ggplot showing names of selected ids - r

Question

I want to plot a scatterplot using ggplot instead of plot and
produce this scatterplot where some IDs have different colors
and labeled:
Some asked for the dput of the data, so I added the dput in the end.
The problem with adding dput, the quesiton won't go through because
it shows too much coding compared to the question content, so that's
why I avoid it, and the head of dput is not helpful in reproducing the
real data.
I want to be able to plot all the values then show the name
for a selected IDs, not all of them.
This what I tried:
library(ggplot2)
library(ggrepel)
fig5ctrial<-read.csv(url("https://raw.githubusercontent.com/learnseq/learning/main/fig5mintab.txt"),sep = '\t',header = TRUE)
a_select<- c("RPL31", "HSPB1", "MAFB", "ALPL1", "VGF","PCSK1N", "BSG", "CALY", "B2M", "SCG5", "TM4SF4")
selalpha <- fig5ctrial[match(rev(a_select), fig5ctrial$geneIDs), ]
alphanames<-fig5ctrial$geneIDs
attach(fig5ctrial)
z1 <-plot(a.donor, a_cells, main="Scatterplot Example", xlab="Spearsman p all cells ", ylab="Spearsman p alpha cells ", pch=19)
text(a_cells~a.donor, labels=alphanames,data=fig5ctrial, font=2)
z01 <-
#Plot
ggplot()+
#assign alpha cell
geom_point(fig5ctrial,aes(a.donor, a_cells))+
#assign all cells
geom_point(fig5ctrial,aes(all_donors, all.cells))+
#assign IDs of interest
geom_point(fig5ctrial,aes(all_donors, all.cells, color = factor(selalpha)))+
#Add labels
geom_text_repel(data=subset(fig5ctrial,
geneIDs %in% a_select),
aes(label=geneIDs),show.legend = F)

Answer 1

The solution from #r2evans is tremendously complete (upvoted for the very well explained details) and better than this. I used the same ggrepel strategy, but with a simulated variable over your data:
library(ggplot2)
library(ggrepel)
#Data
fig5ctrial <- read.csv('https://raw.githubusercontent.com/learnseq/learning/main/alphacell.csv',stringsAsFactors = F)
#Group
fig5ctrial$allcellstypes <- sample(1:2,nrow(fig5ctrial),replace = T)
fig5ctrial$geneIDs <- trimws(fig5ctrial$geneIDs,whitespace = '\'')
#Plot
ggplot(fig5ctrial,aes(X...donor., a.cells, color = factor(allcellstypes)))+
geom_point()+
#Add labels
geom_text_repel(data=subset(fig5ctrial,
geneIDs %in% a_select),
aes(label=geneIDs),show.legend = F)+
labs(color='allcellstypes')
Output:
Update: After playing with data from OP, here a possible sketch to solve the issue (Remember in the last attempt we merged all data):
library(xlsx)
library(ggplot2)
library(ggrepel)
#Data
fig5cwithoutdesc <- read.xlsx('fig5cwithoutdesc.xlsx',1,colIndex = 1:4)
ids <- c("RPL31", "HSPB1", "MAFB", "ALPL1", "VGF","PCSK1N", "BSG", "CALY", "B2M", "SCG5", "TM4SF4")
#Clean gen id
fig5cwithoutdesc$geneIDs <- trimws(gsub('[[:punct:] ]+',' ',fig5cwithoutdesc$geneIDs))
#Plot
ggplot(fig5cwithoutdesc,aes(a_donor, a_cell,color=(geneIDs %in% ids)))+
geom_point()+
scale_color_manual(values = c('gray','blue'))+
geom_text_repel(data=subset(fig5cwithoutdesc,geneIDs %in% ids),
aes(label=geneIDs),force=19)+
theme_bw()+
theme(legend.position = 'none')
Output:

Answer 2

Here's a sample that uses a little dplyr (not essential) and ggrepel (essential).
Sample data:
set.seed(42)
dat <- data.frame(id = c(outer(letters, letters, paste0)), x = runif(26*26))
dat$y <- dat$x + rnorm(26*26, 0.2, 0.2)
dat[1:3,]
# id x y
# 1 aa 0.9148060 0.9611270
# 2 ba 0.9370754 1.0316538
# 3 ca 0.2861395 0.4818541
Code for the plot:
library(ggplot2)
library(ggrepel)
library(dplyr)
# dots we want to highlight
interesting <- c("mg", "qx", "zz")
dat %>%
mutate(id = replace(id, !id %in% interesting, "")) %>%
ggplot(., aes(x, y)) +
geom_point(aes(color = (id %in% interesting))) +
scale_color_manual(guide = FALSE, values = c("FALSE" = "black", "TRUE" = "red")) +
ggrepel::geom_text_repel(
aes(label = id), color = "red",
nudge_x = 0.5, direction = "x", hjust = 0)
Notes:
dplyr can easily be removed here, replaced with transform and perhaps a temp-variable;
the id=replace(...) portion is to remove the label (id) for any uninteresting variables, so that geom_text_repel will only label the interesting ones;
there are other techniques for highlighting specific dots, including adding another geom_point(..., data=~subset(., id %in% interesting)), but that adds more points ... and in some graphic formats (pdf, svg) this produces extra objects and therefore might have unintended consequences. Coloring the points in this way will be more difficult if you are already using aes(color=.) elsewhere.