colorhcplot {colorhcplot}

Colorful Hierarchical Clustering Dendrograms

Description

This function takes a "hclust-class" object and a factor describing the groups as arguments and generates colorful dendrograms in which leaves belonging to different groups are identified by colors. This function produces a plot that allows to easily detect if leaves that clustered together also belong to the same group.

Usage

colorhcplot(hc, fac, hang = 0.1, 
            main = "Cluster Dendrogram", 
            colors = NULL, lab.cex = 1, 
            ylim = NULL, lwd = 3, 
            las = 1, lab.mar = 0.55)

Arguments

hc

hclust-class object, typically the result of a 'hclust()' function call.

fac

factor that defines the grouping.

hang

hang value, as in hclust. hang is the fraction of the plot height by which labels should hang below the rest of the plot. A negative value will align all labels at the bottom of the plot.

main

title of the dendrogram plot.

colors

NULL or a character vector of length 1 or having the same length as the number of levels in fac. This argument defines the palette for the plot.

lab.cex

numeric value for adjusting the font size of the leaf labels (and legend text).

ylim

numeric, defines the minimum and maximum value of the y-axis of the plot.

lwd

numeric value that defines the width (in points) of the lines of the dendogram.

las

graphic parameter for the orientation of the y-axis tick labels.

lab.mar

fraction of the plot area that is reserved for the labels (at the bottom of the plot).

Details

In order to generate a colorful dendrogram, the colorhcplot() function requires 2 mandatory arguments: hc and fac. hc is the result of a hclust() call, while fac is a factor defining the groups. The number of leaves of the dendrogram has to be identical to the length of fac.

Value

Calling colorhcplot() returns a colorful dendrogram plot

Note

Online colorhcplot() function reference at: http://www.biotechworld.it/bioinf/2015/09/30/colorful-hierarchical-clustering-dendrograms-with-r

Author(s)

Damiano Fantini <[e-mail]>

See Also

hclust

Examples

### Example 1, using the USArrests dataset
data(USArrests)
hc <- hclust(dist(USArrests), "ave")
fac <- as.factor(c(rep("group 1", 10), 
                   rep("group 2", 10), 
                   rep("unknown", 30)))
plot(hc)
colorhcplot(hc, fac)
colorhcplot(hc, fac, hang = -1, lab.cex = 0.8)

### Example 2: use the "ward.D2" algorithm and
### the UScitiesD dataset
data(UScitiesD)
hcity.D2 <- hclust(UScitiesD, "ward.D2")
fac.D2 <-as.factor(c(rep("group1", 3), 
                     rep("group2", 7)))
plot(hcity.D2, hang=-1)
colorhcplot(hcity.D2, fac.D2, color = c("chartreuse2", "orange2"))
colorhcplot(hcity.D2, fac.D2, color = "gray30", lab.cex = 1.2, lab.mar = 0.75)

### Example 3: use gene expression data 
data(geneData, package="colorhcplot")
exprs <- geneData$exprs
fac <- geneData$fac
hc <- hclust(dist(t(exprs)))
colorhcplot(hc, fac, main ="default", col = "gray10")
colorhcplot(hc, fac, main="Control vs. Tumor Samples")