autoplot-method: Generic autoplot function

In lawremi/ggbio: Visualization tools for genomic data

Generic autoplot function

Description

autoplot is a generic function to visualize various data object, it tries to give better default graphics and customized choices for each data type, quick and convenient to explore your genomic data compare to low level ggplot method, it is much simpler and easy to produce fairly complicate graphics, though you may lose some flexibility for each layer.

Usage

## S4 method for signature 'GRanges'
autoplot(object, ..., chr, xlab, ylab, main, truncate.gaps = FALSE,
                 truncate.fun = NULL, ratio = 0.0025, space.skip = 0.1,
                 legend = TRUE, geom = NULL, stat = NULL,
                 chr.weight = NULL,
                 coord = c("default", "genome", "truncate_gaps"),
                 layout = c("linear", "karyogram", "circle"))


## S4 method for signature 'GRangesList'
autoplot(object, ..., xlab, ylab, main, indName = "grl_name",
                 geom = NULL, stat = NULL, coverage.col = "gray50",
                 coverage.fill = coverage.col, group.selfish = FALSE)


## S4 method for signature 'IRanges'
autoplot(object, ..., xlab, ylab, main)

## S4 method for signature 'Seqinfo'
autoplot(object, ideogram = FALSE, ... )

## S4 method for signature 'GAlignments'
autoplot(object, ..., xlab, ylab, main, which,
      geom = NULL, stat = NULL)

## S4 method for signature 'BamFile'
autoplot(object, ..., which, xlab, ylab, main,
                 bsgenome, geom = "line", stat = "coverage", method = c("raw",
                 "estimate"), coord = c("linear", "genome"),
                 resize.extra = 10, space.skip = 0.1, show.coverage =
                 TRUE)


## S4 method for signature 'character'
autoplot(object, ..., xlab, ylab, main, which)

## S4 method for signature 'TxDbOREnsDb'
autoplot(object, which, ..., xlab, ylab, main, truncate.gaps =
                 FALSE, truncate.fun = NULL, ratio = 0.0025,
                 mode = c("full", "reduce"),geom =
                 c("alignment"), stat = c("identity", "reduce"),
                 names.expr = "tx_name", label = TRUE)


## S4 method for signature 'BSgenome'
autoplot(object, which, ...,
                    xlab, ylab, main, geom = NULL)


## S4 method for signature 'Rle'
autoplot(object, ..., xlab, ylab, main, binwidth, nbin = 30,
          geom = NULL, stat = c("bin", "identity", "slice"),
          type = c("viewSums", "viewMins", "viewMaxs", "viewMeans"))

## S4 method for signature 'RleList'
autoplot(object, ..., xlab, ylab, main, nbin = 30, binwidth,
         facetByRow = TRUE, stat = c("bin", "identity", "slice"),
         geom = NULL, type = c("viewSums", "viewMins", "viewMaxs", "viewMeans"))

## S4 method for signature 'matrix'
autoplot(object, ..., xlab, ylab, main,
              geom = c("tile", "raster"), axis.text.angle = NULL,
              hjust = 0.5, na.value = NULL,
              rownames.label = TRUE, colnames.label = TRUE,
              axis.text.x = TRUE, axis.text.y = TRUE)


## S4 method for signature 'ExpressionSet'
autoplot(object, ..., type = c("heatmap", "none",
                 "scatterplot.matrix", "pcp", "MA", "boxplot",
                 "mean-sd"), test.method =
                 "t", rotate = FALSE, pheno.plot = FALSE, main_to_pheno
                 = 4.5, padding = 0.2)


## S4 method for signature 'RangedSummarizedExperiment'
autoplot(object, ..., type = c("heatmap", "link", "pcp", "boxplot", "scatterplot.matrix"), pheno.plot = FALSE,
                 main_to_pheno = 4.5, padding = 0.2, assay.id = 1)


## S4 method for signature 'VCF'
autoplot(object, ...,
              xlab, ylab, main,
              assay.id,
              type = c("default", "geno", "info", "fixed"),
              full.string = FALSE,
              ref.show = TRUE,
              genome.axis = TRUE,
              transpose = TRUE)

## S4 method for signature 'OrganismDb'
autoplot(object, which, ...,
                   xlab, ylab, main,
                   truncate.gaps = FALSE,
                   truncate.fun = NULL,
                   ratio = 0.0025,
                   geom = c("alignment"),
                   stat = c("identity", "reduce"),
                   columns = c("TXNAME", "SYMBOL", "TXID", "GENEID"),
                   names.expr = "SYMBOL",
                   label = TRUE,
                   label.color = "gray40")

## S4 method for signature 'VRanges'
autoplot(object, ...,which = NULL,
                             arrow = TRUE, indel.col = "gray30",
                             geom = NULL,
                             xlab, ylab, main)

## S4 method for signature 'TabixFile'
autoplot(object, which, ...)

Arguments

object	object to be plot.
columns	columns passed to method works for TxDb, EnsDb and OrganismDb.
label.color	when label turned on for gene model, this parameter controls label color.
arrow	arrow passed to geome_alignment function to control intron arrow attributes.
indel.col	indel colors.
ideogram	Weather to call plotIdeogram or not, default is FALSE, if TRUE, layout_karyogram will be called.
transpose	logical value, defaut TRUE, always make features from VCF as x, so we can use it to map to genomic position.
axis.text.angle	axis text angle.
axis.text.x	logical value indicates whether to show x axis and labels or not.
axis.text.y	logical value indicates whether to show y axis and labels or not.
hjust	horizontal just for axis text.
rownames.label	logical value indicates whether to show rownames of matrix as y label or not.
colnames.label	logical value indicates whether to show colnames of matrix as y label or not.
na.value	color for NA value.
rotate
pheno.plot	show pheno plot or not.
main_to_pheno	main matrix plot width to pheno plot width ratio.
padding	padding between plots.
assay.id	index for assay you are going to use.
geom	Geom to use (Single character for now). Please see section Geometry for details.
truncate.gaps	logical value indicate to truncate gaps or not.
truncate.fun	shrinkage function. Please see shrinkagefun in package biovizBase.
ratio	used in maxGap.
mode	Display mode for genomic features.
space.skip	space ratio between chromosome spaces in coordate genome.
coord	Coodinate system.
chr.weight	numeric vectors which sum to <1, the names of vectors has to be matched with seqnames in seqinfo, and you can only specify part of the seqnames, other lengths of chromosomes will be assined proportionally to their seqlengths, for example, you could specify chr1 to be 0.5, so the chr1 will take half of the space and other chromosomes squeezed to take left of the space.
legend	A logical value indicates whether to show legend or not. Default is TRUE
which	A GRanges object to subset the result, usually passed to the ScanBamParam function. For autoplot,EnsDb, which can in addition also be an object extending AnnotationFilter, an AnnotationFilterList combining such objects or a formula representing a filter expression. See examples below or documentation of AnnotationFilter for more details.
show.coverage	A logical value indicates whether to show coverage or not. This is used for geom "mismatch.summary".
resize.extra	A numeric value used to add buffer to intervals to compute stepping levels on.
bsgenome	A BSgenome object. Only need for geom "mismatch.summary".
xlab	x label.
ylab	y label.
label	logic value, default TRUE. To show label by the side of features.
facetByRow	A logical value, default is TRUE ,facet RleList by row. If FALSE, facet by column.
type	For Rle/RleList, "raw" plot everything, so be careful, that would be pretty slow if you have too much data. For "viewMins", "viewMaxs", "viewMeans", "viewSums", require extra arguments to slice the object. so users need to at least provide lower, more details and control please refer the the manual of slice function in IRanges. For "viewMins", "viewMaxs", we use viewWhichMin and viewWhichMax to get x scale, for "viewMeans", "viewSums", we use window midpoint as x. For ExpreesionSet, ploting types.
layout	Layout including linear, circular and karyogram. for GenomicRangesList, it only supports circular layout.
method	method used for parsing coverage from bam files. 'estimate' use fast esitmated method and 'raw' use relatively slow parsing method.
test.method	test method
...	Extra parameters. Usually are those parameters used in autoplot to control aesthetics or geometries.
main	title.
stat	statistical transformation.
indName	When coerce GRangesList to GRanges, names created for each group.
coverage.col	coverage stroke color.
coverage.fill	coverage fill color.
group.selfish	Passed to addStepping, control whether to show each group as unique level or not. If set to FALSE, if two groups are not overlapped with each other, they will probably be layout in the same level to save space.
names.expr	names expression used for creating labels. For EnsDb objects either "tx_id", "gene_name" or "gene_id".
binwidth	width of the bins.
nbin	number of bins.
genome.axis	logical value, if TRUE, whenever possible, try to parse genomic postition for each column(e.g. RangedSummarizedExperiment), show column as exatcly the genomic position instead of showing them side by side and indexed from 1.
full.string	logical value. If TRUE, show full string of indels in plot for VCF.
ref.show	logical value. If TRUE, show REF in VCF at bottom track.
chr	characters indicates the seqnames to be subseted.

Value

A ggplot object, so you can use common features from ggplot2 package to manipulate the plot.

Introduction

autoplot is redefined as generic s4 method inside this package, user could use autoplot in the way they are familiar with, and we are also setting limitation inside this package, like

• scales X scales is always genomic coordinates in most cases, x could be specified as start/end/midpoint when it's special geoms for interval data like point/line

• colors Try to use default color scheme defined in biovizBase package as possible as it can

Geometry

We have developed new geom for different objects, some of them may require extra parameters you need to provide. Some of the geom are more like geom + stat in ggplot2 package. e.g. "coverage.line" and "coverage.polygon".We simply combine them together, but in the future, we plan to make it more general.

This package is designed for only biological data, especially genomic data if users want to explore the data in a more flexible way, you could simply coerce the GRanges to a data.frame, then just use formal autoplot function in ggplot2, or autoplot generic for data.frame.

Some objects share the same geom so we introduce all the geom together in this section

full

Showing all the intervals as stepped rectangle, colored by strand automatically.

For TxDb or EnsDb objects, showing full model.

segment

Showing all the intervals as stepped segments, colored by strand automatically.

For object BSgenome, show nucleotides as colored segment.

For Rle/RleList, show histogram-like segments.

line

Showing interval as line, the interval data could also be just single position when start = end, x is one of start/end/midpoint, y value is unquoted name in elementMetadata column names. y value is required.

point

Showing interval as point, the interval data could also be just single position when start = end, x is one of start/end/midpoint, y value is unquoted name in elementMetadata column names. y value is required.

For object BSgenome, show nucleotides as colored point.

coverage.line

Coverage showing as lines for interval data.

coverage.polygon

Coverage showing as polygon for interval data.

splice

Splicing summary. The size and width of the line and rectangle should represent the counts in each model. Need to provide model.

single

For TxDb or EnsDb objects, showing single(reduced) model only.

tx

For TxDb or EnsDb objects, showing transcirpts isoforms.

mismatch.summary

Showing color coded mismatched stacked bar to indicate the proportion of mismatching at each position, the reference is set to gray.

text

For object BSgenome, show nucleotides as colored text.

rectangle

For object BSgenome, show nucleotides as colored rectangle.

Faceting

Faceting in ggbio package is a little differnt from ggplot2 in several ways

• The faceted column could only be seqnames or regions on the genome. So we limited the formula passing to facet argument, e.g something \~ seqnames, is accepted formula, you can change "something" to variable name in the elementMetadata. But you can not change the second part.

• Sometime, we need to view different regions, so we also have a facet_gr argument which accept a GRanges. If this is provided, it will override the default seqnames and use provided region to facet the graphics, this might be useful for different gene centric views.

Author(s)

Tengfei Yin

Examples

set.seed(1)
N <- 1000
library(GenomicRanges)
gr <- GRanges(seqnames =
              sample(c("chr1", "chr2", "chr3"),
                     size = N, replace = TRUE),
              IRanges(
                      start = sample(1:300, size = N, replace = TRUE),
                      width = sample(70:75, size = N,replace = TRUE)),
              strand = sample(c("+", "-", "*"), size = N,
                replace = TRUE),
              value = rnorm(N, 10, 3), score = rnorm(N, 100, 30),
              sample = sample(c("Normal", "Tumor"),
                size = N, replace = TRUE),
              pair = sample(letters, size = N,
                replace = TRUE))

idx <- sample(1:length(gr), size = 50)


###################################################
### code chunk number 3: default
###################################################
autoplot(gr[idx])


###################################################
### code chunk number 4: bar-default-pre
###################################################
set.seed(123)
gr.b <- GRanges(seqnames = "chr1", IRanges(start = seq(1, 100, by = 10),
                  width = sample(4:9, size = 10, replace = TRUE)),
                score = rnorm(10, 10, 3), value = runif(10, 1, 100))
gr.b2 <- GRanges(seqnames = "chr2", IRanges(start = seq(1, 100, by = 10),
                  width = sample(4:9, size = 10, replace = TRUE)),
                score = rnorm(10, 10, 3), value = runif(10, 1, 100))
gr.b <- c(gr.b, gr.b2)
head(gr.b)


###################################################
### code chunk number 5: bar-default
###################################################
p1 <- autoplot(gr.b, geom = "bar")
## use value to fill the bar
p2 <- autoplot(gr.b, geom = "bar", aes(fill = value))
tracks(default = p1, fill = p2)


###################################################
### code chunk number 6: autoplot.Rnw:236-237
###################################################
autoplot(gr[idx], geom = "arch", aes(color = value), facets = sample ~ seqnames)


###################################################
### code chunk number 7: gr-group
###################################################
gra <- GRanges("chr1", IRanges(c(1,7,20), end = c(4,9,30)), group = c("a", "a", "b"))
## if you desn't specify group, then group based on stepping levels, and gaps are computed without
## considering extra group method
p1 <- autoplot(gra, aes(fill = group), geom = "alignment")
## when use group method, gaps only computed for grouped intervals.
## default is group.selfish = TRUE, each group keep one row.
## in this way, group labels could be shown as y axis.
p2 <- autoplot(gra, aes(fill = group, group = group), geom = "alignment")
## group.selfish = FALSE, save space
p3 <- autoplot(gra, aes(fill = group, group = group), geom = "alignment", group.selfish = FALSE)
tracks('non-group' = p1,'group.selfish = TRUE' = p2 , 'group.selfish = FALSE' = p3)


###################################################
### code chunk number 8: gr-facet-strand
###################################################
autoplot(gr, stat = "coverage", geom = "area",
         facets = strand ~ seqnames, aes(fill = strand))


###################################################
### code chunk number 9: gr-autoplot-circle
###################################################
autoplot(gr[idx], layout = 'circle')


###################################################
### code chunk number 10: gr-circle
###################################################
seqlengths(gr) <- c(400, 500, 700)
values(gr)$to.gr <- gr[sample(1:length(gr), size = length(gr))]
idx <- sample(1:length(gr), size = 50)
gr <- gr[idx]
ggplot() + layout_circle(gr, geom = "ideo", fill = "gray70", radius = 7, trackWidth = 3) +
  layout_circle(gr, geom = "bar", radius = 10, trackWidth = 4,
                aes(fill = score, y = score)) +
  layout_circle(gr, geom = "point", color = "red", radius = 14,
                trackWidth = 3, grid = TRUE, aes(y = score)) +
  layout_circle(gr, geom = "link", linked.to = "to.gr", radius = 6, trackWidth = 1)


###################################################
### code chunk number 11: seqinfo-src
###################################################
data(hg19Ideogram, package = "biovizBase")
sq <- seqinfo(hg19Ideogram)
sq


###################################################
### code chunk number 12: seqinfo
###################################################
autoplot(sq[paste0("chr", c(1:22, "X"))])


###################################################
### code chunk number 13: ir-load
###################################################
set.seed(1)
N <- 100
ir <-  IRanges(start = sample(1:300, size = N, replace = TRUE),
               width = sample(70:75, size = N,replace = TRUE))
## add meta data
df <- DataFrame(value = rnorm(N, 10, 3), score = rnorm(N, 100, 30),
              sample = sample(c("Normal", "Tumor"),
                size = N, replace = TRUE),
              pair = sample(letters, size = N,
                replace = TRUE))
values(ir) <- df
ir


###################################################
### code chunk number 14: ir-exp
###################################################
p1 <- autoplot(ir)
p2 <- autoplot(ir, aes(fill = pair)) + theme(legend.position = "none")
p3 <- autoplot(ir, stat = "coverage", geom = "line", facets = sample ~. )
p4 <- autoplot(ir, stat = "reduce")
tracks(p1, p2, p3, p4)


###################################################
### code chunk number 15: grl-simul
###################################################
set.seed(1)
N <- 100
## ======================================================================
##  simmulated GRanges
## ======================================================================
gr <- GRanges(seqnames =
              sample(c("chr1", "chr2", "chr3"),
                     size = N, replace = TRUE),
              IRanges(
                      start = sample(1:300, size = N, replace = TRUE),
                      width = sample(30:40, size = N,replace = TRUE)),
              strand = sample(c("+", "-", "*"), size = N,
                replace = TRUE),
              value = rnorm(N, 10, 3), score = rnorm(N, 100, 30),
              sample = sample(c("Normal", "Tumor"),
                size = N, replace = TRUE),
              pair = sample(letters, size = N,
                replace = TRUE))


grl <- split(gr, values(gr)$pair)


###################################################
### code chunk number 16: grl-exp
###################################################
## default gap.geom is 'chevron'
p1 <- autoplot(grl, group.selfish = TRUE)
p2 <- autoplot(grl, group.selfish = TRUE, main.geom = "arrowrect", gap.geom = "segment")
tracks(p1, p2)


###################################################
### code chunk number 17: grl-name
###################################################
autoplot(grl, aes(fill = ..grl_name..))
## equal to
## autoplot(grl, aes(fill = grl_name))


###################################################
### code chunk number 18: rle-simul
###################################################
library(IRanges)
set.seed(1)
lambda <- c(rep(0.001, 4500), seq(0.001, 10, length = 500),
            seq(10, 0.001, length = 500))

## @knitr create
xVector <- rpois(1e4, lambda)
xRle <- Rle(xVector)
xRle


###################################################
### code chunk number 19: rle-bin
###################################################
p1 <- autoplot(xRle)
p2 <- autoplot(xRle, nbin = 80)
p3 <- autoplot(xRle, geom = "heatmap", nbin = 200)
tracks('nbin = 30' = p1, "nbin = 80" = p2, "nbin = 200(heatmap)" = p3)


###################################################
### code chunk number 20: rle-id
###################################################
p1 <- autoplot(xRle, stat = "identity")
p2 <- autoplot(xRle, stat = "identity", geom = "point", color = "red")
tracks('line' = p1, "point" = p2)


###################################################
### code chunk number 21: rle-slice
###################################################
p1 <- autoplot(xRle, type = "viewMaxs", stat = "slice", lower = 5)
p2 <- autoplot(xRle, type = "viewMaxs", stat = "slice", lower = 5, geom = "heatmap")
tracks('bar' = p1, "heatmap" = p2)


###################################################
### code chunk number 22: rlel-simul
###################################################
xRleList <- RleList(xRle, 2L * xRle)
xRleList


###################################################
### code chunk number 23: rlel-bin
###################################################
p1 <- autoplot(xRleList)
p2 <- autoplot(xRleList, nbin = 80)
p3 <- autoplot(xRleList, geom = "heatmap", nbin = 200)
tracks('nbin = 30' = p1, "nbin = 80" = p2, "nbin = 200(heatmap)" = p3)


###################################################
### code chunk number 24: rlel-id
###################################################
p1 <- autoplot(xRleList, stat = "identity")
p2 <- autoplot(xRleList, stat = "identity", geom = "point", color = "red")
tracks('line' = p1, "point" = p2)


###################################################
### code chunk number 25: rlel-slice
###################################################
p1 <- autoplot(xRleList, type = "viewMaxs", stat = "slice", lower = 5)
p2 <- autoplot(xRleList, type = "viewMaxs", stat = "slice", lower = 5, geom = "heatmap")
tracks('bar' = p1, "heatmap" = p2)


###################################################
### code chunk number 26: txdb
###################################################
library(TxDb.Hsapiens.UCSC.hg19.knownGene)
data(genesymbol, package = "biovizBase")
txdb <- TxDb.Hsapiens.UCSC.hg19.knownGene


###################################################
### code chunk number 27: txdb-visual
###################################################
p1 <- autoplot(txdb, which = genesymbol["ALDOA"], names.expr = "tx_name:::gene_id")
p2 <- autoplot(txdb, which = genesymbol["ALDOA"], stat = "reduce", color = "brown",
               fill = "brown")
tracks(full = p1, reduce = p2, heights = c(5, 1)) + ylab("")


###################################################
### EnsDb
###################################################
## Fetching gene models from an EnsDb object.
library(EnsDb.Hsapiens.v75)
ensdb <- EnsDb.Hsapiens.v75
## We use a GenenameFilter to specifically retrieve all transcripts for that gene.
p1 <- autoplot(ensdb, which = GeneNameFilter("ALDOA"), names.expr = "gene_name")
## Instead of providing the GenenameFilter, we can also use filter expressions
p2 <- autoplot(ensdb, which = ~ genename == "ALDOA", stat = "reduce",
               color = "brown", fill = "brown")
tracks(full = p1, reduce = p2, heights = c(5, 1)) + ylab("")

## Alternatively, we can specify a GRangesFilter and display all genes
## that are (partially) overlapping with that genomic region:
gr <- GRanges(seqnames=16, IRanges(30768000, 30770000), strand="+")
autoplot(ensdb, GRangesFilter(gr, "any"), names.expr="gene_name")
## Just submitting the GRanges object also works.
autoplot(ensdb, gr, names.expr="gene_name")

## Or genes encoded on both strands.
gr <- GRanges(seqnames = 16, IRanges(30768000, 30770000), strand = "*")
autoplot(ensdb, GRangesFilter(gr), names.expr="gene_name")

## Also, we can spefify directly the gene ids and plot all transcripts of these
## genes (not only those overlapping with the region)
autoplot(ensdb, GeneIdFilter(c("ENSG00000196118", "ENSG00000156873")))

###################################################
### code chunk number 28: ga-load
###################################################
library(GenomicAlignments)
data("genesymbol", package = "biovizBase")
bamfile <- system.file("extdata", "SRR027894subRBM17.bam",
                       package="biovizBase")
which <- keepStandardChromosomes(genesymbol["RBM17"])
## need to set use.names = TRUE
ga <- readGAlignments(bamfile,
                      param = ScanBamParam(which = which),
                      use.names = TRUE)


###################################################
### code chunk number 29: ga-exp
###################################################
p1 <- autoplot(ga)
p2 <- autoplot(ga, geom = "rect")
p3 <- autoplot(ga, geom = "line", stat = "coverage")
tracks(default = p1, rect = p2, coverage = p3)


###################################################
### code chunk number 30: bf-load (eval = FALSE)
###################################################
## library(Rsamtools)
## bamfile <- "./wgEncodeCaltechRnaSeqK562R1x75dAlignsRep1V2.bam"
## bf <- BamFile(bamfile)


###################################################
### code chunk number 31: bf-est-cov (eval = FALSE)
###################################################
## autoplot(bamfile)
## autoplot(bamfile, which = c("chr1", "chr2"))
## autoplot(bf)
## autoplot(bf, which = c("chr1", "chr2"))
##
## data(genesymbol, package = "biovizBase")
## autoplot(bamfile,  method = "raw", which = genesymbol["ALDOA"])
##
## library(BSgenome.Hsapiens.UCSC.hg19)
## autoplot(bf, stat = "mismatch", which = genesymbol["ALDOA"], bsgenome = Hsapiens)


###################################################
### code chunk number 32: char-bam (eval = FALSE)
###################################################
## bamfile <- "./wgEncodeCaltechRnaSeqK562R1x75dAlignsRep1V2.bam"
## autoplot(bamfile)


###################################################
### code chunk number 33: char-gr
###################################################
library(rtracklayer)
test_path <- system.file("tests", package = "rtracklayer")
test_bed <- file.path(test_path, "test.bed")
autoplot(test_bed, aes(fill = name))


###################################################
###  matrix
###################################################
volcano <- volcano[20:70, 20:60] - 150
autoplot(volcano)
autoplot(volcano, xlab = "xlab", main = "main", ylab = "ylab")
## special scale theme for 0-centered values
autoplot(volcano, geom = "raster")+scale_fill_fold_change()

## when a matrix has colnames and rownames label them by default
colnames(volcano) <- sort(sample(1:300, size = ncol(volcano), replace = FALSE))
autoplot(volcano)+scale_fill_fold_change()

rownames(volcano) <- letters[sample(1:24, size = nrow(volcano), replace = TRUE)]
autoplot(volcano)

## even with row/col names, you could also disable it and just use numeric index
autoplot(volcano, colnames.label = FALSE)
autoplot(volcano, rownames.label = FALSE, colnames.label = FALSE)

## don't want the axis has label??
autoplot(volcano, axis.text.x = FALSE)
autoplot(volcano, axis.text.y = FALSE)


# or totally remove axis
colnames(volcano) <- lapply(letters[sample(1:24, size = ncol(volcano),
replace = TRUE)],
function(x){
   paste(rep(x, 7), collapse = "")
})

## Oops, overlapped
autoplot(volcano)
## tweak with it.
autoplot(volcano, axis.text.angle =  -45, hjust = 0)

## when character is the value
x <- sample(c(letters[1:3], NA), size = 100, replace = TRUE)
mx <- matrix(x, nrow = 5)
autoplot(mx)
## tile gives you a white margin
rownames(mx) <- LETTERS[1:5]
autoplot(mx, color = "white")
colnames(mx) <- LETTERS[1:20]
autoplot(mx, color = "white")
autoplot(mx, color = "white", size = 2)
## weird in aes(), though works
## default tile is flexible
autoplot(mx, aes(width = 0.6, height = 0.6))
autoplot(mx, aes(width = 0.6, height = 0.6), na.value = "white")
autoplot(mx,  aes(width = 0.6, height = 0.6)) + theme_clear()

###################################################
### Views
###################################################
lambda <- c(rep(0.001, 4500), seq(0.001, 10, length = 500),
            seq(10, 0.001, length = 500))
xVector <- dnorm(1:5e3, mean = 1e3, sd = 200)
xRle <- Rle(xVector)
v1 <- Views(xRle, start = sample(.4e3:.6e3, size = 50, replace = FALSE), width =1000)
autoplot(v1)
names(v1) <- letters[sample(1:24, size = length(v1), replace = TRUE)]
autoplot(v1)
autoplot(v1, geom = "tile", aes(width = 0.5, height = 0.5))
autoplot(v1, geom = "line")
autoplot(v1, geom = "line", aes(color = row)) + theme(legend.position = "none")
autoplot(v1, geom = "line", facets = NULL)
autoplot(v1, geom = "line", facets = NULL, alpha  = 0.1)


###################################################
### ExpressionSet
###################################################
library(Biobase)
data(sample.ExpressionSet)
sample.ExpressionSet
set.seed(1)
## select 50 features
idx <- sample(seq_len(dim(sample.ExpressionSet)[1]), size = 50)
eset <- sample.ExpressionSet[idx,]
eset
autoplot(as.matrix(pData(eset)))

## default heatmap
p1 <- autoplot(eset)
p2 <- p1 + scale_fill_fold_change()
p2
autoplot(eset)
autoplot(eset, geom = "tile", color = "white", size = 2)
autoplot(eset, geom = "tile", aes(width = 0.6, height = 0.6))

autoplot(eset, pheno.plot = TRUE)
idx <- order(pData(eset)[,1])
eset2 <- eset[,idx]
autoplot(eset2, pheno.plot = TRUE)

## parallel coordainte plot
autoplot(eset, type = "pcp")

## boxplot
autoplot(eset, type = "boxplot")


## scatterplot.matrix
## slow, be carefull
##autoplot(eset[, 1:7], type = "scatterplot.matrix")

## mean-sd
autoplot(eset, type = "mean-sd")




###################################################
### RangedSummarizedExperiment
###################################################
library(SummarizedExperiment)
nrows <- 200; ncols <- 6
counts <- matrix(runif(nrows * ncols, 1, 1e4), nrows)
counts2 <- matrix(runif(nrows * ncols, 1, 1e4), nrows)
rowRanges <- GRanges(rep(c("chr1", "chr2"), c(50, 150)),
                   IRanges(floor(runif(200, 1e5, 1e6)), width=100),
                   strand=sample(c("+", "-"), 200, TRUE))
colData <- DataFrame(Treatment=rep(c("ChIP", "Input"), 3),
                     row.names=LETTERS[1:6])
sset <- SummarizedExperiment(assays=SimpleList(counts=counts,
                                               counts2 = counts2),
                             rowRanges=rowRanges, colData=colData)
autoplot(sset) + scale_fill_fold_change()
autoplot(sset, pheno.plot = TRUE)


###################################################
### pcp
###################################################
autoplot(sset, type = "pcp")


###################################################
### boxplot
###################################################
autoplot(sset, type = "boxplot")


###################################################
### scatterplot matrix
###################################################
##autoplot(sset, type = "scatterplot.matrix")


###################################################
### vcf
###################################################

## Not run: 
library(VariantAnnotation)
vcffile <- system.file("extdata", "chr22.vcf.gz", package="VariantAnnotation")
vcf <- readVcf(vcffile, "hg19")
## default use type 'geno'
## default use genome position
autoplot(vcf)
## or disable it
autoplot(vcf, genome.axis = FALSE)
## not transpose
autoplot(vcf, genome.axis = FALSE, transpose = FALSE, rownames.label = FALSE)
autoplot(vcf)
## use
autoplot(vcf, assay.id = "DS")
## equivalent to
autoplot(vcf, assay.id = 2)
## doesn't work when assay.id cannot find
autoplot(vcf, assay.id = "NO")
## use AF or first
autoplot(vcf, type = "info")
## geom bar
autoplot(vcf, type = "info", aes(y  = THETA))
autoplot(vcf, type = "info", aes(y  = THETA, fill = VT, color = VT))
autoplot(vcf, type = "fixed")
autoplot(vcf, type = "fixed", size = 10) + xlim(c(50310860, 50310890)) + ylim(0.75, 1.25)

p1 <- autoplot(vcf, type = "fixed") + xlim(50310860, 50310890)
p2 <- autoplot(vcf, type = "fixed", full.string = TRUE) + xlim(50310860, 50310890)
tracks("full.string = FALSE" = p1, "full.string = TRUE" = p2)+
  scale_y_continuous(breaks = NULL, limits = c(0, 3))
p3 <- autoplot(vcf, type = "fixed", ref.show = FALSE) + xlim(50310860, 50310890) +
    scale_y_continuous(breaks = NULL, limits = c(0, 2))
p3


## End(Not run)


###################################################
### code chunk number 56: bs-v
###################################################
library(BSgenome.Hsapiens.UCSC.hg19)
data(genesymbol, package = "biovizBase")
p1 <- autoplot(Hsapiens, which = resize(genesymbol["ALDOA"], width = 50))
p2 <- autoplot(Hsapiens, which = resize(genesymbol["ALDOA"], width = 50), geom = "rect")
tracks(text = p1, rect = p2)


###################################################
### code chunk number 57: sessionInfo
###################################################
sessionInfo()

lawremi/ggbio documentation built on Nov. 1, 2023, 2:40 p.m.