3.1 plot a DNA sequence logo with different fonts and colors

Users can select different fonts and colors to draw the sequence logo.

suppressPackageStartupMessages(library(motifStack))
pcm <- read.table(file.path(find.package("motifStack"), 
                            "extdata", "bin_SOLEXA.pcm"))
pcm <- pcm[,3:ncol(pcm)]
rownames(pcm) <- c("A","C","G","T")
motif <- new("pcm", mat=as.matrix(pcm), name="bin_SOLEXA")
##pfm object
#motif <- pcm2pfm(pcm)
#motif <- new("pfm", mat=motif, name="bin_SOLEXA")
opar<-par(mfrow=c(4,1))
plot(motif)
#plot the logo with same height
plot(motif, ic.scale=FALSE, ylab="probability")
#try a different font
plot(motif, font="mono,Courier")
#try a different font and a different color group
motif@color <- colorset(colorScheme='basepairing')
plot(motif,font="Times")

Figure 1: Plot a DNA sequence logo with different fonts and colors

par(opar)

3.2 plot a RNA sequence logo

To plot a RNA sequence logo, you only need to change the rowname of the matrix from “T” to “U” as follows.

rna <- pcm
rownames(rna)[4] <- "U"
motif <- new("pcm", mat=as.matrix(rna), name="RNA_motif")
plot(motif)

Figure 2: Plot an RNA sequence logo

3.3 plot an amino acid sequence logo

Given that motifStack allows to use any letters as symbols, it can also be used to draw amino acid sequence logos.

library(motifStack)
protein<-read.table(file.path(find.package("motifStack"),"extdata","cap.txt"))
protein<-t(protein[,1:20])
motif<-pcm2pfm(protein)
motif<-new("pfm", mat=motif, name="CAP", 
            color=colorset(alphabet="AA",colorScheme="chemistry"))
plot(motif)

Figure 3: Plot an sequence logo with any symbols as you want such as amino acid sequence logo

3.4 plot an affinity logo

It can also be used to draw affinity logos given a position specific affinity matrix (PSAM) as described by Foat et al.².

library(motifStack)
motif<-matrix(
  c(
    .846, .631, .593, .000, .000, .000, .434, .410, 1.00, .655, .284, .000, .000, .771, .640, .961,
    .625, .679, .773, 1.00, 1.00, .000, .573, .238, .397, 1.00, 1.00, .000, .298, 1.00, 1.00, .996,
    1.00, 1.00, 1.00, .228, .000, 1.00, 1.00, .597, .622, .630, .000, 1.00, 1.00, .871, .617, 1.00,
    .701, .513, .658, .000, .000, .247, .542, 1.00, .718, .686, .000, .000, .000, .595, .437, .970
  ), nrow=4, byrow = TRUE)
rownames(motif) <- c("A", "C", "G", "T")
motif<-new("psam", mat=motif, name="affinity logo")
plot(motif)

Figure 4: Plot an affinity logo

3.5 plot sequence logo stack

To show multiple motifs on the same canvas as a sequence logo stack, the distance of motifs need to be calculated first. By default, MotIV³::motifDistances ( R implementation of STAMP⁴) is used to calculate the distance. After alignment, users can use plotMotifLogoStack, plotMotifLogoStackWithTree or plotMotifStackWithRadialPhylog to draw sequence logos in different layouts. To make it easy to use, we integrated different functionalities into one workflow function named as motifStack.

suppressPackageStartupMessages(library(motifStack))
#####Input#####
motifs<-importMatrix(dir(file.path(find.package("motifStack"), "extdata"),"pcm$", full.names = TRUE))

## plot stacks
motifStack(motifs, layout="stack", ncex=1.0)

Figure 5: Plot motifs with sequence logo stack style

motif2 <- motif
motif2$mat <- motif$mat[, 5:12]
motif2$name <- "logo2"
psamMotifs <- list(motif, motif2)
motifStack(psamMotifs)

Figure 6: Plot affinity logos with sequence logo stack style

## plot stacks with hierarchical tree
motifStack(motifs, layout="tree")

Figure 7: Sequence logo stack with hierarchical cluster tree

## When the number of motifs is too much to be shown in a vertical stack, 
## motifStack can draw them in a radial style.
## random sample from MotifDb
library("MotifDb")

## See system.file("LICENSE", package="MotifDb") for use restrictions.

matrix.fly <- query(MotifDb, "Dmelanogaster")
motifs2 <- as.list(matrix.fly)
## use data from FlyFactorSurvey
motifs2 <- motifs2[grepl("Dmelanogaster\\-FlyFactorSurvey\\-",
                         names(motifs2))]
## format the names
names(motifs2) <- gsub("Dmelanogaster_FlyFactorSurvey_", "",
                       gsub("_FBgn\\d+$", "",
                            gsub("[^a-zA-Z0-9]","_",
                                 gsub("(_\\d+)+$", "", names(motifs2)))))
motifs2 <- motifs2[unique(names(motifs2))]
pfms <- sample(motifs2, 50)
## creat a list of object of pfm 
motifs2 <- lapply(names(pfms), 
                  function(.ele, pfms){new("pfm",mat=pfms[[.ele]], name=.ele)}
                  ,pfms)
## trim the motifs
motifs2 <- lapply(motifs2, trimMotif, t=0.4)
## setting colors
library(RColorBrewer)
color <- brewer.pal(12, "Set3")
## plot logo stack with radial style
motifStack(motifs2, layout="radialPhylog", 
           circle=0.3, cleaves = 0.2, 
           clabel.leaves = 0.5, 
           col.bg=rep(color, each=5), col.bg.alpha=0.3, 
           col.leaves=rep(color, each=5),
           col.inner.label.circle=rep(color, each=5), 
           inner.label.circle.width=0.05,
           col.outer.label.circle=rep(color, each=5), 
           outer.label.circle.width=0.02, 
           circle.motif=1.2,
           angle=350)

Figure 8: Plot motifs in a radial style when the number of motifs is too much to be shown in a vertical stack

3.6 plot a sequence logo cloud

We can also plot a sequence logo cloud for DNA motifs.

## assign groups for motifs
groups <- rep(paste("group",1:5,sep=""), each=10)
names(groups) <- names(pfms)
## assign group colors
group.col <- brewer.pal(5, "Set3")
names(group.col)<-paste("group",1:5,sep="")
## use MotIV to calculate the distances of motifs
jaspar.scores <- MotIV::readDBScores(file.path(find.package("MotIV"), 
                                               "extdata", 
                                               "jaspar2010_PCC_SWU.scores"))
d <- MotIV::motifDistances(lapply(pfms, pfm2pwm))
hc <- MotIV::motifHclust(d, method="average")
## convert the hclust to phylog object
phylog <- hclust2phylog(hc)
## reorder the pfms by the order of hclust
leaves <- names(phylog$leaves)
pfms <- pfms[leaves]
## create a list of pfm objects
pfms <- lapply(names(pfms), function(.ele, pfms){
                                new("pfm",mat=pfms[[.ele]], name=.ele)}
               ,pfms)
## extract the motif signatures
motifSig <- motifSignature(pfms, phylog, groupDistance=0.01, min.freq=1)
## draw the motifs with a tag-cloud style.
motifCloud(motifSig, scale=c(6, .5), 
           layout="rectangles", 
           group.col=group.col, 
           groups=groups, 
           draw.legend=TRUE)

Figure 9: Sequence logo cloud with rectangle packing layout

3.7 plot grouped sequence logo

Grouped sequence logo can also be plotted in radial phylogeny tree style.

## get the signatures from object of motifSignature
sig <- signatures(motifSig)
## set the inner-circle color for each signature
gpCol <- sigColor(motifSig)
## plot the logo stack with radial style.
plotMotifStackWithRadialPhylog(phylog=phylog, pfms=sig, 
                              circle=0.4, cleaves = 0.3, 
                              clabel.leaves = 0.5, 
                              col.bg=rep(color, each=5), col.bg.alpha=0.3, 
                              col.leaves=rep(rev(color), each=5),
                              col.inner.label.circle=gpCol, 
                              inner.label.circle.width=0.03,
                              angle=350, circle.motif=1.2, 
                              motifScale="logarithmic")

Figure 10: Grouped sequence logo with radialPhylog style layout

3.8 motifCircos

We can also plot it with circos style. In circos style, we can plot two group of motifs and with multiple color rings.

## plot the logo stack with cirsoc style.
motifCircos(phylog=phylog, pfms=pfms, pfms2=sig, 
            col.tree.bg=rep(color, each=5), col.tree.bg.alpha=0.3, 
            col.leaves=rep(rev(color), each=5),
            col.inner.label.circle=gpCol, 
            inner.label.circle.width=0.03,
            col.outer.label.circle=gpCol, 
            outer.label.circle.width=0.03,
            r.rings=c(0.02, 0.03, 0.04), 
            col.rings=list(sample(colors(), 50), 
                           sample(colors(), 50), 
                           sample(colors(), 50)),
            angle=350, motifScale="logarithmic")

Figure 11: Grouped sequence logo with circos style layout

3.9 motifPiles

We can also plot the motifs in pile style. In pile style, we can plot two group of motifs with multiple types of annotation.

## plot the logo stack with pile style.
motifPiles(phylog=phylog, pfms=pfms, pfms2=sig, 
            col.tree=rep(color, each=5),
            col.leaves=rep(rev(color), each=5),
            col.pfms2=gpCol, 
            r.anno=c(0.02, 0.03, 0.04), 
            col.anno=list(sample(colors(), 50), 
                          sample(colors(), 50), 
                          sample(colors(), 50)),
            motifScale="logarithmic",
            plotIndex=TRUE,
            groupDistance=0.01)

Figure 12: Grouped sequence logo with piles style layout

We can also use motifPiles to plot motifs with a heatmap.

## plot the logo stack with heatmap.
df <- data.frame(A=runif(n = 50), B=runif(n = 50), C=runif(n = 50), D=runif(n = 50))
map2col <- function(x, pal){
  rg <- range(x)
  pal[findInterval(x, seq(rg[1], rg[2], length.out = length(pal)+1), 
                   all.inside = TRUE)]
}
dl <- lapply(df, map2col, pal=heat.colors(10))
motifPiles(phylog=phylog, pfms=pfms, 
            col.tree=rep(color, each=5),
            col.leaves=rep(rev(color), each=5),
            col.pfms2=gpCol, 
            r.anno=rep(0.02, length(dl)), 
            col.anno=dl,
            motifScale="logarithmic",
            plotIndex=TRUE,
            groupDistance=0.01)

Figure 13: Grouped sequence logo with a heatmap