0
$\begingroup$

I want to create a stacked bar-graph with different cell cycles for each cell type within each condition. I have uploaded the file for it. Don't know how to go about it as now I have another condition (cell cycle). I was able to do the bar graph comparing cell cycle per condition but now I want to compare cell type and cell cycle in each condition

output of dput(head(all.combined@metadata))

structure(list(orig.ident = c("treated", "treated", "treated", 
    "treated", "treated", "treated"), nCount_RNA = c(1892, 307, 1348, 
    3699, 4205, 4468), nFeature_RNA = c(960L, 243L, 765L, 1612L, 
    1341L, 1644L), percent.mt = c(0.211416490486258, 1.62866449511401, 
    4.45103857566766, 4.4065963773993, 0.0713436385255648, 3.87197851387645
    ), RNA_snn_res.0.5 = structure(c(11L, 11L, 5L, 6L, 11L, 13L), .Label = c("0", "1", "2", "3", "4", "5", "6", "7", "8", "9", "10", "11", "12", 
    "13", "14", "15", "16", "17", "18", "19"), class = "factor"), seurat_clusters = structure(c(11L, 11L, 5L, 6L, 11L, 13L), .Label = c("0", "1", "2", "3", "4", "5", "6", "7", "8", "9", "10", "11", "12", "13", "14", "15", "16", "17", "18", "19"), class = "factor"), S.Score = c(0.476893835992198, -0.0200784617568548, -0.0335915198305002, -0.0247184276246385, 0.010785196602457, 0.0190008903712199), G2M.Score = c(0.204441469200986, 0.173804859670862, -0.0313235510969097, -0.0376796363661889, -0.0559526905696905, -0.0122031631356698), Phase = structure(c(3L, 2L, 1L, 1L, 3L, 3L), .Label = c("G1", "G2M", "S"), class = "factor"), old.ident = structure(c(7L,7L, 1L, 4L, 7L, 9L), .Label = c("Fibroblast", "T cell", "Macrophage", "Stellate", "Acinar", "Endothelial", "Tumor", "B cell", "Mast cell", "Ductal", "Islets of Langerhans"), class = "factor")), row.names = c("treated_AAACGCTAGCGGGTTA-1", "treated_AAAGGTAAGTACAGAT-1", "treated_AAAGTGAGTTTGATCG-1", "treated_AAATGGACAAAGTGTA-1", 
    "treated_AACAAAGGTCGACTTA-1", "treated_AACAGGGTCCTAGCCT-1"), class = "data.frame")

output of dput(tail(all.combined@metadata))

structure(list(orig.ident = c("untreated", "untreated", "untreated", 
"untreated", "untreated", "untreated"), nCount_RNA = c(901, 823, 
1184, 1835, 1147, 1407), nFeature_RNA = c(482L, 479L, 649L, 1043L, 
604L, 709L), percent.mt = c(1.77580466148724, 2.91616038882138, 
4.22297297297297, 3.86920980926431, 2.0052310374891, 4.05117270788913
), RNA_snn_res.0.5 = structure(c(7L, 7L, 7L, 14L, 7L, 7L), .Label = c("0", 
"1", "2", "3", "4", "5", "6", "7", "8", "9", "10", "11", "12", 
"13", "14", "15", "16", "17", "18", "19"), class = "factor"), 
    seurat_clusters = structure(c(7L, 7L, 7L, 14L, 7L, 7L), .Label = c("0", 
    "1", "2", "3", "4", "5", "6", "7", "8", "9", "10", "11", 
    "12", "13", "14", "15", "16", "17", "18", "19"), class = "factor"), 
    S.Score = c(-0.0320858200243315, 0.0304725660342869, 0.0215996091745327, 
    0.0384166213301423, 0.144956251122548, -0.0242770509986111
    ), G2M.Score = c(0.0904224391544142, 0.050148242050667, -0.0178041670730754, 
    -0.0112596867977946, -0.0519554524339088, -0.0136533184257381
    ), Phase = structure(c(2L, 2L, 3L, 3L, 3L, 1L), .Label = c("G1", 
    "G2M", "S"), class = "factor"), old.ident = structure(c(5L, 
    5L, 5L, 5L, 5L, 5L), .Label = c("Fibroblast", "T cell", "Macrophage", 
    "Stellate", "Acinar", "Endothelial", "Tumor", "B cell", "Mast cell", 
    "Ductal", "Islets of Langerhans"), class = "factor")), row.names = c("untreated_TTTGGTTGTCTAATCG-18", 
"untreated_TTTGGTTTCCCGAGGT-18", "untreated_TTTGTTGAGAACTGAT-18", 
"untreated_TTTGTTGAGCTCGGCT-18", "untreated_TTTGTTGAGTGCCTCG-18", 
"untreated_TTTGTTGCACGGTGCT-18"), class = "data.frame")

Been using this code to generate the previous graph.

ggplot(CC, aes(x = Condition, y = Percent, fill = Cell_Cycle))+
  geom_bar(stat = "identity")+
  geom_text(aes(label = paste(round(Percent,2),"%")), position = position_stack(vjust =  0.5))

cell cycle by condition

$\endgroup$
4
  • $\begingroup$ What is DF ? Can you provide an image of the graph when you are comparing cell cycle in each condition ? Your example "all.combined@metadata" contains only "untreated" conditions, maybe you should update your example to provide few lines of treated conditions too. $\endgroup$
    – dc37
    Jan 20 '20 at 23:25
  • $\begingroup$ geom_bar(stat = "identity") is rarely correct, as bar plots should be used for count-level data only. If you want to use a value, then use geom_col() instead - this is explained in the help for geom_bar. $\endgroup$
    – gringer
    Jan 20 '20 at 23:35
  • $\begingroup$ You don't have to post all the metadata file, but few lines of each condition "treated / untreated" could be helpful. $\endgroup$
    – dc37
    Jan 20 '20 at 23:52
  • $\begingroup$ @dc37 I posted the head and tail of the dataset. $\endgroup$
    – user6774
    Jan 20 '20 at 23:52
1
$\begingroup$

Based on your example, you can use count for each phase of cell cycle for each cell types in each condition. Here, I'm using dplyr package to do that but most likely, you could have the same output using various methods:

DF <- rbind(dfhead, dftail)

library(dplyr)
DF_Count <- DF %>%group_by(orig.ident,Phase,old.ident) %>%
  count() %>%
  ungroup() %>%
  group_by(orig.ident,old.ident) %>%
  mutate(Freq = n/sum(n)*100)

# A tibble: 8 x 5
# Groups:   orig.ident, old.ident [5]
  orig.ident Phase old.ident      n  Freq
  <chr>      <fct> <fct>      <int> <dbl>
1 treated    G1    Fibroblast     1 100  
2 treated    G1    Stellate       1 100  
3 treated    G2M   Tumor          1  33.3
4 treated    S     Tumor          2  66.7
5 treated    S     Mast cell      1 100  
6 untreated  G1    Acinar         1  16.7
7 untreated  G2M   Acinar         2  33.3
8 untreated  S     Acinar         3  50  

As you can see, DF_Count has the frequency for each cell types of each phase of the cell cycle in function of the condition. We can use DF_Count to get the following plot. Using facet_wrap, you can create two panels based on the condition column and thus represent the cell cycle of each cell types in function of the treatment condition:

ggplot(DF_Count, aes(x = old.ident, y = Freq, fill = Phase))+
  geom_col()+
  geom_text(aes(label = paste(round(Freq, 2),"%")),position = position_stack(vjust = 0.5))+
  facet_wrap(~orig.ident)

enter image description here

Does it look what you are expecting ?

NB: DF here is the addition of your head and tail of your all.combined dataframe.

$\endgroup$
2
  • $\begingroup$ How about the other cell types? Can I just make DF <- all.combined@metadata? That way all the cell types are included? $\endgroup$
    – user6774
    Jan 21 '20 at 0:22
  • $\begingroup$ It should work. Here the graph represent only the portion of data you have provided. If you think this answer was useful, do not hesitate to vote or validate it. Learn more here: bioinformatics.stackexchange.com/help/someone-answers $\endgroup$
    – dc37
    Jan 21 '20 at 0:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.