absoluteRanges {GenomicRanges}R Documentation

Transform genomic ranges into "absolute" ranges

Description

absoluteRanges transforms the genomic ranges in x into absolute ranges i.e. into ranges counted from the beginning of the virtual sequence obtained by concatenating all the sequences in the underlying genome (in the order reported by seqlevels(x)).

relativeRanges performs the reverse transformation.

NOTE: These functions only work on small genomes. See Details section below.

Usage

absoluteRanges(x)
relativeRanges(x, seqlengths)

## Related utility:
isSmallGenome(seqlengths)

Arguments

x

For absoluteRanges: a GenomicRanges object with ranges defined on a small genome (see Details section below).

For relativeRanges: a Ranges object.

seqlengths

An object holding sequence lengths. This can be a named integer (or numeric) vector with no duplicated names as returned by seqlengths(), or any object from which sequence lengths can be extracted with seqlengths().

For relativeRanges, seqlengths must describe a small genome (see Details section below).

Details

Because absoluteRanges returns the absolute ranges in an IRanges object, and because an IRanges object cannot hold ranges with an end > .Machine$integer.max (i.e. >= 2^31 on most platforms), absoluteRanges cannot be used if the size of the underlying genome (i.e. the total length of the sequences in it) is > .Machine$integer.max. Utility function isSmallGenome is provided as a mean for the user to check upfront whether the genome is small (i.e. its size is <= .Machine$integer.max) or not, and thus compatible with absoluteRanges or not.

relativeRanges applies the same restriction by looking at the seqlengths argument.

Value

An IRanges object for absoluteRanges.

A GRanges object for relativeRanges.

absoluteRanges and relativeRanges both return an object that is parallel to the input object (i.e. same length and names).

isSmallGenome returns TRUE if the total length of the underlying sequences is <= .Machine$integer.max (e.g. Fly genome), FALSE if not (e.g. Human genome), or NA if it cannot be computed (because some sequence lengths are NA).

Author(s)

H. Pagès

See Also

Examples

## ---------------------------------------------------------------------
## TOY EXAMPLE
## ---------------------------------------------------------------------

gr <- GRanges(Rle(c("chr2", "chr1", "chr3", "chr1"), 4:1),
              IRanges(1:10, width=5),
              seqinfo=Seqinfo(c("chr1", "chr2", "chr3"), c(100, 50, 20)))

ar <- absoluteRanges(gr)
ar

gr2 <- relativeRanges(ar, seqlengths(gr))
gr2

## Sanity check:
stopifnot(all(gr == gr2))

## ---------------------------------------------------------------------
## ON REAL DATA
## ---------------------------------------------------------------------

## With a "small" genome

library(TxDb.Dmelanogaster.UCSC.dm3.ensGene)
txdb <- TxDb.Dmelanogaster.UCSC.dm3.ensGene
ex <- exons(txdb)
ex

isSmallGenome(ex)

## Note that because isSmallGenome() can return NA (see Value section
## above), its result should always be wrapped inside isTRUE() when
## used in an if statement:
if (isTRUE(isSmallGenome(ex))) {
    ar <- absoluteRanges(ex)
    ar

    ex2 <- relativeRanges(ar, seqlengths(ex))
    ex2  # original strand is not restored

    ## Sanity check:
    strand(ex2) <- strand(ex)  # restore the strand
    stopifnot(all(ex == ex2))
}

## With a "big" genome (but we can reduce it)

library(TxDb.Hsapiens.UCSC.hg19.knownGene)
txdb <- TxDb.Hsapiens.UCSC.hg19.knownGene
ex <- exons(txdb)
isSmallGenome(ex)
## Not run: 
    absoluteRanges(ex)  # error!

## End(Not run)

## However, if we are only interested in some chromosomes, we might
## still be able to use absoluteRanges():
seqlevels(ex, pruning.mode="coarse") <- paste0("chr", 1:10)
isSmallGenome(ex)  # TRUE!
ar <- absoluteRanges(ex)
ex2 <- relativeRanges(ar, seqlengths(ex))

## Sanity check:
strand(ex2) <- strand(ex) 
stopifnot(all(ex == ex2))

[Package GenomicRanges version 1.30.1 Index]