In teunbrand/tornadoplot: Tornado Plot

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-"
)

tornadoplot

A tornadoplot* is a heatmap of coverage from genome wide assays centred at several features. The depth of coverage is encoded in a sequential colour scale. Every row in the heatmap is a genomic locus and every column is a position relative to the centre of some feature. Arranging the rows from high to low coverage gives this heatmap a characteristic shape resembling a tornado. This type of plot is common in the epigenetics and chromatin fields of molecular biology.

The {tornadoplot} package facilitates making these types of heatmaps in R with Bioconductor infrastructure. It can use precalculated coverage data from bigwig files, or calculate coverage de novo from tabix files or GRanges objects in R. The data can then be plotted with help from the {ggplot2} package.

* This is not a widely used term. It is just more imaginative than yet another type of 'heatmap', which may display anything really.

Installation

You can install the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("teunbrand/tornadoplot")

Example

This is a basic example which shows you how to make a tornado plot. First, we need to decide what data we want to plot. If you're downloading data from the internet in a script, you can use the {BiocFileCache} package to avoid repeatedly downloading data and alleviate traffic on the provider. Below, we download and cache two bigwig files from Sox2 and Oct4 ChIP-seq experiments [1].

library(BiocFileCache)
bfc <- BiocFileCache()

# Declare files
# These bigwigs are derived from PMID: 28212747
sources <- c(
  "sox2" = "http://dc2.cistrome.org/genome_browser/bw/73468_treat.bw",
  "oct4" = "http://dc2.cistrome.org/genome_browser/bw/73466_treat.bw"
)
bigwigs <- setNames(bfcrpath(bfc, sources), c("Sox2 ChIP", "Oct4 ChIP"))

The following decision is what parts of the data to plot. For purposes of this example, the octsox_peaks object is a preprocessed set of genomic loci where Oct4 and Sox2 bind alone or together.

library(tornadoplot)

# These are a preprocessed set of ChIP-seq peaks
feats <- octsox_peaks

# Recall that the files are named, which the next function uses for labels
bigwigs <- BigWigFileList(bigwigs)

# Give the function locations and data
tornado_plot(features = feats, data = bigwigs)

Related works

• The deepTools suite of python tools is perhaps best known for making these types of plots through a combination of their computeMatrix and plotHeatmap combination.

References

[1] Liu, Ziyang et al. (2017) Catalytic-Independent Functions of PARP-1 Determine Sox2 Pioneer Activity at Intractable Genomic Loci. Molecular Cell 65 4 p589-603 https://doi.org/10.1016/j.molcel.2017.01.017

teunbrand/tornadoplot documentation built on Dec. 23, 2021, 8:48 a.m.