compHclust.heatmap {compHclust}

Heat Map for Complementary Hierarchical Clustering

Description

Displays a heat map of X, a dendrogram of the clustering of the columns of X and a bar plot of the relative gene importances.

Usage

compHclust.heatmap(x, xhc, gi, d.title = "Cluster Dendrogram",
                   hm.lab = TRUE, hm.lab.cex = 1, d.ht = 0.25,
                   gi.width = 0.5, d.mar = c(0, 4, 4, 2),
                   hm.mar = c(5, 4, 2, 2))

Arguments

x

A numeric matrix X, where interest lies in clustering its columns.

xhc

An object of class hclust, specifically, a hierarchical clustering of the columns of X.

gi

A vector of the relative gene importances, as returned by compHclust.

d.title

The title for the dendrogram.

hm.lab

Logical. If true, the columns of the heat map are labeled with column numbers.

hm.lab.cex

The magnification to be used for the column labels relative to the current setting of cex. See axis and par.

d.ht

The relative height of the plotting region for the dendrogram. Note that the relative height of the plotting region for the heat map is set to 1. See layout.

gi.width

The relative width of the plotting region for the relative gene importance plot. Note that the relative width of the plotting region for the heat map is set to 1. See layout.

d.mar

The margins of the plotting region for the dendrogram. See par.

hm.mar

The margins of the plotting region for the heat map. See par.

Details

Given a numeric matrix X, a hierarchical clustering of the columns of X and a vector of the relative gene importances as returned by compHclust, this function displays a heat map of X with a dendrogram above and a bar plot of the relative gene importances to the right. The columns of X are reordered to correspond with the leaves of the dendrogram.

This function can be fragile - depending on the dimensions of X, some of the arguments such as the margins, heights and widths of the plotting regions may need to be tweaked in order for the figure to look nice. However, it provides a quick and easy way of displaying the output of compHclust and seeing which genes (rows) may be most influential in the clustering of the samples (columns).

For examples of its usage, see the help file for compHclust.

Author(s)

Gen Nowak [e-mail] and Robert Tibshirani

References

Nowak, G. and Tibshirani, R. (2008) Complementary hierarchical clustering. Biostatistics, 9(3), 467–483.

See Also

compHclust