Interactive Data Visualization with R Shiny for Life Sciences

Day 1: ggplot2 recap

Michael Teske & Jonas Schmid, Zurich 08 April 2024

grammar of graphics

ggplot2_logo

 


Data
Sepal.Length Sepal.Width Petal.Length Petal.Width Species
5.1 3.5 1.4 0.2 setosa
4.9 3 1.4 0.2 setosa
4.7 3.2 1.3 0.2 setosa
4.6 3.1 1.5 0.2 setosa
5 3.6 1.4 0.2 setosa
...
Aesthetics
  • X axis
  • Y axis
  • Point colour
  • ...
Geometry
  • Barplot
  • Boxplot
  • Scatterplot
  • Lineplot
  • ...

grammar of graphics

ggplot2_logo

 


Data
Sepal.Length Sepal.Width Petal.Length Petal.Width Species
5.1 3.5 1.4 0.2 setosa
4.9 3 1.4 0.2 setosa
4.7 3.2 1.3 0.2 setosa
4.6 3.1 1.5 0.2 setosa
5 3.6 1.4 0.2 setosa
...

 

ggplot(data, aes(x=Petal.Length y=Petal.Width)) + 
	geom_point(aes(color=Species))

grammar of graphics

  • Package: library(ggplot2) OR library(tidyverse)
  • Data: ggplot2 works with data frames, allowing you to map variables in your data to visual elements in your plot.
  • Aesthetic Mapping: Aesthetic mappings define how variables in your dataset are mapped to visual properties of the plot, such as x and y axes, color, shape, size, etc.
  • Geometric Objects (geoms): Geoms are the actual graphical representations of data points in your plot, such as points, lines, bars, and polygons.
  • Layers: In ggplot2, plots are built up in layers, where each layer adds additional information to the plot. Layers typically consist of geoms (geometric objects) and stats (statistical transformations).
  • Themes: Themes control the overall appearance of the plot, including axes, background, text, etc. ggplot2 provides several built-in themes, and you can customize themes to match your preferences or the requirements of your visualization.

Further Reading

Background

Resources

mystical_book

Let's create a plot!

Iris_ViolinPlot
Iris Dataset :: Violin Plot 
								
									data(iris)

									ggplot(iris, aes(x=Species, y=Petal.Length)) + 
										geom_violin(aes(fill=Species)) + 
										geom_point() + 
										geom_boxplot(width=0.1)

Let's create a plot!

Iris_ViolinPlot_quasirandom
Iris Dataset :: Violin Plot 
								
									library(ggbeeswarm)

									data(iris)

									ggplot(iris, aes(x=Species, y=Petal.Length)) + 
										geom_violin(aes(fill=Species)) + 
										geom_quasirandom() + 
										geom_boxplot(width=0.1)

Let's create a plot! - Variant 2

Ruhland2016_VolcanoPlot
Ruhland2016 Dataset :: Volcano Plot 
							
								# Load data
								ruhland2016 <- read.csv('Ruhland2016.csv', row.names=1)
								
								# Mark differentially expressed genes
								log2FCthreshold <- log2(1.5)
								p_threshold     <- 0.05
								
								ruhland2016$Expression <- 'not sign.'
								ruhland2016$Expression[ruhland2016$log2FoldChange >= log2FCthreshold & ruhland2016$pval <= p_threshold]  <- 'Up'
								ruhland2016$Expression[ruhland2016$log2FoldChange <= -log2FCthreshold & ruhland2016$pval <= p_threshold] <- 'Down'
								
								# Plot
								ggplot(ruhland2016, aes(x=log2FoldChange, y=-log10(pvalue), col=Expression)) +
										geom_point() +
										geom_vline(
											xintercept  = c(-log2FCthreshold, log2FCthreshold),
											linetype    ='dotted'
										) +
										geom_hline(
											yintercept  = -log10(p_threshold),
											linetype    = 'dotted'
										) +
										ggtitle('senescent vs wt') +
										# Optional
										scale_colour_manual(values=c(
											'not sign.'  = '#777777',
											'Up'         = '#FF3344',
											'Down'       = '#0077CC'
										))

Questions?

Exercise 2

Task: Create a plot with ggplot2!

Iris_ViolinPlot_quasirandom
Iris Dataset :: Violin Plot 
								
									library(tidyverse)
									library(ggbeeswarm)

									data(iris)

									ggplot(iris, aes(x=..., y=...)) + 
										...
Ruhland2016_VolcanoPlot

15 min

alarm_clock
Ruhland 2016 RNA-seq :: Violin Plot 
								
									# Load data
									ruhland2016 <- read.csv('Ruhland2016.csv', row.names=1)

									# Mark differentially expressed genes - Wrap this part in a reactive({})
									log2FCthreshold <- log2(1.5)
									p_threshold     <- 0.05

									ruhland2016$Expression <- 'not sign.'
									ruhland2016$Expression[ruhland2016$log2FoldChange >= log2FCthreshold & ruhland2016$pval <= p_threshold]  <- 'Up'
									ruhland2016$Expression[ruhland2016$log2FoldChange <= -log2FCthreshold & ruhland2016$pval <= p_threshold] <- 'Down'
										...