Create interactive circos plots from tidy data.frames.

Circos plots can be built up layer by layer.

You can install the development version of ggcircos from GitHub with:

# install.packages("pak")
pak::pak("selkamand/ggcircos")

Lets build a standard genomic SNV circos plot up from scratch.

1. Start by creating the basic reference structure. You can either supply a custom .fai index of your reference genome, or use one of our prebuilt templates for common reference genomes (e.g. reference_hg38).

library(ggcircos)

## Read in Data
fai <- system.file("hg38.fai", package = "ggcircos")
centromeres_tsv <- system.file("hg38.centromeres.fai", package = "ggcircos")
telomeres_tsv <- system.file("hg38.telomers.fai", package = "ggcircos")
mutations_tsv <- system.file("hg38.mutations.fai", package = "ggcircos")
copynumber_tsv <- system.file("hg38.segments.tsv", package = "ggcircos")
methylation_tsv <- system.file("hg38.methylation.tsv", package = "ggcircos")

# Read in data
hg38 <- fai_to_reference(fai)
centromeres <- read.csv(centromeres_tsv, sep = "\t")
telomeres <- read.csv(telomeres, sep = "\t")
mutations <- read.csv(mutations_tsv, sep = "\t")
methylation <- read.csv(methylation_tsv, sep = "\t")

# Create Circos Plot
multiomic_circos <- ggcircos(hg38) |>
  add_centromeres(centromeres) |>
  add_telomeres(telomeres) |>
  add_mutations(mutations) |>
  add_copynumber(copynumber_segments) |>
  add_methylation(methylation_tsv)

# Make interactive
make_interactive(multiomic_circos)

ggcircos provides tools to help parse different fileformats into the data.frames ggcircos expects.

Circos plots rendered as vector graphics are larger and more computationaly demanding than raster alternatives, especially when viewing highly mutated samples. Adding interactivity bloats the files even more, since SVGs must be embedded with tooltip information. If your circos plot is too large or not responsive, consider only adding tooltips to a subset of your variants, or decreasing the size of the tooltips.

One day we may add the ability to render ggcircos output as raster graphics, or export the data required to plug into alternate interactive circos visualisation apps that make better use locally available compute (CPU & GPU).