generate heatmap based on pair-wise distance matrix using R

If you have a pair-wise distance matrix which was gained outside R, how could you generate heatmap based on this matrix. Here is an example

1. The pair-wise distance matrix outside R

2. This is the heatmap generated by R ggplot2 package.

3. R code

# load packages needed

library(ggplot2)



# read in the data frame including pair-wise distances

dist.m <- read.csv("dist_matrix.csv", header=T)

head(dist.m) # in fact, dist.m is a data.frame of R



# convert the data.frame of distance to a real matrix of distance

# be care of setting the right row/column names

dist.m2 <- as.matrix(dist.m[,2:13], dimnames=lsit(as.vector(dist.m[,1]), as.vector(dist.m[,1])))

row.names(dist.m2) <- as.vector(dist.m[,1])



head(dist.m2)

dist.m2[1,1]



# convert the distance matrix to data.frame 

dist.m3 <- melt(dist.m2)

dist.m3 <- dist.m3

head(dist.m3)



# set the order of rows/columns

dist.m3$X1 <- factor(dist.m3$X1, levels = c("Ath", "AL3b", "AL13a", "AL34a", "AL25b", "AL8b", "AL10b",

                      "AL38a", "AL35a", "AL4a", "AL17a", "AL16a"))

dist.m3$X2 <- factor(dist.m3$X2, levels = c("Ath", "AL3b", "AL13a", "AL34a", "AL25b", "AL8b", "AL10b",

                      "AL38a", "AL35a", "AL4a", "AL17a", "AL16a"))



# set up heatmap plot

pdf("heatmap.pdf")

g.m <- ggplot(dist.m3, aes(X1, X2)) + geom_tile(aes(fill=value))

g.m

dev.off()

Created by Pretty R at inside-R.org