basic barplot with ggplot: assign specific color to column with specific values

Question

[camillab.]

Hi,

I want to draw a barplot with basic ggplot but I would like to color the bar in the plot (so genes) that have a specific value (e.g. above 1) with a different color from the other. I found this webiste where they show how to do it but I was wondering if there is a simple/faster way to do it without me going through the dataset and check which row has the value, I am interested and then change the name and so on.. since I am planning to do multiple barplots for multiple genes?

or if there is any other similar question that I might have missed on the web that could answer my question would be much appreciated.

here my code to generate a normal barplot only with one sample (just to give an idea):

#define new dataset
d1 <- mouse$`Log2FC p0`
x= mouse$GeneSymbl

#tell what to plot
to_plot <- data.frame(x=x, "Mouse"=d1)
melted<-melt(to_plot, id="x")

#plot
f <- ggplot(melted,aes(x=x,y=value,fill=variable)) + geom_bar(stat="identity", colour="black", size=0.4, width = 0.8, position="dodge", show.legend = T)+
  labs(y="LOG2FC", x = "Genes shared", title ="Apoptotic genes comparion") + scale_fill_manual(values=c("blue"))
f + theme(legend.text=element_text(size=8), legend.title = element_blank(), legend.position="bottom", panel.background = element_blank(), axis.line = element_line(colour = "black"),axis.text.x = element_text(face="plain", color="black", size=9, angle=45, hjust=1), axis.text.y = element_text(face="plain", color="black",  size=10, hjust=1))

I have also tried to assign a new column to "melted" dataset but it's not working at all, it doesn't ever show the bars to be blue (like in the previous example):

 #define new dataset
    d1 <- mouse$`Log2FC p0`
    x= mouse$GeneSymbl

#tell what to plot
to_plot <- data.frame(x=x, "Mouse"=d1)
melted<-melt(to_plot, id="x")
melted[["sign"]] = ifelse(melted[["value"]] <= -1, "positive", "negative")

#plot
f <- ggplot(melted,aes(x=x,y=value,fill=variable)) + geom_bar(stat="identity", colour="black", size=0.4, width = 0.8, position="dodge", show.legend = T)+
  labs(y="LOG2FC", x = "Genes shared", title ="Apoptotic genes comparion") + scale_fill_manual(values = c("positive" = "darkblue", "negative" = "red"))
f + theme(legend.text=element_text(size=8), legend.title = element_blank(), legend.position="bottom", panel.background = element_blank(), axis.line = element_line(colour = "black"),axis.text.x = element_text(face="plain", color="black", size=9, angle=45, hjust=1), axis.text.y = element_text(face="plain", color="black",  size=10, hjust=1))

thank you very much

Camilla

Answer 1

[rpolicastro]

I'm not completely sure what you are asking, but all you need to do for most cases is to make a column with factor levels corresponding to what you want colored differently. If you want a more detailed response post some example data for us to work with, and explain better what you want your output to be.

Here's an example using toy data.

df <- data.frame(gene=sprintf("ENS%05d", seq(100, 105, 1)), count=rnorm(6, 1, 0.25))

> df
      gene     count
1 ENS00100 0.9943232
2 ENS00101 0.8794445
3 ENS00102 0.8482693
4 ENS00103 1.2455513
5 ENS00104 1.1781073
6 ENS00105 0.6342954

Create your column with the factor levels.

df <- mutate(df, condition=if_else(count>1, "greater", "less"))

> df
      gene     count condition
1 ENS00100 0.9943232      less
2 ENS00101 0.8794445      less
3 ENS00102 0.8482693      less
4 ENS00103 1.2455513   greater
5 ENS00104 1.1781073   greater
6 ENS00105 0.6342954      less

Use the new factor levels to color the plot.

  ggplot(df, aes(x=gene, y=count, fill=condition)) +
    geom_col() +
    theme(axis.text.x=element_text(angle=45, hjust=1))