In areyesq89/GenomeMatrix: Plot rotated matrices

knitr::opts_chunk$set(tidy = FALSE,
                      cache = TRUE,
                      dev = "png",
                      message = FALSE,
                      error = FALSE,
                      warning = TRUE)

Quick start

The minimum necessary to plot a matrix is simply a matrix. The function matrixPlotter() will rotate and plot only the upper half of the matrix. We specify the parameter heightProp=1/2 for the height of the plot (y-axis) to be one half of the x-axis.

library(GenomeMatrix)
set.seed(100)
mat <- matrix( rnorm( 100, 0, 1 ), ncol=10 )
matrixPlotter(mat, heightProp=1/2) 

This is an ugly plot, but in the following sections I show a more useful example, with HiC data.

HiC data

Below, I use a publicly available HiC dataset to demonstrate the main function. By providing a matrix, a GenomicRanges object where each element corresponds to a row (or a column) of the matrix and specifying the right zlim= and heightProp= parameters, we can visualize TADs horizontally.

library(HiTC)
data(Dixon2012_IMR90, package="HiCDataHumanIMR90")
mat <- as.matrix( hic_imr90_40@.Data[[1]]@intdata )
granges <- hic_imr90_40@.Data[[1]]@xgi
matrixPlotter( log2(mat+1), granges, 
               plotGR=GRanges("chr1", IRanges( 50000000, 60000000 ) ), 
               zlim=5, heightProp=1/4) 

Note that the output is a ggplot2 object. Thus it is straightforward to arrange this into a single panel with other ggplot2 plots, such as gene annotations from the ggbio package, using packages like ggpubr.

areyesq89/GenomeMatrix documentation built on May 24, 2019, 2:51 p.m.