09: Data visualization with ggplot2

Juan Caballero

What is ggplot2?

ggplot2 is part of the tidyverse, a set of packages created by Hadley Wickham.
ggplot2 implements a grammar of graphics to enable creation of plots from modular building blocks
ggplot2 is designed to work with tidy data (we’ll get into this later today).

Graph components

Plots in ggplot2 consist of 3 main components:
- Data: The dataset being summarized
- Aesthetic mapping: Variables mapped to visual cues, such as x-axis and y-axis values and colors
- Geometry: The type of plot (scatterplot, boxplot, barplot, histogram, qqplot, smooth density, etc.)

ggplot(data = <DATA>, mapping = aes(<MAPPINGS>)) +  <GEOM_FUNCTION>()

Graph components

geoms

geom_point() for scatter plots, dot plots, etc.
geom_histogram() for histograms
geom_boxplot() for, well, boxplots!
geom_line() for trend lines, time series, etc.
and many more …

Graph components

There are additional components:
- Scale
- Labels, Title, Legend
- Theme/Style

Graph components

Our dataset

Expression matrix from the mouse influenza experiment

library(tidyverse)

rnaseq_file = "https://raw.githubusercontent.com/maxplanck-ie/Rintro/2024.04/qmd/data/rnaseq_counts_wide.csv"

rna = read_csv(rnaseq_file)

rna

# A tibble: 1,474 × 23
   gene    GSM2545336 GSM2545337 GSM2545338 GSM2545339 GSM2545340 GSM2545341
   <chr>        <dbl>      <dbl>      <dbl>      <dbl>      <dbl>      <dbl>
 1 Asl           1170        361        400        586        626        988
 2 Apod         36194      10347       9173      10620      13021      29594
 3 Cyp2d22       4060       1616       1603       1901       2171       3349
 4 Klk6           287        629        641        578        448        195
 5 Fcrls           85        233        244        237        180         38
 6 Slc2a4         782        231        248        265        313        786
 7 Exd2          1619       2288       2235       2513       2366       1359
 8 Gjc2           288        595        568        551        310        146
 9 Plp1         43217     101241      96534      58354      53126      27173
10 Gnb4          1071       1791       1867       1430       1355        798
# ℹ 1,464 more rows
# ℹ 16 more variables: GSM2545342 <dbl>, GSM2545343 <dbl>, GSM2545344 <dbl>,
#   GSM2545345 <dbl>, GSM2545346 <dbl>, GSM2545347 <dbl>, GSM2545348 <dbl>,
#   GSM2545349 <dbl>, GSM2545350 <dbl>, GSM2545351 <dbl>, GSM2545352 <dbl>,
#   GSM2545353 <dbl>, GSM2545354 <dbl>, GSM2545362 <dbl>, GSM2545363 <dbl>,
#   GSM2545380 <dbl>

Our dataset

Let’s plot a MA-like plot for GSM2545336 vs GSM2545380, first, we generate the tibble as:

# log2fc:     M = log2(x/y) = log2(x) - log2(y)
# norm_mean:  A = 1/2 ( log2(x) + log2(y) )
ma_data = rna %>% 
          select(gene, GSM2545336, GSM2545380) %>% 
          mutate(
              norm_mean = ( log2(GSM2545336 + 1) + log2(GSM2545380 + 1) ) / 2,
              log2fc = log2(GSM2545336 + 1) - log2(GSM2545380 + 1)
            ) %>% 
            select(gene, log2fc, norm_mean)
ma_data

# A tibble: 1,474 × 3
   gene      log2fc norm_mean
   <chr>      <dbl>     <dbl>
 1 Asl     -0.0269      10.2 
 2 Apod    -0.0759      15.2 
 3 Cyp2d22  0.0146      12.0 
 4 Klk6     0.148        8.10
 5 Fcrls    0.318        6.27
 6 Slc2a4  -0.0256       9.63
 7 Exd2    -0.0653      10.7 
 8 Gjc2     0.136        8.11
 9 Plp1     0.386       15.2 
10 Gnb4     0.00810     10.1 
# ℹ 1,464 more rows

ggplot2 in action - `geom_point()`

geom_point() are basic scatter plots

ggplot(ma_data, aes(x = norm_mean, y = log2fc)) +
  geom_point()

ggplot2 in action - `geom_point()`

We can modify graphic elements, for color

ggplot(ma_data, aes(x = norm_mean, y = log2fc)) +
  geom_point(color = 'blue')

ggplot2 in action - `geom_point()`

Overplotting can be fixed using transparency with alpha

ggplot(ma_data, aes(x = norm_mean, y = log2fc)) +
  geom_point(alpha = 0.1)

ggplot2 in action - `geom_point()`

Point marks can be changed with shape, the numbers are predefined forms

ggplot(ma_data, aes(x = norm_mean, y = log2fc)) +
  geom_point(shape = 3)

ggplot2 in action - `geom_point()`

shape also accept any char

ggplot(ma_data, aes(x = norm_mean, y = log2fc)) +
  geom_point(shape = '?')

ggplot2 in action - `geom_point()`

We can overlap new elements, such as an horizontal line geom_hline()

ggplot(ma_data, aes(x = norm_mean, y = log2fc)) +
  geom_point(color = 'blue', alpha = 0.1) +
  geom_hline(yintercept =  1, color = 'red') +
  geom_hline(yintercept = -1, color = 'red')

Hands on

Generate a similar plot comparing another 2 samples, use red “triangles” as points with a vertical green line in norm_mean = 5

Hands on

ggplot(ma_data, aes(x = norm_mean, y = log2fc)) +
  geom_point(color = 'red', shape = 17, alpha = 0.1) +
  geom_vline(xintercept = 5, color = 'green')

ggplot2 in action - `geom_histogram()`

Histogram of expression values for sample GSM2545336

ggplot(rna, aes(x = GSM2545336)) +
  geom_histogram()

ggplot2 in action - `geom_histogram()`

We can adjust the histogram bins

ggplot(rna, aes(x = GSM2545336)) +
  geom_histogram(bins = 50)

ggplot2 in action - `geom_histogram()`

We can use a variable to store a template

# Assign plot to a variable
hist_plot = ggplot(rna, aes(x = GSM2545336))
# Draw the plot
hist_plot + geom_histogram(bins = 100, color = 'blue')

Hands on

Create a new histogram for another sample values but in log2 scale

Hands on

hist_plot = ggplot(rna, aes(x = log2(GSM2545336 + 1)))
hist_plot + geom_histogram(bins = 100, color = 'blue')

New dataset

rnaseq.csv is the long-format for the mouse influenza experiment. Let’s add a log2-expression column

rnaseq_file = "https://raw.githubusercontent.com/maxplanck-ie/Rintro/2024.04/qmd/data/rnaseq.csv"
rna_data = read_csv(rnaseq_file)
rna_long = rna_data %>% 
           mutate(expression_log = log2(expression + 1))
rna_long

# A tibble: 32,428 × 20
   gene    sample  expression organism   age sex   infection strain  time tissue
   <chr>   <chr>        <dbl> <chr>    <dbl> <chr> <chr>     <chr>  <dbl> <chr> 
 1 Asl     GSM254…       1170 Mus mus…     8 Fema… Influenz… C57BL…     8 Cereb…
 2 Apod    GSM254…      36194 Mus mus…     8 Fema… Influenz… C57BL…     8 Cereb…
 3 Cyp2d22 GSM254…       4060 Mus mus…     8 Fema… Influenz… C57BL…     8 Cereb…
 4 Klk6    GSM254…        287 Mus mus…     8 Fema… Influenz… C57BL…     8 Cereb…
 5 Fcrls   GSM254…         85 Mus mus…     8 Fema… Influenz… C57BL…     8 Cereb…
 6 Slc2a4  GSM254…        782 Mus mus…     8 Fema… Influenz… C57BL…     8 Cereb…
 7 Exd2    GSM254…       1619 Mus mus…     8 Fema… Influenz… C57BL…     8 Cereb…
 8 Gjc2    GSM254…        288 Mus mus…     8 Fema… Influenz… C57BL…     8 Cereb…
 9 Plp1    GSM254…      43217 Mus mus…     8 Fema… Influenz… C57BL…     8 Cereb…
10 Gnb4    GSM254…       1071 Mus mus…     8 Fema… Influenz… C57BL…     8 Cereb…
# ℹ 32,418 more rows
# ℹ 10 more variables: mouse <dbl>, ENTREZID <dbl>, product <chr>,
#   ensembl_gene_id <chr>, external_synonym <chr>, chromosome_name <chr>,
#   gene_biotype <chr>, phenotype_description <chr>,
#   hsapiens_homolog_associated_gene_name <chr>, expression_log <dbl>

ggplot2 in action - `geom_boxplot()`

Boxplots are good to visualize distributions, let’s see gene log2-expression over samples

ggplot(rna_long, aes(y = expression_log, x = sample)) +
  geom_boxplot()

ggplot2 in action - `geom_boxplot()`

We can add the points over the boxplot with geom_jitter()

ggplot(rna_long, aes(y = expression_log, x = sample)) +
  geom_jitter(alpha = 0.2, color = "tomato") +
  geom_boxplot(alpha = 0)

ggplot2 in action - `geom_boxplot()`

I can’t read the labels!! theme can modify the axis labels

ggplot(rna_long, aes(y = expression_log, x = sample)) +
  geom_jitter(alpha = 0.2, color = "tomato") +
  geom_boxplot(alpha = 0) +
  theme(axis.text.x = element_text(angle = 90,  hjust = 0.5, vjust = 0.5))

ggplot2 in action - `geom_boxplot()`

Can we color each sample per time point? Sure …

# time as integer
ggplot(rna_long, aes(y = expression_log, x = sample)) +
  geom_jitter(alpha = 0.2, aes(color = time)) +
  geom_boxplot(alpha = 0) +
  theme(axis.text.x = element_text(angle = 90,  hjust = 0.5, vjust = 0.5))

Is this what we want?

ggplot2 in action - `geom_boxplot()`

Can we color each sample per time point? Sure …

# time as factor
ggplot(rna_long, aes(y = expression_log, x = sample)) +
  geom_jitter(alpha = 0.2, aes(color = as.factor(time))) +
  geom_boxplot(alpha = 0) +
  theme(axis.text.x = element_text(angle = 90,  hjust = 0.5, vjust = 0.5))

Better …

ggplot2 in action - `facet_wrap()`

We can plot multiple elements in a single plot separately with facet_wrap()

ggplot(rna_long,aes(x = expression_log)) +
  geom_histogram(bins = 50) +
  facet_wrap(~ sample)

ggplot2 in action - `facet_wrap()`

By default, all scales are the same, to avoid this:

ggplot(rna_long, aes(x = expression_log)) +
  geom_histogram(bins = 50) +
  facet_wrap(~ sample, scales = "free_y")

ggplot2 in action - `theme_bw()`

ggplot2 supports themes such as theme_bw()

ggplot(rna_long, aes(x = expression_log)) + 
  geom_histogram(bins = 50) +
  facet_wrap(~ sample, scales = "free_y") +
  theme_bw() +
  theme(panel.grid = element_blank())

ggplot2 in action - `ggsave()`

At the end, we would like to save our gorgeous plot, we can use ggsave().

my_plot = ggplot(rna_long, aes(x = expression_log)) +
          geom_histogram(bins = 50) +
          facet_wrap(~ sample, scales = "free_y") +
          theme_bw() +
          theme(panel.grid = element_blank())

ggsave("samples_expr_histograms.png", 
        my_plot,
        width = 15,
        height = 10)

Tip of the iceberg

See this handy cheatsheet for lots more!

09: Data visualization with ggplot2

What is ggplot2?

Graph components

Graph components

Graph components

Graph components

Our dataset

Our dataset

ggplot2 in action - geom_point()

ggplot2 in action - geom_point()

ggplot2 in action - geom_point()

ggplot2 in action - geom_point()

ggplot2 in action - geom_point()

ggplot2 in action - geom_point()

Hands on

Hands on

ggplot2 in action - geom_histogram()

ggplot2 in action - geom_histogram()

ggplot2 in action - geom_histogram()

Hands on

Hands on

New dataset

ggplot2 in action - geom_boxplot()

ggplot2 in action - geom_boxplot()

ggplot2 in action - geom_boxplot()

ggplot2 in action - geom_boxplot()

ggplot2 in action - geom_boxplot()

ggplot2 in action - facet_wrap()

ggplot2 in action - facet_wrap()

ggplot2 in action - theme_bw()

ggplot2 in action - ggsave()

Tip of the iceberg

Any questions?

ggplot2 in action - `geom_point()`

ggplot2 in action - `geom_point()`

ggplot2 in action - `geom_point()`

ggplot2 in action - `geom_point()`

ggplot2 in action - `geom_point()`

ggplot2 in action - `geom_point()`

ggplot2 in action - `geom_histogram()`

ggplot2 in action - `geom_histogram()`

ggplot2 in action - `geom_histogram()`

ggplot2 in action - `geom_boxplot()`

ggplot2 in action - `geom_boxplot()`

ggplot2 in action - `geom_boxplot()`

ggplot2 in action - `geom_boxplot()`

ggplot2 in action - `geom_boxplot()`

ggplot2 in action - `facet_wrap()`

ggplot2 in action - `facet_wrap()`

ggplot2 in action - `theme_bw()`

ggplot2 in action - `ggsave()`