Boxplot with ggplot2 in R-CodePudding

I have a dataframe called cells, with cells in rows and samples in columns, here is a dput sample:

structure(c(8.10937548981953e-20, 0.095381661829093, 0.054868371418562, 
0.0523687378840825, 0.0100173293159538, 0.0332395245437795, 3.37811149975583e-20, 
0.048191378909587, 0.13314908462763, 0, 0.00612878313809124, 
0, 0.00209460699328093, 0.205599458004829, 0.318048653115709, 
4.21796249339787e-05, 0.00844407692255898, 0, 0.00613007026042523, 
0.0300024082993193, 0.0405191646567986, 0.00654087887823056, 
0.0111094954094255, 1.30617589099212e-19, 0.0398730537850546, 
0.0390946117756341, 0.239413780024853, 2.07521807718399e-19, 
0.00116980239850497, 0, 0, 0.00971921247886335, 0.0588607291731613, 
3.8563512241696e-21, 0.00247621905821516, 0), .Dim = c(6L, 6L
), .Dimnames = list(c("Adipocytes", "B-cells", "Basophils", "CD4  memory T-cells", 
"CD4  naive T-cells", "CD4  T-cells"), c("Pt1", "Pt10", "Pt103", 
"Pt106", "Pt11", "Pt17")))

number of rows in cells is 38 and number of columns is 49. My target cell type that I want to build a boxplot for is: CD4 memory T-cells

In addition, have another data frame called Metadata, which contains the response column, if a sample responded to a certain therapy or not - Metadata$Benefit. So for each sample in cells I can know whether they responded or not.

Goal: make a boxplot with Response/No Response in the x axis and values on the y axis, and I need the boxplot with the indiviudal data points on it.

My code:

cells %>%
  ggplot( aes(x = as.factor(Metadata$Benefit), y = as.numeric(cells['CD4  memory T-cells',]), fill= c('red','lightblue')))  
  geom_boxplot()  
  scale_fill_viridis(discrete = TRUE, alpha=0.6)  
  geom_jitter(color="black", size=0.4, alpha=0.9)  
  theme_ipsum()  
  theme(
    legend.position="none",
    plot.title = element_text(size=11)
  )  
  ggtitle("A boxplot with jitter")  
  xlab("")

Error: Aesthetics must be either length 1 or the same as the data (38): x, y and fill

I don't get what's the problem because both length(as.factor(Metadata$Benefit)) and length(as.numeric(cells['CD4 memory T-cells',])) is 49, it's the same so what's the problem?

CodePudding user response：

There are a few issues with your attempt here as mentioned in the comments. You can't take x and y aesthetics from different data sources so you should merge them first. This also means rotating (for which I used t() while it was still a matrix) the original data prior to merging so each sample is a row. To demonstrate this I made up some response metadata.

library(tidyverse)

d <- structure(c(8.10937548981953e-20, 0.095381661829093, 0.054868371418562, 0.0523687378840825, 0.0100173293159538, 0.0332395245437795, 3.37811149975583e-20, 0.048191378909587, 0.13314908462763, 0, 0.00612878313809124, 0, 0.00209460699328093, 0.205599458004829, 0.318048653115709, 4.21796249339787e-05, 0.00844407692255898, 0, 0.00613007026042523, 0.0300024082993193, 0.0405191646567986, 0.00654087887823056, 0.0111094954094255, 1.30617589099212e-19, 0.0398730537850546, 0.0390946117756341, 0.239413780024853, 2.07521807718399e-19, 0.00116980239850497, 0, 0, 0.00971921247886335, 0.0588607291731613, 3.8563512241696e-21, 0.00247621905821516, 0), .Dim = c(6L, 6L), .Dimnames = list(c("Adipocytes", "B-cells", "Basophils", "CD4  memory T-cells", "CD4  naive T-cells", "CD4  T-cells"), c("Pt1", "Pt10", "Pt103", "Pt106", "Pt11", "Pt17")))

metadata <- data.frame(sample = c("Pt1", "Pt10", "Pt103", "Pt106", "Pt11", "Pt17"), response = c(T, T, T, F, F, F))

d %>% 
  t() %>% 
  as.data.frame() %>% 
  rownames_to_column("sample") %>% 
  right_join(metadata, .) %>% 
  ggplot(aes(x = response, y = `CD4  memory T-cells`))  
  geom_boxplot(aes(fill = response))  
  scale_fill_manual(values = c('red','lightblue'))
#> Joining, by = "sample"

^{Created on 2022-02-16 by the reprex package (v2.0.1)}