ggefp

ggefp is an extension of the ggplot2 package that provides functionality to plot line drawings with regions filled according to provided data. With the right line drawings, ggefp can be used to provide data visualization similar to those known from the eFP browser that shows gene expression values in various tissues in plants and other species.

To install or update, run:

library(devtools)
install_github("ggefp", subdir="pkg")

Usage

The line drawings to display are denoted exhibits and define the line drawing as well as which paths correspond to which tissue. To see the exhibits bundled with the package (currently the main Arabidopsis developmental map kindly provided by Nicholas Provart)

library(ggefp)
head(exhibits())

##             exhibit       tissue  from
## 1 ath_dormant_seeds imbibed-seed ggefp
## 2 ath_dormant_seeds     dry-seed ggefp
## 3 ath_flower_series        bud-9 ggefp
## 4 ath_flower_series       bud-10 ggefp
## 5 ath_flower_series       bud-11 ggefp
## 6 ath_flower_series       bud-12 ggefp

We can define a combination of exhibits to use to a collection by specifiying the names of the exhibits and where to plot them.

collection <- data.frame(exhibit=c('ath_seedling', 'ath_young_plant'),
                 x=c(.2, .7), y=c(.5, .5), width=c(.4,.6), height=c(1,1))

we can then define some data and generate the plot

values <-
  data.frame(tissue=rep(c('young-root', 'hypocotyl', 'cotelydons',
               'vegetative-rosette', 'root'), 2),
             value=c(1,200,20, 10, 10, 900, 100, 40,60,900),
             treatment=rep(c('control', 'cold'), each=5))

ggplot(values, aes(tissue=tissue, fill=value)) +
  geom_efp(collection) +
  facet_wrap(~treatment) + expand_limits(collection) +
  xlim(0,1) + ylim(0,1) + theme_efp()

Or we can draw the entire Arabidopsis developmental map for some fake data:

dd <- data.frame(tissue=exhibits()$tissue)
dd$val <- abs(rnorm(nrow(dd),sd=40))

ggplot(dd, aes(tissue=tissue, fill=val)) +
  geom_efp(ath_developmental_map) +
  expand_limits(ath_developmental_map) + 
  xlim(0,1) + ylim(0,1)

New exhibits

New exhibits should first be generated in a vector graphics program such as inkscape where each tissue is filled with a unique color. These colors are matched with the tissue name and exhibits are created according to the trace.R file. Exhibits from outside the ggefp package are named as paths to saved R-objects (one object per rda file).