31, 2019, 4:52 a.m.