gene: Annotate genomic positions with gene transcript information
In tobyjohnson/gtx: Genetics ToolboX

Description Usage Arguments Details Value Author(s) Examples

Annotates genomic positions (chromosome and position) with information about nearby transcripts from a UCSC genome browser table such as GENCODE.

gene.annotate(chr, pos, genetable, win.size = 10000)
gene.nearest(chr, pos, genetable)
gene.draw(chr, leftpos, rightpos,
          genetable, nodraw = NULL, 
          genesep = 10000, hlines.min = NULL, yhi = -1, ylo = -5,
          exony = 0.05, genecex = 1)

`chr`	A vector of chromosome numbers
`pos`	A vector of chromosomal base pair positions (one-based)
`genetable`	A data frame of a UCSC genome browser table
`win.size`	Size of window to include transcripts for
`leftpos`	Leftmost position to draw transcripts for
`rightpos`	Rightmost position to draw transcripts for
`nodraw`	A list of gene names not to draw
`genesep`	Desired space between genes drawn on same horizontal line
`hlines.min`	Minimum number of horizontal lines to use
`yhi`	Y coordinate of highest line
`ylo`	Y coordinate of lowest line
`exony`	Y coordinate height for exons
`genecex`	cex parameter for gene names

genetable could be the refgene or GENCODE table from UCSC. It must have columns “chrom”, “txStart”, “txEnd”, “name2”, and for gene.draw it must also have columns “exonStarts” and “exonEnds”. Note that all positions in genetable are assumed to be zero-based, but function argument pos is assumed to be one-based.

The algorithm for gene.nearest is peculiar and may change in future versions. If the queried position is inside one or more transcripts, their names are concatenated (comma separated). Otherwise, the names of the nearest transcripts to the left and right are concatendated (hyphen separated). This may produce curious output for transcript names that themselves contain a hyphen.

gene.draw draws representations of all genes in the specified region. For each gene (determined by unique values of “name2”), all transcripts are drawn in an overlapping style, with yellow blocks for exons and black lines for introns, and “name2” written below. Different genes are distributed over a number of lines, chosen adaptively to avoid horizontal overlap. In large or gene dense regions, this may result in overlap of the gene names with the vertical overlap.

Character string, or plot.

Toby Johnson Toby.x.Johnson@gsk.com

## Not run: 
## genome wide annotation 
genetable <- read.table(gzfile("path/to/UCSC-genome-tables/hg19.GENCODE14.gz"),
                          header = TRUE, comment.char = "", sep = "\t", as.is = TRUE)

## End(Not run)

data(gencode14.UGT1A1)
leftrightpos <- c(234568893, 234781945)
plot.new()
plot.window(leftrightpos, c(-5, 5))
abline(h = 0, col = "grey")
gene.draw(2, leftrightpos[1], leftrightpos[2], gencode14.UGT1A1, genesep = 5000)
axis(1, at = pretty(leftrightpos * 1e-6)*1e6, labels = pretty(leftrightpos * 1e-6))
title(xlab = "chr2 genomic position (Mb)")
axis(2, at = 0:5, las = 1)
box()

querypos <- c(234680000, 234685000)
points(querypos, rep(0, 2), pch = 23)
gene.nearest(c("chr2", "chr2"), querypos, gencode14.UGT1A1)

gene.annotate("chr2", querypos[1], gencode14.UGT1A1) # genes within 10kb
gene.annotate("chr2", querypos[1], gencode14.UGT1A1, win.size = 0) # genes within 0kb
gene.annotate(2, querypos[1], gencode14.UGT1A1, win.size = 0) # same