Tracks for genomic graphics

Description

tracks is a conventient constructor for bindind graphics as trakcs. You dont' have to worry about adjusting different graphics, tracks did that for you. It's NOT just limited to bind genomic tracks, you can use this function to bind any tracks with the same defination of x axis, for example, sets of time series plots you made.

Tracks view is most common way to viewing genome features and annotation data and widely used by most genome browsers. Our assumption is that, most graphics you made with ggbio or by yourself using ggplot2, are almost always sitting on the genomic coordinates or the same x axis. And to compare annotation information along with genome features, we need to align those plots on exactly the same x axis in order to form your hypothesis. This function leaves you the flexibility to construct each tracks separately with worrying your alignments later.

Usage

tracks(..., heights, xlim, xlab = NULL, main = NULL,
            title = NULL, theme = NULL, fixed = NULL,
            track.plot.color = NULL,
            track.bg.color = NULL,
            main.height = unit(2, "lines"),
            scale.height = unit(2, "lines"),
            xlab.height = unit(2, "lines"),
            padding = unit(0, "lines"),
            label.bg.color =  "white",
            label.bg.fill = "gray80",
            label.text.color = "black",
            label.text.cex = 1,                   
            label.width = unit(2.5, "lines"))

Arguments

Details

tracks did following modification for passed plots.

• remove x-axis, ticks, xlab and tile, for each track and add a scale track at bottom. We suppose a new xlab and title would be provided by the tracks function for the whole tracks, but we still keep y axis.

• align x-scale limits to make sure every plots sitting on exactly the same x scale.

• squeezing plots together a little bit.

• labeling tracks if names provided, please check utilities section about labeled method.

• return a track object. This would allow many features introduced in this manual.

Value

A Tracks object.

Track class

constructor tracks will return a Tracks object, which has following slots.

grobs

a ggplotGrobList object contains a list of ggplot object, which is our passed graphics.

backup

a backup of all the slots for holding the original tracks, so users could edit it and reset it back at any time later, and backup method will reset the backupped copy.

ylim

y limits for each plot.

labeled

vector of logical value indicates whether a track is labeled or not, for labeled attributes please check utilities section.

mutable

vector of logical value indicates whether a track is mutable for theme editing or not, for mutable attributes please check utilities section.

hasAxis

vector of logical value indicates whether a track has axis or not, for hasAxis attributes please check utilities section.

heights, xlim, xlab, main, title, theme, fixed, track.plot.color, track.bg.color, main.height, scale.height, xlab.height, padding, label.bg.color, label.bg.fill, label.text.color, label.text.cex, label.width

those slots are described in arguments section for constructor.

Utilities

Please check examples for usage.

summary(object)

summary information about tracks object.

fixed(x), fixed(x) <- value

x is the ggplot object, this controls if a track has a fixed x scale or not, if the fixed attributes is TRUE, then when you pass this plot to a tracks, this plot won't be re-aligned with other tracks and will keep the original x-axis, this allow you to pass some plot like ideogram. fixed function will return a logical value

labeled(x), labeled(x) <- value

x is the ggplot object, if you pass named graphics into tracks, it will create the labels on the left for you. Several ways supported to name it. You can pass a list of graphics with names. Or you can use tracks('name1' = p1, 'name 2' = p2, ...) with quotes for complicated words or simply tracks(part1 = p1, part = p2, ...).

mutable(x), mutable(x) <- value

x is the ggplot object, this controls whether a plot in the tracks mutable to theme changing or not, when you use + method for Tracks object, add-on edit will only be applied to the the mutable plots.

bgColor(x), bgColor(x) <- value

x is the ggplot object, this change the background color for single plot shown in the tracks.

xlim(x), xlim(x) <- value

when x is the numeric value, it calls ggplot2::coord_cartesian(xlim = ...) method, we doesn't use ggplot2::xlim() for the reason it will cut data outside the range, and we believe the best behavior would be zoom-in/out like most browser. when x is IRanges, GRanges, it get the range and passed to ggplot2::coord_cartesian function.

when x is Tracks object, xlim(x) will return x limits for that tracks. xlim(x) <- value replace method only works for Tracks object. value could be numeric, IRanges, GRanges object. This will change the x limits associated with tracks.

+ xlim(obj):obj is the numeric range, or IRanges, GRanges object.

+ coord_cartesian(): please read manual in ggplot2, this controls both xlim an ylim, only accept numerical range.

+

The most nice features about Tracks object is the one inherited from ggplot2's components additive features, with + method you can use any theme object and utilities in ggplot2 package, to add them on a Tracks object, for example, if x is our Tracks object, x + theme would apply theme to any plots in the tracks except those are immutable.

Backup and reset

reset(obj)

obj is the Tracks object, this reset the tracks back to original or backuped version.

backup(obj)

obj is the Tracks object, this clear previous backup and use current setting for a new backup.

update(obj, xlim)

obj is the Tracks object, this allow a quick tweak with xlim when the plot is shown on your screen.

Author(s)

Tengfei Yin

See Also

align.plots

Examples

## make a simulated time series data set
df1 <- data.frame(time = 1:100, score = sin((1:100)/20)*10)
p1 <- qplot(data = df1, x = time, y = score, geom = "line")
df2 <- data.frame(time = 30:120, score = sin((30:120)/20)*10, value = rnorm(120-30 + 1))
p2 <- ggplot(data = df2, aes(x = time, y = score)) + 
  geom_line() + geom_point(size = 4, aes(color = value))
## check p2
p1

## check p2
p2

## binding
tracks(p1, p2)

## or
tks <- tracks(p1, p2)
tks

## labeling: default labeling a named graphic
## simply pass a name with it
tracks(time1 = p1, time2 = p2)

## or pass a named list with it
lst <- list(time1 = p1, time2 = p2)
tracks(lst)

## more complicated case please use quotes
tracks(time1 = p1, "second time" = p2)

## set heights
tracks(time1 = p1, time2 = p2, heights = c(1, 3))

## if you want to disable label arbitrarily
## default label is always TRUE
labeled(p2) 

## [1] TRUE

labeled(p2) <- FALSE
## set labeled to FALSE, remove label even the plot has a name
tracks(time1 = p1, time2 = p2)

labeled(p2) <- TRUE
## fix a plot, not synchronize with other plots
p3 <- p1
## default is always FALSE
fixed(p3)

## [1] FALSE

## set to TRUE
fixed(p3) <- TRUE
fixed(p3)

## [1] TRUE

tracks(time1 = p1, time2 = p2, "time3(fixed)" = p3) 

fixed(p3) <- FALSE
## otherwise you could run
tracks(time1 = p1, time2 = p2, "time3(fixed)" = p3, fixed = c(FALSE, FALSE, TRUE))

## control axis
hasAxis(p1)

## [1] FALSE

hasAxis(p1) <- TRUE
# ready for weird looking
tracks(time1 = p1, time2 = p2)

# set it back
hasAxis(p1) <- FALSE
## mutable
mutable(p1)

## [1] TRUE

tracks(time1 = p1, time2 = p2) + theme_bw()

mutable(p1) <- FALSE
# mutable for "+" method
tracks(time1 = p1, time2 = p2) + theme_bw()

mutable(p1) <- TRUE
## bgColor
bgColor(p1)

## [1] "white"

tracks(time1 = p1, time2 = p2)

bgColor(p1) <- "brown"
# mutable for "+" method
tracks(time1 = p1, time2 = p2)

# set it back
bgColor(p1) <- "white"
## apply a theme to each track
tks <- tracks(time1 = p1, time2 = p2) + theme_bw()
tks

reset(tks)

## store it with tracks
tks <- tracks(time1 = p1, time2 = p2, theme = theme_bw())
tks

tks <- tks + theme_gray()
tks

## reset will be introduced later
reset(tks)

## apply a pre-defiend theme for tracks!
tracks(time1 = p1, time2 = p2) + theme_tracks_sunset()

 tracks(p1, p2) + theme_tracks_sunset()

## change limits
tracks(time1 = p1, time2 = p2) + xlim(c(1, 40))

tracks(time1 = p1, time2 = p2) + xlim(1, 40)

tracks(time1 = p1, time2 = p2) + coord_cartesian(xlim = c(1, 40))

# change y
tracks(time1 = p1, time2 = p2) + xlim(1, 40) + ylim(0, 10)

## Warning: Removed 38 rows containing missing values (geom_path).

## Warning: Removed 58 rows containing missing values (geom_path).

## Warning: Removed 58 rows containing missing values (geom_point).

## Warning: Removed 38 rows containing missing values (geom_path).

library(GenomicRanges)

## Loading required package: BiocGenerics

## Attaching package: 'BiocGenerics'

## The following object(s) are masked from 'package:stats':
## 
## xtabs

## The following object(s) are masked from 'package:base':
## 
## anyDuplicated, cbind, colnames, duplicated, eval, Filter, Find, get,
## intersect, lapply, Map, mapply, mget, order, paste, pmax, pmax.int, pmin,
## pmin.int, Position, rbind, Reduce, rep.int, rownames, sapply, setdiff,
## table, tapply, union, unique

## Loading required package: IRanges

gr <- GRanges("chr", IRanges(1, 40))
# GRanges
tracks(time1 = p1, time2 = p2) + xlim(gr)

# IRanges
tracks(time1 = p1, time2 = p2) + xlim(ranges(gr))

tks <- tracks(time1 = p1, time2 = p2)
xlim(tks)

## [1] -10.9 131.9

xlim(tks) <- c(1, 35)
xlim(tks) <- gr
xlim(tks) <- ranges(gr)
## xlab, title
tracks(time1 = p1, time2 = p2, xlab = "time")

tracks(time1 = p1, time2 = p2, main = "title")

tracks(time1 = p1, time2 = p2, title = "title")

tracks(time1 = p1, time2 = p2, xlab = "time", title = "title") + theme_tracks_sunset()

## backup and restore
tks <- tracks(time1 = p1, time2 = p2)
tks

tks <- tks + xlim(1, 40)
tks

reset(tks)

tks <- tks + xlim(1, 40)
tks

tks <- backup(tks)
tks <- tks + theme_bw()
tks

reset(tks)

## update
## show it on the fly, save your time to re-print
tks <- tracks(time1 = p1, time2 = p2)
tks

update(tks, xlim = c(1, 40))

## padding(need to be fixed for more delicate control)
tracks(time1 = p1, time2 = p2, padding = 2)

## track color
tracks(time1 = p1, time2 = p2, track.bg.color = "yellow")

tracks(time1 = p1, time2 = p2, track.plot.color = c("yellow", "brown"))

## alignPlots simply align the panel, without adjusting the xlims
alignPlots(p1, p2)

alignPlots(p1, p2 + theme(legend.position = "none"), vertical = FALSE)