How can a line to indicate significance be overlaid on a bar plot with multiple groups using ggplot2?

Question

I am trying to add lines to a bar plot to indicate significance between two observations. For my plot my bar plot , I want to add a line above first two x-axis observations indicating that there is a significant difference ie between BDD angry and Control angry, much like what has been done in other threads, but not ones with multiple groups, eg: Example bar plot

Similar to what has been done here: Indicating the statistically significant difference in bar graph USING R

MY PLOT CODE:

p <- ggplot(faces_data_accuracy, aes(x=Condition, y=Mean, fill=Group)) + 
  geom_bar(position=position_dodge(), stat="identity") + 
  geom_errorbar(aes(ymin=Mean-se, ymax=Mean+se), #ADD ERROR BARS
                width=.2,                    # Width of the error bars
                position=position_dodge(.9)) +
  ylab("Percentage of Correct Responses")+
  xlab("Emotion")+
  theme_bw()+
  theme(
    plot.background = element_blank()
    ,panel.grid.major = element_blank()
    ,panel.grid.minor = element_blank()
    ,panel.border = element_blank() 
  ) +
  theme(axis.line = element_line(color = 'grey')) +
  scale_fill_brewer(palette="Paired") 

So I have gotten as far as to create a data frame with the coordinates of the p-value and plot that as text:

  label.df <- data.frame(Condition = c("Angry", "Angry"), Mean = c(86, 87), Group = c("BDD","Control"))
  arc.df <- data.frame(Condition = x, Mean = y)
  p+geom_text(data = label.df, label = "p=0.028")+
    geom_line(data = arc.df, aes(Condition+1, Mean+10))

But no matter what I do I cannot seem to add a line. Can you please help me add a line at position 80 on the y-axis that connects the two angry observations like the example plot?

Answer 1

I think the package ggsignif is exactly what you are after ( https://cran.r-project.org/web/packages/ggsignif/index.html ). You haven't provided your data to work with but I am guessing that +geom_signif(comparisons=list(c('BDD','Control'))) should be close to what you need.

Consider the example provided in the vignette:

library(ggplot2)
library(ggsignif)
ggplot(mpg, aes(class, hwy)) +
  geom_boxplot() +
  geom_signif(comparisons = list(c("compact", "pickup"),
                                 c("subcompact", "suv")))