semiPheatmap: A function to draw clustered heatmaps.

In campbio/celda: CEllular Latent Dirichlet Allocation

A function to draw clustered heatmaps.

Description

A function to draw clustered heatmaps where one has better control over some graphical parameters such as cell size, etc.

The function also allows to aggregate the rows using kmeans clustering. This is advisable if number of rows is so big that R cannot handle their hierarchical clustering anymore, roughly more than 1000. Instead of showing all the rows separately one can cluster the rows in advance and show only the cluster centers. The number of clusters can be tuned with parameter kmeansK.

Usage

semiPheatmap(
  mat,
  color = colorRampPalette(rev(brewer.pal(n = 7, name = "RdYlBu")))(100),
  kmeansK = NA,
  breaks = NA,
  borderColor = "grey60",
  cellWidth = NA,
  cellHeight = NA,
  scale = "none",
  clusterRows = TRUE,
  clusterCols = TRUE,
  clusteringDistanceRows = "euclidean",
  clusteringDistanceCols = "euclidean",
  clusteringMethod = "complete",
  clusteringCallback = .identity2,
  cutreeRows = NA,
  cutreeCols = NA,
  treeHeightRow = ifelse(clusterRows, 50, 0),
  treeHeightCol = ifelse(clusterCols, 50, 0),
  legend = TRUE,
  legendBreaks = NA,
  legendLabels = NA,
  annotationRow = NA,
  annotationCol = NA,
  annotation = NA,
  annotationColors = NA,
  annotationLegend = TRUE,
  annotationNamesRow = TRUE,
  annotationNamesCol = TRUE,
  dropLevels = TRUE,
  showRownames = TRUE,
  showColnames = TRUE,
  main = NA,
  fontSize = 10,
  fontSizeRow = fontSize,
  fontSizeCol = fontSize,
  displayNumbers = FALSE,
  numberFormat = "%.2f",
  numberColor = "grey30",
  fontSizeNumber = 0.8 * fontSize,
  gapsRow = NULL,
  gapsCol = NULL,
  labelsRow = NULL,
  labelsCol = NULL,
  fileName = NA,
  width = NA,
  height = NA,
  silent = FALSE,
  rowLabel,
  colLabel,
  rowGroupOrder = NULL,
  colGroupOrder = NULL,
  ...
)

Arguments

mat	numeric matrix of the values to be plotted.
color	vector of colors used in heatmap.
kmeansK	the number of kmeans clusters to make, if we want to agggregate the rows before drawing heatmap. If NA then the rows are not aggregated.
breaks	Numeric vector. A sequence of numbers that covers the range of values in the normalized 'counts'. Values in the normalized 'matrix' are assigned to each bin in 'breaks'. Each break is assigned to a unique color from 'col'. If NULL, then breaks are calculated automatically. Default NULL.
borderColor	color of cell borders on heatmap, use NA if no border should be drawn.
cellWidth	individual cell width in points. If left as NA, then the values depend on the size of plotting window.
cellHeight	individual cell height in points. If left as NA, then the values depend on the size of plotting window.
scale	character indicating if the values should be centered and scaled in either the row direction or the column direction, or none. Corresponding values are "row", "column" and "none".
clusterRows	boolean values determining if rows should be clustered or hclust object,
clusterCols	boolean values determining if columns should be clustered or hclust object.
clusteringDistanceRows	distance measure used in clustering rows. Possible values are "correlation" for Pearson correlation and all the distances supported by dist, such as "euclidean", etc. If the value is none of the above it is assumed that a distance matrix is provided.
clusteringDistanceCols	distance measure used in clustering columns. Possible values the same as for clusteringDistanceRows.
clusteringMethod	clustering method used. Accepts the same values as hclust.
clusteringCallback	callback function to modify the clustering. Is called with two parameters: original hclust object and the matrix used for clustering. Must return a hclust object.
cutreeRows	number of clusters the rows are divided into, based on the hierarchical clustering (using cutree), if rows are not clustered, the argument is ignored
cutreeCols	similar to cutreeRows, but for columns
treeHeightRow	the height of a tree for rows, if these are clustered. Default value 50 points.
treeHeightCol	the height of a tree for columns, if these are clustered. Default value 50 points.
legend	logical to determine if legend should be drawn or not.
legendBreaks	vector of breakpoints for the legend.
legendLabels	vector of labels for the legendBreaks.
annotationRow	data frame that specifies the annotations shown on left side of the heatmap. Each row defines the features for a specific row. The rows in the data and in the annotation are matched using corresponding row names. Note that color schemes takes into account if variable is continuous or discrete.
annotationCol	similar to annotationRow, but for columns.
annotation	deprecated parameter that currently sets the annotationCol if it is missing.
annotationColors	list for specifying annotationRow and annotationCol track colors manually. It is possible to define the colors for only some of the features. Check examples for details.
annotationLegend	boolean value showing if the legend for annotation tracks should be drawn.
annotationNamesRow	boolean value showing if the names for row annotation tracks should be drawn.
annotationNamesCol	boolean value showing if the names for column annotation tracks should be drawn.
dropLevels	logical to determine if unused levels are also shown in the legend.
showRownames	boolean specifying if column names are be shown.
showColnames	boolean specifying if column names are be shown.
main	the title of the plot
fontSize	base fontsize for the plot
fontSizeRow	fontsize for rownames (Default: fontsize)
fontSizeCol	fontsize for colnames (Default: fontsize)
displayNumbers	logical determining if the numeric values are also printed to the cells. If this is a matrix (with same dimensions as original matrix), the contents of the matrix are shown instead of original values.
numberFormat	format strings (C printf style) of the numbers shown in cells. For example "%.2f" shows 2 decimal places and "%.1e" shows exponential notation (see more in sprintf).
numberColor	color of the text
fontSizeNumber	fontsize of the numbers displayed in cells
gapsRow	vector of row indices that show shere to put gaps into heatmap. Used only if the rows are not clustered. See cutreeRow to see how to introduce gaps to clustered rows.
gapsCol	similar to gapsRow, but for columns.
labelsRow	custom labels for rows that are used instead of rownames.
labelsCol	similar to labelsRow, but for columns.
fileName	file path where to save the picture. Filetype is decided by the extension in the path. Currently following formats are supported: png, pdf, tiff, bmp, jpeg. Even if the plot does not fit into the plotting window, the file size is calculated so that the plot would fit there, unless specified otherwise.
width	manual option for determining the output file width in inches.
height	manual option for determining the output file height in inches.
silent	do not draw the plot (useful when using the gtable output)
rowLabel	row cluster labels for semi-clustering
colLabel	column cluster labels for semi-clustering
rowGroupOrder	Vector. Specifies the order of feature clusters when semisupervised clustering is performed on the y labels.
colGroupOrder	Vector. Specifies the order of cell clusters when semisupervised clustering is performed on the z labels.
...	graphical parameters for the text used in plot. Parameters passed to grid.text, see gpar.

Value

Invisibly a list of components

• treeRow the clustering of rows as hclust object

• treeCol the clustering of columns as hclust object

• kmeans the kmeans clustering of rows if parameter kmeansK was specified

Author(s)

Raivo Kolde <[e-mail]> #@examples # Create test matrix test = matrix(rnorm(200), 20, 10) test[seq(10), seq(1, 10, 2)] = test[seq(10), seq(1, 10, 2)] + 3 test[seq(11, 20), seq(2, 10, 2)] = test[seq(11, 20), seq(2, 10, 2)] + 2 test[seq(15, 20), seq(2, 10, 2)] = test[seq(15, 20), seq(2, 10, 2)] + 4 colnames(test) = paste("Test", seq(10), sep = "") rownames(test) = paste("Gene", seq(20), sep = "")

# Draw heatmaps pheatmap(test) pheatmap(test, kmeansK = 2) pheatmap(test, scale = "row", clusteringDistanceRows = "correlation") pheatmap(test, color = colorRampPalette(c("navy", "white", "firebrick3"))(50)) pheatmap(test, cluster_row = FALSE) pheatmap(test, legend = FALSE)

# Show text within cells pheatmap(test, displayNumbers = TRUE) pheatmap(test, displayNumbers = TRUE, numberFormat = "%.1e") pheatmap(test, displayNumbers = matrix(ifelse(test > 5, "*", ""), nrow(test))) pheatmap(test, cluster_row = FALSE, legendBreaks = seq(-1, 4), legendLabels = c("0", "1e-4", "1e-3", "1e-2", "1e-1", "1"))

# Fix cell sizes and save to file with correct size pheatmap(test, cellWidth = 15, cellHeight = 12, main = "Example heatmap") pheatmap(test, cellWidth = 15, cellHeight = 12, fontSize = 8, fileName = "test.pdf")

# Generate annotations for rows and columns annotationCol = data.frame(CellType = factor(rep(c("CT1", "CT2"), 5)), Time = seq(5)) rownames(annotationCol) = paste("Test", seq(10), sep = "")

annotationRow = data.frame(GeneClass = factor(rep(c("Path1", "Path2", "Path3"), c(10, 4, 6)))) rownames(annotationRow) = paste("Gene", seq(20), sep = "")

# Display row and color annotations pheatmap(test, annotationCol = annotationCol) pheatmap(test, annotationCol = annotationCol, annotationLegend = FALSE) pheatmap(test, annotationCol = annotationCol, annotationRow = annotationRow)

# Specify colors ann_colors = list(Time = c("white", "firebrick"), CellType = c(CT1 = "#1B9E77", CT2 = "#D95F02"), GeneClass = c(Path1 = "#7570B3", Path2 = "#E7298A", Path3 = "#66A61E"))

pheatmap(test, annotationCol = annotationCol, annotationColors = ann_colors, main = "Title") pheatmap(test, annotationCol = annotationCol, annotationRow = annotationRow, annotationColors = ann_colors) pheatmap(test, annotationCol = annotationCol, annotationColors = ann_colors[2])

# Gaps in heatmaps pheatmap(test, annotationCol = annotationCol, clusterRows = FALSE, gapsRow = c(10, 14)) pheatmap(test, annotationCol = annotationCol, clusterRows = FALSE, gapsRow = c(10, 14), cutreeCol = 2)

# Show custom strings as row/col names labelsRow = c("", "", "", "", "", "", "", "", "", "", "", "", "", "", "", "", "", "Il10", "Il15", "Il1b")

pheatmap(test, annotationCol = annotationCol, labelsRow = labelsRow)

# Specifying clustering from distance matrix drows = stats::dist(test, method = "minkowski") dcols = stats::dist(t(test), method = "minkowski") pheatmap(test, clusteringDistanceRows = drows, clusteringDistanceCols = dcols)

# Modify ordering of the clusters using clustering callback option callback = function(hc, mat) sv = svd(t(mat))$v[, 1] dend = reorder(as.dendrogram(hc), wts = sv) as.hclust(dend)

pheatmap(test, clusteringCallback = callback)

campbio/celda documentation built on April 5, 2024, 11:47 a.m.