Heatmaps
Zanin Pavel
March 11, 2016
Link to project files on GitHUB
Link to projects on RPubs
Introduction
Here I’m going to show some basic heatmaps.
Simple heatmap 1
library(gplots)
library(RColorBrewer)
# 1. Reading in data and transform it to matrix format
data <- read.csv("dataset.csv", comment.char="#")
rnames <- data[,1] # assign labels in column 1 to "rnames"
mat_data <- data.matrix(data[,2:ncol(data)]) # transform column 2-5 into a matrix
rownames(mat_data) <- rnames # assign row names
# 2. Customizing and plotting heatmap
# 2.1 Creates a own color palette from red to green
my_palette <- colorRampPalette(c("white", "#FCF14C", "red"))(n = 799)
# 2.2 Defining the color breaks manually for a "skewed" color transition
breaks <- c(seq(-1, 0,length=300),seq(0.001,0.5,length=300),seq(0.501,1,length=200))
# 2.3 Creating heatmap
heatmap.2(mat_data,
cellnote = mat_data, # same data set for cell labels
main = "Correlation", # heat map title
notecol="black", # change font color of cell labels to black
density.info="none", # turns off density plot inside color legend
trace="none", # turns off trace lines inside the heat map
margins =c(12,9), # widens margins around plot
col=my_palette, # use on color palette defined earlier
breaks=breaks, # enable color transition at specified limits
dendrogram="row", # only draw a row dendrogram
Colv="NA") # turn off column clustering)
dev.off()
## null device
## 1Let’s make the same graph by d3heatmap package:
library(d3heatmap)
d3heatmap(mat_data, scale="column", color = scales::col_quantile("Reds", NULL, 5))Heatmap with clustering
# 1. Reading in data and transform it to matrix format
data <- read.csv("dataset.csv", comment.char="#")
rnames <- data[,1] # assign labels in column 1 to "rnames"
mat_data <- data.matrix(data[,2:ncol(data)]) # transform column 2-5 into a matrix
rownames(mat_data) <- rnames # assign row names
# 2. Customizing and plotting heatmap
# 2.1 Creates a own color palette from red to green
my_palette <- colorRampPalette(c("white", "#FCF14C", "red"))(n = 799)
# 2.2 Defining the color breaks manually for a "skewed" color transition
breaks <- c(seq(-1, 0,length=300),seq(0.001,0.5,length=300),seq(0.501,1,length=200))
# 3. Changes the distance measure and clustering method
# NOTE: Matrix here not symmetrical. For symmetrical matrices
# only one distance and cluster could and SHOULD be defined.
# Distance options: euclidean (default), maximum, canberra, binary, minkowski, manhattan
# Cluster options: complete (default), single, average, mcquitty, median, centroid, ward
row_distance = dist(mat_data, method = "manhattan")
row_cluster = hclust(row_distance, method = "ward")
col_distance = dist(t(mat_data), method = "manhattan")
col_cluster = hclust(col_distance, method = "ward")
heatmap.2(mat_data,
cellnote = mat_data, # same data set for cell labels
main = "Correlation", # heat map title
notecol = "black", # change font color of cell labels to black#
density.info = "none", # turns off density plot inside color legend
trace = "none", # turns off trace lines inside the heat map
margins = c(12,9), # widens margins around plot
col = my_palette, # use on color palette defined earlier
breaks = breaks, # enable color transition at specified limits
Rowv = as.dendrogram(row_cluster), # apply default clustering method
Colv = as.dendrogram(col_cluster)) # apply default clustering method
dev.off()
## null device
## 1Heatmap with clustering and categorazing
library(gplots)
library(RColorBrewer)
# 1. Reading in data and transform it to matrix format
data <- read.csv("dataset.csv", comment.char="#")
rnames <- data[,1] # assign labels in column 1 to "rnames"
mat_data <- data.matrix(data[,2:ncol(data)]) # transform column 2-5 into a matrix
rownames(mat_data) <- rnames # assign row names
# 2. Customizing and plotting heatmap
# 2.1 Creates a own color palette from red to green
my_palette <- colorRampPalette(c("white", "#FCF14C", "red"))(n = 799)
# 2.2 Defining the color breaks manually for a "skewed" color transition
breaks <- c(seq(-1, 0,length=300),seq(0.001,0.5,length=300),seq(0.501,1,length=200))
# 3. Changes the distance measure and clustering method
# NOTE: Matrix here not symmetrical. For symmetrical matrices
# only one distance and cluster could and SHOULD be defined.
# Distance options: euclidean (default), maximum, canberra, binary, minkowski, manhattan
# Cluster options: complete (default), single, average, mcquitty, median, centroid, ward
row_distance = dist(mat_data, method = "manhattan")
row_cluster = hclust(row_distance, method = "ward")
col_distance = dist(t(mat_data), method = "manhattan")
col_cluster = hclust(col_distance, method = "ward")
heatmap.2(mat_data,
cellnote = mat_data, # same data set for cell labels
RowSideColors = c( # grouping row-variables into different
rep("blue", 3), # categories, Measurement 1-3: green
rep("black", 5), # Measurement 4-8: blue
rep("green", 2)), # Measurement 9-10: red
main = "Categories", # heat map title
notecol="black", # change font color of cell labels to black
density.info="none", # turns off density plot inside color legend
trace="none", # turns off trace lines inside the heat map
margins =c(16,16), # widens margins around plot
col=my_palette, # use on color palette defined earlier
breaks=breaks, # enable color transition at specified limits
Rowv = as.dendrogram(row_cluster), # apply default clustering method
Colv = as.dendrogram(col_cluster)) # apply default clustering method
# adding a color legend for the categories
par(lend = 1) # square line ends for the color legend
legend("topright", # location of the legend on the heatmap plot
legend = c("A", "B", "C"), # category labels
col = c("blue", "black", "green"), # color key
lty = 1, # line style
lwd = 10, # line width
) 
dev.off()
## null device
## 1