humarray (version 1.0.0)

get.gene.annot: Get human gene names and locations from biomart

Description

Various R packages assist in downloading genomic information but often the input required is complex, or several lines of code are required to initiate, returning an object that might require some manipulation to be useful. This function simplifies the job considerably, not necessarily requiring any arguments. The object returned can be a standard data.frame or a bioconductor GRanges/RangedData object. The raw annotation file downloaded will be kept in the working directory so that subsequent calls to this function run very quickly, and also allow use offline.

Usage

get.gene.annot(dir = NULL, build = NULL, bioC = TRUE, duplicate.report = FALSE, one.to.one = FALSE, remap.extra = FALSE, discard.extra = TRUE, only.named = FALSE, ens.id = FALSE, refresh = FALSE, GRanges = TRUE, host.txt = "")

Arguments

dir

character, location to store file with the gene annotation. If NULL then getOption("save.annot.in.current")>=1 will result in this file being stored in the current directory, or if

build

string, currently 'hg18' or 'hg19' to specify which annotation version to use. Default is build-36/hg-18. Will also accept integers 36,37 as alternative arguments.

bioC

logical, whether to return the annotation as a ranged S4 object (GRanges or RangedData), or as a data.frame

duplicate.report

logical, whether to provide a report on the genes labels that are listed in more than 1 row - this is because some genes span ranges with substantial gaps within them

one.to.one

logical, as per above, some genes have duplicate entries, sometimes for simplicity you want just one range per gene, if this parameter is set TRUE, one range per gene is enforced, and only the widest range will be kept by default for each unique gene label

remap.extra

logical, whether to remap chromosome annotation for alternative builds and unconnected segments to the closest regular chromosome, e.g, mapping MHC mappings to chromosome 6

discard.extra

logical, similar to above, but if TRUE, then any non-standard chromosome genes will just be discarded

only.named

logical, biomart annotation contains some gene segments without names, if TRUE, then such will not be included in the returned object (note that this will happen also if one.to.one is TRUE)

ens.id

logical, whether to include the ensembl id in the dataframe

refresh

logical, if you already have the file in the current directory, this argument will let you re-download and re-generate this file, e.g, if the file is modified or corrupted this will make a new one without having to manually delete it

GRanges

logical, if TRUE and bioC is also TRUE, then returned object will be GRanges, otherwise it will be RangedData

host.txt

character, the argument to pass to biomaRt::useMart(). Default for build 36 is 'may2009.archive.ensembl.org', and for build 37, "feb2014.archive.ensembl.org" but for recent builds the recommended link is 'www.ensembl.org'

Value

Returns a data.frame, GRanges or RangedData object, depending on input parameters. Contained will be HGNC gene labels, chromosome and start and end positions, other information depends on specific parameters documented above

Examples

setwd(tempdir())
get.gene.annot()