How to visualize multiple bar plots in one (or splitted) pdf - r

Question

I'm using the tidyverse-ggplot2 combination to plot multiple bar plots. In one of my comparisons i would like to have even up to 300 single plots. I was wondering if there is a possibility to make sure that the plots will be visible in the pdf file and not look like the attached example
If possible I would prefer to have all the plots in one single pdf file, but if not, also multiple pages will be ok.
The command to plot the bar charts is
common %>%
as_tibble(rownames="gene") %>%
left_join(x= ., y = up[,1:2], by = c("gene" = "ensembl_gene_id") ) %>%
pivot_longer(starts_with("S"), names_to="sample", values_to="counts") %>%
left_join(groups, by="sample") %>%
group_by(mgi_symbol, group, cond, time) %>%
summarize(mean_count=mean(counts)) %>%
ggplot( aes(x = time, y = mean_count, fill=cond)) +
geom_bar(stat = "identity", position = position_dodge(width=0.9) ) +
scale_fill_manual(values=c("darkblue", "lightblue", "black")) +
facet_wrap(~mgi_symbol, scales = "free", ncol = 5) +
theme_bw()
I forgot to add the group table
groups <- tibble(
sample= colnames(normCounts),
group = rep(seq(1, ncol(normCounts)/3), each=3),
cond = rep(c("WT", "GCN2-KO", "GCN1-KO"), each = 12),
time = rep(rep(c("0h", "1h", "4h", "8h"), each=3), times = 3 )
)
thanks
Adding the command with the group_map was as such
common %>%
as_tibble(rownames="gene") %>%
left_join(x= ., y = up[,1:2], by = c("gene" = "ensembl_gene_id") ) %>%
pivot_longer(starts_with("S"), names_to="sample", values_to="counts") %>%
left_join(groups, by="sample") %>%
group_by(mgi_symbol, group, cond, time) %>%
summarize(mean_count=mean(counts)) %>%
group_map(function(g, ...)
ggplot(g, aes(x = time, y = mean_count, fill=cond)) +
geom_bar(stat = "identity", position = position_dodge(width=0.9) ) +
scale_fill_manual(values=c("darkblue", "lightblue", "black")) +
facet_wrap(~mgi_symbol, scales = "free", ncol = 5) +
theme_bw()
)
EDIT
This is how the data looks like in the input table (after summarizing the means)
df <-
common %>%
as_tibble(rownames="gene") %>%
left_join(x= ., y = up[,1:2], by = c("gene" = "ensembl_gene_id") ) %>%
pivot_longer(starts_with("S"), names_to="sample", values_to="counts") %>%
left_join(groups, by="sample") %>%
group_by(mgi_symbol, group, cond, time) %>%
summarize(mean_count=mean(counts)) %>%
ungroup()
df
#>`summarise()` regrouping output by 'mgi_symbol', 'group', 'cond' (override with `.groups` argument)
#> # A tibble: 1,212 x 5
#> mgi_symbol group cond time mean_count
#> <chr> <int> <chr> <chr> <dbl>
#> 1 0610031O16Rik 1 WT 0h 14.4
#> 2 0610031O16Rik 2 WT 1h 30.9
#> 3 0610031O16Rik 3 WT 4h 45.5
#> 4 0610031O16Rik 4 WT 8h 56.0
#> 5 0610031O16Rik 5 GCN2-KO 0h 18.9
#> 6 0610031O16Rik 6 GCN2-KO 1h 39.4
#> 7 0610031O16Rik 7 GCN2-KO 4h 13.9
#> 8 0610031O16Rik 8 GCN2-KO 8h 13.3
#> 9 0610031O16Rik 9 GCN1-KO 0h 12.3
#> 10 0610031O16Rik 10 GCN1-KO 1h 25.3
#> # … with 1,202 more rows

Answer 1

Start with some dummy data. This is the data after you've finished running left_join, pivot_longer, group_by, summarize.
library(tidyverse)
df <- tibble(
time = 1:5,
mean_count = 1:5,
cond = "x"
) %>%
expand_grid(mgi_symbol = c(letters, LETTERS))
Create a column group which represents what page the mgi_symbol belongs on.
plots_per_page <- 20
df <-
df %>%
mutate(group = (dense_rank(mgi_symbol) - 1) %/% plots_per_page)
Create all the plots with group_map.
plots <-
df %>%
group_by(group) %>%
group_map(function(g, ...) {
ggplot(g, aes(x = time, y = mean_count, fill=cond)) +
geom_bar(stat = "identity", position = position_dodge(width=0.9) ) +
scale_fill_manual(values=c("darkblue", "lightblue", "black")) +
facet_wrap(~mgi_symbol, scales = "free", ncol = 5) +
theme_bw()
})
Save as multiple pages using ggpubr
ggpubr::ggexport(
ggpubr::ggarrange(plotlist = plots, nrow = 1, ncol = 1),
filename = "plots.pdf"
)