Data Visualization with ggplot2

Introduction to ggplot2

ggplot2 is a powerful package for creating beautiful visualizations in R. It’s based on the “Grammar of Graphics,” which means we build plots layer by layer. Think of it like making a sandwich:

1. First, you need a base (the data)
2. Then add your main ingredients (geometric shapes like points, lines, or bars)
3. Finally, add your toppings (colors, labels, themes)

Let’s start by loading the required packages and our sample data:

# Load required packages for data visualization
library(tidyverse)  # Includes ggplot2 and data manipulation tools
library(viridis)    # Color-blind friendly color palettes

# Create sample gene expression dataset
gene_data <- tibble(
  gene_id = c("BRCA1", "TP53", "EGFR", "KRAS", "HER2"),        # Gene names
  control_1 = c(100, 150, 80, 200, 120),                        # Control replicate 1
  control_2 = c(110, 140, 85, 190, 125),                        # Control replicate 2
  treated_1 = c(200, 300, 90, 180, 240),                        # Treatment replicate 1
  treated_2 = c(190, 280, 95, 185, 230),                        # Treatment replicate 2
  chromosome = c("17", "17", "7", "12", "17"),                  # Chromosome location
  pathway = c("DNA repair", "Cell cycle", "Growth", "Signaling", "Growth")  # Biological pathway
)

# Convert data from wide to long format for plotting
gene_data_long <- gene_data %>%
  pivot_longer(
    cols = c(control_1, control_2, treated_1, treated_2),  # Columns to convert
    names_to = "sample",                                   # New column for sample names
    values_to = "expression"                              # New column for expression values
  )

# Basic bar plot showing gene expression levels
ggplot(gene_data_long, aes(x = gene_id, y = expression)) +     # Map genes to x-axis, expression to y-axis
  geom_bar(stat = "identity") +                                # Create bars with heights = y values
  labs(title = "Gene Expression Levels",                       # Add plot title
       x = "Gene",                                            # Label x-axis
       y = "Expression Level")                                # Label y-axis

# Grouped bar plot showing expression by sample
ggplot(gene_data_long, aes(x = gene_id, y = expression, fill = sample)) +  # Add color grouping
  geom_bar(stat = "identity", position = "dodge") +           # Place bars side by side
  labs(title = "Gene Expression Levels by Sample",            # Add title
       x = "Gene",                                           # Label x-axis
       y = "Expression Level") +                             # Label y-axis
  theme_minimal() +                                          # Use minimal theme
  theme(axis.text.x = element_text(angle = 45, hjust = 1))  # Rotate x-axis labels

# Calculate summary statistics for each gene
gene_summary <- gene_data_long %>%
  group_by(gene_id) %>%                                      # Group data by gene
  summarise(
    mean_expr = mean(expression),                            # Calculate mean expression
    sd_expr = sd(expression),                                # Calculate standard deviation
    sem_expr = sd(expression) / sqrt(n())                    # Calculate standard error
  )

# Basic box plot showing expression distribution
ggplot(gene_data_long, aes(x = gene_id, y = expression)) +   # Map genes and expression
  geom_boxplot() +                                           # Create box plot
  labs(title = "Distribution of Expression Levels",          # Add title
       x = "Gene",                                          # Label x-axis
       y = "Expression Level")                              # Label y-axis

# Box plot with individual points and error bars
ggplot(gene_data_long, aes(x = gene_id, y = expression, fill = gene_id)) +  # Add color by gene
  geom_boxplot(alpha = 0.7) +                               # Add semi-transparent boxes
  geom_point(position = position_jitter(width = 0.2)) +     # Add jittered data points
  labs(title = "Gene Expression Distribution with Data Points",  # Add title
       x = "Gene",                                          # Label x-axis
       y = "Expression Level") +                            # Label y-axis
  theme_minimal() +                                         # Use minimal theme
  theme(legend.position = "none")                          # Remove legend

# Calculate mean expression for control and treated conditions
expression_means <- gene_data %>%
  mutate(
    mean_control = (control_1 + control_2) / 2,             # Average control replicates
    mean_treated = (treated_1 + treated_2) / 2              # Average treated replicates
  )

# Create scatter plot comparing control vs treated expression
ggplot(expression_means, 
       aes(x = mean_control, y = mean_treated, label = gene_id)) +  # Map control vs treated
  geom_point() +                                            # Add points
  geom_text(vjust = -0.5) +                                # Add labels above points
  geom_abline(intercept = 0, slope = 1,                    # Add y=x reference line
              linetype = "dashed", color = "red") +         # Make line dashed and red
  labs(title = "Control vs Treated Expression",            # Add title
       x = "Mean Control Expression",                      # Label x-axis
       y = "Mean Treated Expression") +                    # Label y-axis
  theme_minimal()                                         # Use minimal theme

# Scatter plot with additional pathway information
ggplot(expression_means, 
       aes(x = mean_control, y = mean_treated, 
           color = pathway, label = gene_id)) +             # Color points by pathway
  geom_point(size = 3) +                                   # Add larger points
  geom_text(vjust = -0.5) +                               # Add labels above points
  geom_abline(intercept = 0, slope = 1,                   # Add reference line
              linetype = "dashed", color = "gray") +       # Gray dashed line
  labs(title = "Control vs Treated Expression by Pathway", # Add title
       x = "Mean Control Expression",                     # Label x-axis
       y = "Mean Treated Expression") +                   # Label y-axis
  theme_minimal()                                        # Use minimal theme

# Create basic histogram of expression values
ggplot(gene_data_long, aes(x = expression)) +              # Map expression to x-axis
  geom_histogram(bins = 10) +                              # Create histogram with 10 bins
  labs(title = "Distribution of Expression Values",        # Add title
       x = "Expression Level",                            # Label x-axis
       y = "Count")                                       # Label y-axis

# Histogram with density curve
ggplot(gene_data_long, aes(x = expression)) +
  geom_histogram(aes(y = ..density..),                    # Convert to density scale
                 bins = 10,                               # Number of bins
                 fill = "lightblue",                      # Bar color
                 alpha = 0.7) +                           # Bar transparency
  geom_density(color = "red") +                          # Add density curve in red
  labs(title = "Distribution of Expression Values with Density Curve",  # Add title
       x = "Expression Level",                           # Label x-axis
       y = "Density") +                                  # Label y-axis
  theme_minimal()                                        # Use minimal theme

# Faceted histogram by sample type
ggplot(gene_data_long, aes(x = expression, fill = sample)) +  # Map expression and color
  geom_histogram(bins = 10,                              # Number of bins
                 alpha = 0.7,                            # Bar transparency
                 position = "identity") +                # Overlay histograms
  facet_wrap(~sample) +                                 # Create separate plot per sample
  labs(title = "Expression Distribution by Sample",     # Add title
       x = "Expression Level",                         # Label x-axis
       y = "Count") +                                  # Label y-axis
  theme_minimal()                                      # Use minimal theme

Basic ggplot2 Syntax

Every ggplot2 plot has three key components:

1. Data: The dataset you want to visualize
2. Aesthetics (aes): How your data maps to visual properties (x, y, color, size, etc.)
3. Geometries (geom_*): The type of plot you want to create

The basic template is:

ggplot(data = your_data, aes(x = x_variable, y = y_variable)) +
  geom_something()

1. Bar Plots

Let’s start with a simple bar plot showing gene expression levels:

# Basic bar plot
ggplot(gene_data_long, aes(x = gene_id, y = expression)) +
  geom_bar(stat = "identity") +
  labs(title = "Gene Expression Levels",
       x = "Gene",
       y = "Expression Level")

# Grouped bar plot
ggplot(gene_data_long, aes(x = gene_id, y = expression, fill = sample)) +
  geom_bar(stat = "identity", position = "dodge") +
  labs(title = "Gene Expression Levels by Sample",
       x = "Gene",
       y = "Expression Level") +
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))  # Rotate x-axis labels

2. Box Plots

Box plots are great for showing the distribution of data and identifying outliers:

# Calculate some summary statistics
gene_summary <- gene_data_long %>%
  group_by(gene_id) %>%
  summarise(
    mean_expr = mean(expression),
    sd_expr = sd(expression),
    sem_expr = sd(expression) / sqrt(n())  # Standard Error of Mean
  )

# Basic box plot
ggplot(gene_data_long, aes(x = gene_id, y = expression)) +
  geom_boxplot() +
  labs(title = "Distribution of Expression Levels",
       x = "Gene",
       y = "Expression Level")

# Box plot with individual points and error bars
ggplot(gene_data_long, aes(x = gene_id, y = expression, fill = gene_id)) +
  geom_boxplot(alpha = 0.7) +  # Alpha controls transparency
  geom_point(position = position_jitter(width = 0.2)) +  # Add individual points
  labs(title = "Gene Expression Distribution with Data Points",
       x = "Gene",
       y = "Expression Level") +
  theme_minimal() +
  theme(legend.position = "none")  # Remove legend

# Understanding the Box Plot Elements:
# - Box: Shows the Interquartile Range (IQR) - middle 50% of data
# - Line in box: Median
# - Whiskers: Extend to most extreme points within 1.5 * IQR
# - Points beyond whiskers: Potential outliers

3. Scatter Plots

Scatter plots are useful for showing relationships between two variables:

# Calculate mean expression for control and treated conditions
expression_means <- gene_data %>%
  mutate(
    mean_control = (control_1 + control_2) / 2,
    mean_treated = (treated_1 + treated_2) / 2
  )

# Basic scatter plot
ggplot(expression_means, 
       aes(x = mean_control, y = mean_treated, label = gene_id)) +
  geom_point() +
  geom_text(vjust = -0.5) +  # Add labels above points
  geom_abline(intercept = 0, slope = 1, linetype = "dashed", color = "red") +  # Add y=x line
  labs(title = "Control vs Treated Expression",
       x = "Mean Control Expression",
       y = "Mean Treated Expression") +
  theme_minimal()

# Scatter plot with additional information
ggplot(expression_means, 
       aes(x = mean_control, y = mean_treated, 
           color = pathway, label = gene_id)) +
  geom_point(size = 3) +
  geom_text(vjust = -0.5) +
  geom_abline(intercept = 0, slope = 1, linetype = "dashed", color = "gray") +
  labs(title = "Control vs Treated Expression by Pathway",
       x = "Mean Control Expression",
       y = "Mean Treated Expression") +
  theme_minimal()

4. Histograms

Histograms show the distribution of a continuous variable:

# Basic histogram
ggplot(gene_data_long, aes(x = expression)) +
  geom_histogram(bins = 10) +
  labs(title = "Distribution of Expression Values",
       x = "Expression Level",
       y = "Count")

# Histogram with density curve
ggplot(gene_data_long, aes(x = expression)) +
  geom_histogram(aes(y = ..density..), bins = 10, fill = "lightblue", alpha = 0.7) +
  geom_density(color = "red") +
  labs(title = "Distribution of Expression Values with Density Curve",
       x = "Expression Level",
       y = "Density") +
  theme_minimal()

# Faceted histogram by sample type
ggplot(gene_data_long, aes(x = expression, fill = sample)) +
  geom_histogram(bins = 10, alpha = 0.7, position = "identity") +
  facet_wrap(~sample) +
  labs(title = "Expression Distribution by Sample",
       x = "Expression Level",
       y = "Count") +
  theme_minimal()

5. Heatmaps

Heatmaps are excellent for visualizing expression patterns across multiple conditions:

# Prepare data for heatmap
heatmap_data <- gene_data %>%
  select(gene_id, control_1, control_2, treated_1, treated_2) %>%
  pivot_longer(
    cols = -gene_id,
    names_to = "sample",
    values_to = "expression"
  )

# Create heatmap
ggplot(heatmap_data, aes(x = sample, y = gene_id, fill = expression)) +
  geom_tile() +
  scale_fill_viridis() +  # Color-blind friendly palette
  labs(title = "Gene Expression Heatmap",
       x = "Sample",
       y = "Gene",
       fill = "Expression") +
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

# Heatmap with centered and scaled values
heatmap_data_scaled <- heatmap_data %>%
  group_by(gene_id) %>%
  mutate(scaled_expression = scale(expression)[,1]) %>%
  ungroup()

ggplot(heatmap_data_scaled, 
       aes(x = sample, y = gene_id, fill = scaled_expression)) +
  geom_tile() +
  scale_fill_viridis(option = "magma") +
  labs(title = "Gene Expression Heatmap (Scaled)",
       x = "Sample",
       y = "Gene",
       fill = "Z-score") +
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

Customizing Plots

You can customize almost every aspect of your plots. Here are some common modifications:

# Example of a highly customized plot
ggplot(gene_data_long, aes(x = gene_id, y = expression, fill = sample)) +
  geom_boxplot(alpha = 0.7) +
  geom_point(position = position_jitterdodge(jitter.width = 0.2)) +
  scale_fill_viridis_d() +  # Discrete color-blind friendly palette
  labs(
    title = "Gene Expression Analysis",
    subtitle = "Comparing Control and Treated Samples",
    x = "Gene",
    y = "Expression Level",
    fill = "Sample Type"
  ) +
  theme_minimal() +
  theme(
    plot.title = element_text(size = 16, face = "bold"),
    plot.subtitle = element_text(size = 12, color = "gray50"),
    axis.title = element_text(size = 12),
    axis.text.x = element_text(angle = 45, hjust = 1),
    legend.position = "top"
  )

Common Plot Modifications

Here are some useful modifications you might want to apply to your plots:

1. Changing Themes: ```r

   Same plot with different themes

   base_plot <- ggplot(gene_data_long, aes(x = gene_id, y = expression)) + geom_boxplot()

Default theme

base_plot

Minimal theme

base_plot + theme_minimal()

Classic theme

base_plot + theme_classic()

Black and white theme

base_plot + theme_bw()

2. **Modifying Axes**:
```r
base_plot +
  scale_y_log10() +  # Log scale for y-axis
  theme_minimal() +
  theme(
    axis.text.x = element_text(angle = 45, hjust = 1),  # Rotate x labels
    axis.title = element_text(size = 12, face = "bold")  # Bold axis titles
  )

1. Adding Labels and Annotations:

   ggplot(expression_means, 
       aes(x = mean_control, y = mean_treated)) +
     geom_point() +
     geom_text(aes(label = gene_id), vjust = -0.5) +  # Add labels
     geom_abline(intercept = 0, slope = 1, linetype = "dashed", color = "red") +  # Add line
     annotate("text", x = 100, y = 200, 
           label = "Upregulated", color = "blue", fontface = "italic") +  # Add annotation
     theme_minimal()
   

Practice Exercises

1. Create a scatter plot showing the relationship between control_1 and control_2 values
2. Make a box plot comparing expression levels across pathways
3. Create a heatmap of log2-transformed expression values
4. Make a histogram of fold changes between treated and control conditions

# Exercise 1: Scatter plot of technical replicates
ggplot(gene_data, aes(x = control_1, y = control_2, label = gene_id)) +
  geom_point() +
  geom_text(vjust = -0.5) +
  geom_abline(intercept = 0, slope = 1, linetype = "dashed", color = "red") +
  labs(title = "Technical Replicates Comparison",
       x = "Control Replicate 1",
       y = "Control Replicate 2") +
  theme_minimal()

# Exercise 2: Box plot by pathway
ggplot(gene_data_long, aes(x = pathway, y = expression, fill = pathway)) +
  geom_boxplot() +
  labs(title = "Expression Levels by Pathway",
       x = "Pathway",
       y = "Expression Level") +
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

Tips for Creating Good Plots

1. Keep it Simple
   • Don’t overload your plots with unnecessary information
   • Use clear, readable fonts and appropriate sizes
   • Choose colors wisely (consider color-blind friendly options)
2. Label Everything
   • Always include axis labels
   • Use informative titles
   • Add units where appropriate
3. Consider Your Audience
   • Make sure your plot tells a clear story
   • Use appropriate scales
   • Add explanatory notes if needed
4. Technical Tips
   • Save high-resolution versions for publications
   • Be consistent with styling across related plots
   • Test your plots with different data sizes

Next Steps

After mastering these basics, you can explore:

• Interactive plots with plotly
• Multiple plots with patchwork
• Custom themes and color palettes
• Statistical visualizations
• Publication-ready figure preparation