extractUpstreamSeqs: Extract sequences upstream of a set of genes or transcripts
In jmacdon/GenomicFeatures: Conveniently import and query gene models

Description Usage Arguments Value Note Author(s) See Also Examples

extractUpstreamSeqs is a generic function for extracting sequences upstream of a supplied set of genes or transcripts.

extractUpstreamSeqs(x, genes, width=1000, ...)

## Dispatch is on the 2nd argument!

## S4 method for signature 'GenomicRanges'
extractUpstreamSeqs(x, genes, width=1000)

## S4 method for signature 'TxDb'
extractUpstreamSeqs(x, genes, width=1000, exclude.seqlevels=NULL)

`x`	An object containing the chromosome sequences from which to extract the upstream sequences. It can be a BSgenome, TwoBitFile, or FaFile object, or any genome sequence container. More formally, `x` must be an object for which `seqinfo` and `getSeq` are defined.
`genes`	An object containing the locations (i.e. chromosome name, start, end, and strand) of the genes or transcripts with respect to the reference genome. Only GenomicRanges and TxDb objects are supported at the moment. If the latter, the gene locations are obtained by calling the `genes` function on the TxDb object internally.
`width`	How many bases to extract upstream of each TSS (transcription start site).
`...`	Additional arguments, for use in specific methods.
`exclude.seqlevels`	A character vector containing the chromosome names (a.k.a. sequence levels) to exclude when the genes are obtained from a TxDb object.

A DNAStringSet object containing one upstream sequence per gene (or per transcript if genes is a GenomicRanges object containing transcript ranges).

More precisely, if genes is a GenomicRanges object, the returned object is parallel to it, that is, the i-th element in the returned object is the upstream sequence corresponding to the i-th gene (or transcript) in genes. Also the names on the GenomicRanges object are propagated to the returned object.

If genes is a TxDb object, the names on the returned object are the gene IDs found in the TxDb object. To see the type of gene IDs (i.e. Entrez gene ID or Ensembl gene ID or ...), you can display genes with show(genes).

In addition, the returned object has the following metadata columns (accessible with mcols) that provide some information about the gene (or transcript) corresponding to each upstream sequence:

gene_seqnames: the chromosome name of the gene (or transcript);
gene_strand: the strand of the gene (or transcript);
gene_TSS: the transcription start site of the gene (or transcript).

IMPORTANT: Always make sure to use a TxDb package (or TxDb object) that contains a gene model compatible with the genome sequence container x, that is, a gene model based on the exact same reference genome as x.

See http://bioconductor.org/packages/release/BiocViews.html#___TxDb for the list of TxDb packages available in the current release of Bioconductor. Note that you can make your own custom TxDb object from various annotation resources by using one of the makeTxDbFrom*() functions listed in the "See also" section below.

Hervé Pagès

makeTxDbFromUCSC, makeTxDbFromBiomart, and makeTxDbFromEnsembl, for making a TxDb object from online resources.
makeTxDbFromGRanges and makeTxDbFromGFF for making a TxDb object from a GRanges object, or from a GFF or GTF file.
The available.genomes function in the BSgenome package for checking avaibility of BSgenome data packages (and installing the desired one).
The BSgenome, TwoBitFile, and FaFile classes, defined and documented in the BSgenome, rtracklayer, and Rsamtools packages, respectively.
The TxDb class.
The genes function for extracting gene ranges from a TxDb object.
The GenomicRanges class defined and documented in the GenomicRanges package.
The DNAStringSet class defined and documented in the Biostrings package.
The seqinfo getter defined and documented in the GenomeInfoDb package.
The getSeq function for extracting subsequences from a sequence container.

## Load a genome:
library(BSgenome.Dmelanogaster.UCSC.dm3)
genome <- BSgenome.Dmelanogaster.UCSC.dm3
genome

## Use a TxDb object:
library(TxDb.Dmelanogaster.UCSC.dm3.ensGene)
txdb <- TxDb.Dmelanogaster.UCSC.dm3.ensGene
txdb  # contains Ensembl gene IDs

## Because the chrU and chrUextra sequences are made of concatenated
## scaffolds (see http://genome.ucsc.edu/cgi-bin/hgGateway?db=dm3),
## extracting the upstream sequences for genes located on these
## scaffolds is not reliable. So we exclude them:
exclude <- c("chrU", "chrUextra")
up1000seqs <- extractUpstreamSeqs(genome, txdb, width=1000,
                                  exclude.seqlevels=exclude)
up1000seqs  # the names are Ensembl gene IDs
mcols(up1000seqs)

## Upstream sequences for genes close to the chromosome bounds can be
## shorter than 1000 (note that this does not happen for circular
## chromosomes like chrM):
table(width(up1000seqs))
mcols(up1000seqs)[width(up1000seqs) != 1000, ]