Creating a grouped barplot in ggplot2 showing mean values (that I do not want to enter manually)

I'm trying to create a grouped bar plot showing the mean values for the data that I am inputting, and at the same time I want to make it easily reproducible so that I can swap in other data sets that are arranged similarly. I don't want to have to calculate the mean manually and input it in a data frame - is there a way to do this?

My data consist of numerical values between 100 - 2000 grouped into 3 different drug treatment groups, which are then subdivided into 3 groups (based on their anatomical location in an organism, termed "Inner", "Middle", "Outer"). The final plot should be 3 groups of 3 bars (each representing the mean values of cell survival in each of the 3 locations). So far I have managed to make individual barplots, but I want to combine them.

Here is some code that I have, and below that is a small excerpt from the data set:

 ggplot(Survival, aes(Treatment, Inner)) +
stat_summary(fun.y = mean,
             geom = "bar",
             position = "dodge")

Treatment   Inner   Middle  Outer
RAD          317    373      354
RAD          323    217      174
RAD          236    255      261
HUTS        1411    1844     1978
HUTS        1922    1756     1856
HUTS        1478    1711     1433
RGD         1433    1489     1633
RGD         1400    1500     1544
RGD         1222    1333     1444

Here's two options: one uses group_by and summarise from dplyr to calculate means, then pipes into ggplot to make dodged bars. The second uses stat_summary like your example code does. Both require gathering your data into a long format.

library(tidyverse)

df %>%
  gather(key = group, value = value, -Treatment) %>%
  group_by(Treatment, group) %>%
  summarise(mean_val = mean(value, na.rm = T)) %>%
  ggplot(aes(x = Treatment, y = mean_val, fill = group)) +
    geom_col(position = position_dodge())

df %>%
  gather(key = group, value = value, -Treatment) %>%
  ggplot(aes(x = Treatment, y = value, fill = group)) +
    stat_summary(fun.y = mean, geom = "col", position = position_dodge())