How to add duplicate label using directlabels package in R - r

Question

I have the following data frame:
library(tidyverse)
library(directlabels)
dat <- structure(list(time.course = c("CONTROL", "DAY03", "DAY06", "DAY09",
"DAY12", "DAY15", "CONTROL", "DAY03", "DAY06", "DAY09", "DAY12",
"DAY15", "CONTROL", "DAY03", "DAY06", "DAY09", "DAY12", "DAY15",
"CONTROL", "DAY03", "DAY06", "DAY09", "DAY12", "DAY15"), log_delta = c(0,
0.620163956872191, 0.97251217133899, 0.788819459139427, 0.412543422847407,
0.401621905837411, 0, -0.168711062429047, -0.973481367557294,
-1.46433243027353, -1.34771037206345, -1.77709667157235, 0, -0.187344700204557,
-0.254280909246003, -0.335330756378048, -0.655121382977672, -1.1733031812697,
0, -0.0160729795971869, -0.628563089917479, -1.43060414378064,
-1.466051599194, -2.57510172892555), `UMAP cluster` = structure(c(1L,
1L, 1L, 1L, 1L, 1L, 5L, 5L, 5L, 5L, 5L, 5L, 7L, 7L, 7L, 7L, 7L,
7L, 13L, 13L, 13L, 13L, 13L, 13L), .Label = c("1", "2", "3",
"4", "5", "6", "7", "8", "9", "10", "11", "12", "13", "14", "15"
), class = "factor"), cell_name = structure(c(1L, 1L, 1L, 1L,
1L, 1L, 5L, 5L, 5L, 5L, 5L, 5L, 7L, 7L, 7L, 7L, 7L, 7L, 7L, 7L,
7L, 7L, 7L, 7L), .Label = c("Macrophage", "Enteroendocrine",
"Endothelial", "Lymphatic", "Fibroblast", "T cell", "Myofibroblast",
"Absorbtice & secrectory cell", "Plasmacytoid DC", "Neutrophil",
"Plasma cell", "Cajal intestinal cell", "Glial cell", "Germinal center B cell"
), class = "factor")), row.names = c(NA, -24L), class = c("tbl_df",
"tbl", "data.frame"))
dat
It looks like this:
# A tibble: 24 x 4
time.course log_delta `UMAP cluster` cell_name
<chr> <dbl> <fct> <fct>
1 CONTROL 0 1 Macrophage
2 DAY03 0.620 1 Macrophage
3 DAY06 0.973 1 Macrophage
4 DAY09 0.789 1 Macrophage
5 DAY12 0.413 1 Macrophage
6 DAY15 0.402 1 Macrophage
7 CONTROL 0 5 Fibroblast
8 DAY03 -0.169 5 Fibroblast
9 DAY06 -0.973 5 Fibroblast
10 DAY09 -1.46 5 Fibroblast
11 DAY12 -1.35 5 Fibroblast
12 DAY15 -1.78 5 Fibroblast
13 CONTROL 0 7 Myofibroblast
14 DAY03 -0.187 7 Myofibroblast
15 DAY06 -0.254 7 Myofibroblast
16 DAY09 -0.335 7 Myofibroblast
17 DAY12 -0.655 7 Myofibroblast
18 DAY15 -1.17 7 Myofibroblast
19 CONTROL 0 13 Myofibroblast
20 DAY03 -0.0161 13 Myofibroblast
21 DAY06 -0.629 13 Myofibroblast
22 DAY09 -1.43 13 Myofibroblast
23 DAY12 -1.47 13 Myofibroblast
24 DAY15 -2.58 13 Myofibroblast
Notice that "Myofibroblast" occurs twice as UMAP cluster 7 and 13.
I tried to plot that using this code with directlabels package:
ggplot(dat, aes(x = time.course, y = log_delta,
color = cell_name)) +
geom_line(aes(group = `UMAP cluster`)) +
scale_x_discrete(expand = c(0, 2.5)) +
theme_bw() +
theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank()) +
theme(axis.text.x = element_text(angle = 60, hjust = 1)) +
xlab("") +
ylab("log(proportion / control proportion)") +
theme(legend.title = element_blank()) +
geom_dl(aes(label = cell_name), method = list(dl.trans(x = x + 0.3), "last.bumpup", cex = 0.8))
The plot looks like this:
Notice that Myofibroblast doesn't occur at the end of the two lines (blue).
What I want to do is
to color Myofibroblast with two colors
each of Myofibroblast lines also tagged with labels.
How can I achieve that?

Answer 1

It's not possible to do it with geom_dl because it inherits the aes and directly takes the column variable to be the label. I can think of two solutions, first, is to create a new variable by fusing the cluster id with cell type:
# the column name is giving some problems
colnames(dat)[3] = "UMAPcluster"
dat <- dat %>% mutate(new=paste(cell_name,UMAPcluster))
dat$new <- factor(dat$new,levels=unique(dat$new))
ggplot(dat, aes(x = time.course, y = log_delta,,group=new,col = new)) +
geom_line() +
scale_x_discrete(expand = c(0, 2.5)) +
theme_bw() +
theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank()) +
theme(axis.text.x = element_text(angle = 60, hjust = 1)) +
xlab("") +
ylab("log(proportion / control proportion)") +
theme(legend.title = element_blank()) +
geom_dl(aes(label = new), method = list(dl.trans(x = x + 0.3), "last.bumpup", cex = 0.8))
Or you create a new data frame and annotate with geom_text (or geom_label if you like boxes). Preferably you keep the legend for the cluster so that it is clear what the colors mean.
LAB = dat %>% group_by(UMAPcluster) %>% top_n(1,wt=time.course)
ggplot(dat, aes(x = time.course, y = log_delta,color = UMAPcluster)) +
geom_line(aes(group = UMAPcluster)) +
scale_x_discrete(expand = c(0, 2.5)) +
theme_bw() +
theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank()) +
theme(axis.text.x = element_text(angle = 60, hjust = 1)) +
xlab("") +
ylab("log(proportion / control proportion)") +
geom_label(data=LAB,aes(label=cell_name),show.legend=FALSE,nudge_x=0.7)