Gradient bar plot in ggplot2

Question

[suvratha]

I've 3 variables, for a set of genes - the log2FC, the log2(basemeanB) and the difference in expression of a gene in fraction of cells. below are my vectors -

df <- data.frame(Genes=c('ADCYAP1R’, ‘GRM1’, ‘GRM3’, ‘HTR1F’, ‘HTR2A’, ‘MCHR2’, ‘RXFP1’), Log2FC=c(-2.299, 0.52, 2.439, 1.783, 2.459, 5.196, 4.071))
df2 <- data.frame(Genes=c('ADCYAP1R’, ‘GRM1’, ‘GRM3’, ‘HTR1F’, ‘HTR2A’, ‘MCHR2’, ‘RXFP1’), grad=c(7.20, 7.79, 6.45, 7.43, 5.79, 0.79, 5.48))
df3 <- data.frame(Genes=c('ADCYAP1R’, ‘GRM1’, ‘GRM3’, ‘HTR1F’, ‘HTR2A’, ‘MCHR2’, ‘RXFP1’), beta=c(0.283, 0.35, 0.405, 0.451, 0.594, 0.523, 0.65))

I want a gradient bar plot with the height of the bar from the 1st vector, the gradient from the 2nd vector and the points on the bar from the 3rd vector.

I was following the below code, but i'm not able to get the output i want.

p <- ggplot(df, aes(x=Genes, y=Log2FC))

p <- p + geom_bar(stat = "identity", aes(fill = Log2FC)) + scale_fill_gradient2(low='plum1', mid='plum', high='plum4')
p + geom_point(data=df3, aes(x=as.numeric(Genes), y=total), colour = "blue", shape = 15, size = 3) + scale_y_continuous(sec.axis = sec_axis(~./3, name = "Beta Value"))

can anyone please help me?

[Ram]

Your code contains stylized single quotes mixed in with regular ASCII single quotes. This could result in unexpected errors. Please make sure that your code only has ASCII characters.

df <- data.frame(Genes=c('ADCYAP1R’
__________________________________^

should be

df <- data.frame(Genes=c('ADCYAP1R'

Answer 1

[venu]

Here is something for you to start with, you don't need 3 different dataframes.

df <- data.frame(Genes = c('ADCYAP1R', 'GRM1', 'GRM3', 'HTR1F', 'HTR2A', 'MCHR2', 'RXFP1'),
                 Log2FC = c(-2.299, 0.52, 2.439, 1.783, 2.459, 5.196, 4.071),
                 grad = c(7.20, 7.79, 6.45, 7.43, 5.79, 0.79, 5.48),
                 beta = c(0.283, 0.35, 0.405, 0.451, 0.594, 0.523, 0.65))

df <- df %>% 
  dplyr::arrange(Log2FC) %>% 
  tibble::rowid_to_column(var = "POS")

df$Genes <- factor(df$Genes, levels = df$Genes)

ggplot(df, aes(x = Genes, y = Log2FC, fill = grad))+
  geom_bar(stat = "identity")+
  geom_point(aes(x = POS, y = beta))+
  scale_fill_viridis_c()

This produces

You can play with secondary y-axis for geom_point.

P.S: I'd rather suggest to add text labels of beta on top of bars than adding points on the bars.

[suvratha]

Thank you very much!

[Ram]

If an answer was helpful, you should upvote it; if the answer resolved your question, you should mark it as accepted. You can accept more than one answer if they work.

[suvratha]

is there a function to use the gradient colours as different shades of grey with a white background?

[Ram]

There are functions, such as colorRamp and colorRampPalette that take a set of discrete colors and make a continuous scale out of them. See here:

You may want to use them in conjunction with gray.colors

[venu]

Official ggplot2 website has all this info and how to modify components of a plot. E.g. Sequential grey colour scales