The bio3d.ggplot package currently provides two categories of functions:
gg.*() and gg_*() functions. This first produce ggplot objects from
bio3d package related data. The later add layers to the former.
Currently the main function is gg.mat(), which returns a
ggplot version of the old bio3d plot.dmat() function for NxN matrix data.
This function also works well for difference distance matrices (DDMs),
correlation matrices (substituting for plot.dccm()) and contact maps
(substituting for plot.cmap()).
Also included is a basic secondary structure element annotation function
called gg_sse() that provides the classic alpha helix and beta
strand rectangles seen in bio3d plot functions since day one. Options for
more fancy helices will follow.
This a work in progress.
As things develop I will likely introduce S3 based classes that call this one
master matrix plot function with suitable settings. As well as one master
vector plot function (the later like plot.bio3d(). The the gg_*()
annotation functions should work for all of these in the same way.
To install:
devtools::install_github("bioboot/bio3d.ggplot")
To make some nice figures (i.e. nicer than base graphics versions):
library(bio3d.ggplot)
library(bio3d)
library(ggplot2)
## Run the examples below
example(bio3d.ggplot)
Or run each example in turn (these are in the man/help page for the package):
##- Single structure distance matrix
pdb <- read.pdb( "5p21" )
k <- dm(pdb, inds="calpha", mask.lower=FALSE)
gg.mat(k)
## Add secondary structure 'annotation layer' to plot
p <- gg.mat(k) + gg_sse(pdb) ## save in an object 'p'
p ## produce the plot
## Data driven axis from SSE boundaries
sse_labels <- c(pdb$helix$start, pdb$sheet$start,
pdb$helix$end, pdb$sheet$end)
x <- sort(sse_labels); x <- x[diff(x)> 4]
## Add your customization with additional layers
p + theme_grey() + ## The ggplot2 gray background theme
scale_fill_gradient(high = "orange", low = "white") +
scale_y_continuous("My Y label",breaks=x, labels=x, expand=c(0,0.5)) +
scale_x_continuous("My X label",breaks=x, labels=x, expand=c(0,0.5))
gg.mat(k) + theme_minimal(base_size = 14) +
scale_x_continuous(expand = c(0, 0)) +
scale_y_continuous(expand = c(0, 0)) ## start axis at zero
##- Calculate and plot correlation matrix
cij <- dccm.nma(nma(pdb))
q <- gg.mat(cij) + gg_sse(pdb)
q
q + scale_fill_gradient2(limit = c(-1,1),
high = "red", mid = "white", low = "blue")
##- Difference distance matrices (DDM) of heterogeneous structures
pdbs <- pdbaln( c("5p21","4q21") )
mat <- dm(pdbs, mask.lower=FALSE)
ddm <- mat[,,1] - mat[,,2]
gg.mat(ddm) + gg_sse(pdbs)
##- Contact maps
cm <- cmap(pdb, scut=0, mask.lower=FALSE)
gg.mat(cm) + gg_sse(pdb) + theme(legend.position="none")