makePvalueAnnotation: Initialize a PvalueAnnotation

Description

This function initializes a PvalueAnnotation using a gene BED file and optional BED files corresponding to interval datasets. This is a necessary first step in order to establish for each gene the gene promoter, body and associated intervals.

Usage

makePvalueAnnotation(data, other_data = NULL, other_tss_distance = 10000,  promoter_upstream_distance = 1000, promoter_downstream_distance = 1000,  strand_col = NULL, gene_name_col = NULL)

Arguments

data

A required gene annotation BED file like that obtained from the UCSC Table Browser. At a minimum BED files must have the first three columns as (chromosome, start, end). Additional required columns should correspond to the strand and gene name.The gene name needs to match the gene format desired for the interaction network. Duplicated gene names and associated gene annotations are removed.

other_data

A list of BED files corresponding to each additional interval file to be associated with genes. The function will use the other_tss_distance variable and the gene transcription start site (TSS) to find for each gene all intervals within [tss-other_tss_distance, tss+other_tss_distance].

other_tss_distance

A vector specifying for each element of otherdata a distance from the gene TSS to consider that interval as related to a gene. If the length of other_tss_distance does not match the length of the otherdata list, then the first value is used for all datasets in the otherdata list. DEFAULTS to 10,000 base pairs.

promoter_upstream_distance

A numeric specifying how far upstream from the gene TSS is considered part of the gene promoter. DEFAULTS to 1,000 base pairs.

promoter_downstream_distance

A numeric specifying how far downstream from the gene TSS is considered part of the gene promoter. Gene bodies subtract the promoter_downstream region. DEFAULTS to 1,000 base pairs.

strand_col

A numeric specifying the column of the gene BED file (data) corresponding to the gene strand. If this is not provided, the function will attempt to determine the strand.

gene_name_col

A numeric specifying the column of the gene BED file (data) corresponding to the gene name.

Value

An S4 object of class PvalueAnnotation containing slots for an annotation (GRangesList), an expression set, modifications (GRangesList), and a PvalueObject.

Details

The required only input file is the gene annotation BED file that should have (as all BED files) the chromosome, start and end in columns 1,2 and 3, respectively. Also, there should be a column for gene name and gene strand. The user needs to determine distance from the gene transcription start site that will define the gene promoter. The gene body will then be calculated as the non-promoter overlapping sequence. If optional BED files are given as otherdata (e.g. transcription factor binding sites, histone modification peaks), then the user will also decide a distance from the gene TSS to associate each BED interval with a gene. For a particular BED file, each genes may have more than one interval that falls within the desired range around a TSS. Unique gene names are required and the function will automatically remove duplicated genes. We recommend deciding on an interaction network first and then loading a gene annotation BED file with the same names. This will likely necessitate allowing the function to pick one annotation of a gene, or pre- processing using some criteria (e.g. longest transcript).

Examples

Run this code

## Note: Commented out below. See vignette for more detailed usage information##

## Load genome bed file ##
data(hg19_genes_bed)

## Create a PvalueAnnotation with defaults for promoter size##
test_annotation <- makePvalueAnnotation(data=hg19_genes, gene_name_col=5)