Heatmap trouble in R

Question

[Jack]

I have a dataset in this format:

id              drug_stat locus   location
5               0          4      755
37133           0         13      755
24086           0          4       41
7               1          4      243
27057           0         13      243
20              1          4      580
21022           1          4      1650

And I want to generate a heat map from it. value1 and value2 are actually the gene's locus type and value2 is location.

my R code for this is the following:

# install libraries
install.packages("gplots")
library(gplots)

# load the dataframe
df <- read.table("data.csv", header = TRUE, sep='\t')
df[df=="n/a"] = NA
df$drug_binary <- ifelse(df$drug_stat == "druggable", 1, 0)
#df[order(df$id_numeric), ]
df_final <- df[c("id_numeric", "drug_binary", "locus_type", "location")]


# change strings to ints for faster read.
df_final$locus_type <- as.numeric(factor(df_final$locus_type))
df_final$location <- as.numeric(factor(df_final$location))

# Plot Heatmap
png(file="plot.png", bg = "transparent")
heat<-heatmap.2(data.matrix(), 
                main = "Sample Heatmap", 
                trace = "none", 
                margin = c(12,30), 
                col = colorRampPalette(c("blue", "yellow", "green")),
                Rowv = NA,
                Colv = NA)

plot(heat)
dev.off()

Question1: Why the binary column is entirely BLUE? It should at least have some GREEN spots in it. Same with the locus.

Question2: Why my plot is not exported as a PNG?

Answer 1

[Irsan]

Because it scales the data. Use the scale="none" option

Answer 2

[ssv.bio]

Q2:Try adding type="cairo", after file name.