made4 (version 1.46.0)

heatplot: Draws heatmap with dendrograms.

Description

heatplot calls heatmap.2 using a red-green colour scheme by default. It also draws dendrograms of the cases and variables using correlation similarity metric and average linkage clustering as described by Eisen. heatplot is useful for a quick overview or exploratory analysis of data

Usage

heatplot(dataset, dend = c("both", "row", "column", "none"),  cols.default = TRUE, lowcol = "green", highcol = "red", scale="none",  classvec=NULL, classvecCol=NULL, classvec2=NULL, distfun=NULL, returnSampleTree=FALSE,method="ave", dualScale=TRUE, zlim=c(-3,3),scaleKey=TRUE, ...)

Arguments

dataset

a matrix, data.frame, ExpressionSet or marrayRaw-class. If the input is gene expression data in a matrix or data.frame. The rows and columns are expected to contain the variables (genes) and cases (array samples) respectively.

dend

A character indicating whether dendrograms should be drawn for both rows and columms "both", just rows "row" or column "column" or no dendrogram "none". Default is both.

cols.default

Logical. Default is TRUE. Use blue-brown color scheme.

lowcol, highcol

Character indicating colours to be used for down and upregulated genes when drawing heatmap if the default colors are not used, that is cols.default = FALSE.

scale

Default is row. Scale and center either "none","row", or "column").

classvec,classvec2

A factor or vector which describes the classes in columns or rows of the dataset. Default is NULL. If included, a color bar including the class of each column (array sample) or row (gene) will be drawn. It will automatically add to either the columns or row, depending if the length(as.character(classvec)) ==nrow(dataset) or ncol(dataset).

classvecCol

A vector of length the number of levels in the factor classvec. These are the colors to be used for the row or column colorbar. Colors should be in the same order, as the levels(factor(classvec))

distfun

A character, indicating function used to compute the distance between both rows and columns. Defaults to 1- Pearson Correlation coefficient

method

The agglomeration method to be used. This should be one of '"ward"', '"single"','"complete"', '"average"', '"mcquitty"', '"median"' or '"centroid"'. See hclust for more details. Default is "ave"

dualScale

A logical indicating whether to scale the data by row and columns. Default is TRUE

zlim

A vector of length 2, with lower and upper limits using for scaling data. Default is c(-3,3)

scaleKey

A logical indicating whether to draw a heatmap color-key bar. Default is TRUE

returnSampleTree

A logical indicating whether to return the sample (column) tree. If TRUE it will return an object of class dendrogram. Default is FALSE

...

further arguments passed to or from other methods.

Value

dendrogram which can be manipulated using

Details

The hierarchical plot is produced using average linkage cluster analysis with a correlation metric distance. heatplot calls heatmap.2 in the R package gplots.

NOTE: We have changed heatplot scaling in made4 (v 1.19.1) in Bioconductor v2.5. Heatplot by default dual scales the data to limits of -3,3. To reproduce older version of heatplot, use the parameters dualScale=FALSE, scale="row".

References

Eisen MB, Spellman PT, Brown PO and Botstein D. (1998). Cluster Analysis and Display of Genome-Wide Expression Patterns. Proc Natl Acad Sci USA 95, 14863-8.

See Also

See also as hclust, heatmap and dendrogram

Examples

data(khan)

## Change color scheme
heatplot(khan$train[1:30,])
heatplot(khan$train[1:30,], cols.default=FALSE, lowcol="white", highcol="red")

## Add labels to rows, columns
heatplot(khan$train[1:26,], labCol = c(64:1), labRow=LETTERS[1:26])

## Add a color bar
heatplot(khan$train[1:26,], classvec=khan$train.classes)
heatplot(khan$train[1:26,], classvec=khan$train.classes, classvecCol=c("magenta", "yellow", "cyan", "orange"))

## Change the scaling to the older made4 version (pre Bioconductor 2.5)
heatplot(khan$train[1:26,], classvec=khan$train.classes, dualScale=FALSE, scale="row")


## Getting the members of a cluster and manuipulating the tree
sTree<-heatplot(khan$train, classvec=khan$train.classes, returnSampleTree=TRUE)
class(sTree)
plot(sTree)

## Cut the tree at the height=1.0
lapply(cut(sTree,h=1)$lower, labels)

## Zoom in on the first cluster
plot(cut(sTree,1)$lower[[1]])
str(cut(sTree,1.0)$lower[[1]])



## Visualizing results from an ordination using heatplot
if (require(ade4, quiet = TRUE)) {
res<-ord(khan$train, ord.nf=5)  # save 5 components from correspondence analysis
khan.coa = res$ord
}

# Provides a view of the components of the Correspondence analysis (gene projection) 
heatplot(khan.coa$li,  dend="row", dualScale=FALSE)    # first 5 components, do not cluster columns, only rows.

# Provides a view of the components of the Correspondence analysis (sample projection) 
# The difference between tissues and cell line samples are defined in the first axis.
heatplot(khan.coa$co, margins=c(4,20), dend="row")  # Change the margin size. The default is c(5,5)

# Add a colorbar, change the heatmap color scheme and no scaling of data
heatplot(khan.coa$co,classvec2=khan$train.classes, cols.default=FALSE, lowcol="blue", dend="row", dualScale=FALSE)
apply(khan.coa$co,2, range)