Skip to content

Advice/Tips data visualization

Covering

  • Layout
  • Gestalt principles (how people organize visual information)
  • Color usage
  • Plot specific parts
    • Heatmaps
    • Networks
  • Plot horrors

BBGallery of plots (legacy from BBGcloud)

Other galleries

Other resources

  • awesome-genome-visualization: A curated list of tools and libraries for genome visualization, including interactive and static visualization options.

Color usage

Perception of color

Some important considerations when selecting a color palette include:

  • Does the palette retain its integrity when printed in black and white? Gradients are a good idea to satisfy this criteria, e.g. the viridis palettes.
  • Are people with colorblindness able to understand it? Color blindness simulators. Can help with that. Also, the colorblindr R package can simulate color blindness to ggplot objects.

Color palettes

  • ColoRBrewer: Classic tool for picking colorblind-safe and print-friendly palettes.
  • Viridis: Perceptually uniform and colorblind-friendly.
  • Color Palette Finder (Also accessible with the paletteer R package).
  • Coolors: Nice tool to explore and create palettes (but limited to ~5 colors).

Creating reusable themes in R

Creating reusable themes in R is a great way to ensure consistency and save time across visualizations.

1. Create a Custom Theme Function

library(ggplot2)

theme_tidy <- function(base_size = 12, base_family = "sans") {
  theme_minimal(base_size = base_size, base_family = base_family) +
    theme(
      plot.title = element_text(face = "bold", size = base_size + 2, hjust = 0.5),
      plot.subtitle = element_text(size = base_size),
      plot.caption = element_text(size = base_size - 2, color = "gray40"),
      axis.title = element_text(face = "bold"),
      axis.text = element_text(size = base_size - 1),
      panel.grid.major = element_line(color = "gray85"),
      panel.grid.minor = element_blank(),
      legend.position = "bottom",
      legend.title = element_text(face = "bold"),
      strip.text = element_text(face = "bold", size = base_size),
      plot.background = element_rect(fill = "white", color = NA)
    )
}

2. Apply to any ggplot

ggplot(mtcars, aes(x = mpg, y = hp)) +
  geom_point(color = "#2c7fb8") +
  labs(title = "Horsepower vs. MPG") +
  theme_tidy()

3. Make It Reusable Across Projects

Save theme and palette functions in a script (e.g., mythemes.R) and source it:

source("path/to/mythemes.R")

4. Set Your Theme Globally

Use your theme across all plots in a session:

theme_set(theme_tidy())

Colour resources for plots

  • Predefined color names
  • Predefined color palettes
  • Generate customized palettes
    • iWantHue: generate and refine palettes of 1 to more than 30 optimally distinct colors, with option to set up the number of colours and hue. Option to export the colour IDs from your palettes as json or lists in different formats (rgb, hex...) and save your own palettes when registered.
    • coolors: set up from 1 to 10 colour palettes searching for colours, topics, styles, popularity, hex values... Possible to save your own palettes when registered.

Reference

  • Axel Rosendahl Huber
  • Elisabet Figuerola
  • Laura Torrens