getNearestFeature: Get nearest annotation boundary for a position range.

In hiAnnotator: Functions for annotating GRanges objects

Description

Given a query object, the function retrieves the nearest feature and its properties from a subject and then appends them as new columns within the query object. When used in genomic context, the function can be used to retrieve the nearest gene 5' or 3' end relative to genomic position of interest.

Usage

getNearestFeature(sites.rd, features.rd, colnam = NULL,
  side = "either", feature.colnam = NULL, dists.only = FALSE,
  parallel = FALSE, relativeTo = "subject")

Arguments

sites.rd	GRanges object to be used as the query.
features.rd	GRanges object to be used as the subject or the annotation table.
colnam	column name to be added to sites.rd for the newly calculated annotation...serves a core!
side	boundary of annotation to use to calculate the nearest distance. Options are '5p','3p', 'either'(default), or 'midpoint'.
feature.colnam	column name from features.rd to be used for retrieving the nearest feature name. By default this is NULL assuming that features.rd has a column that includes the word 'name' somewhere in it.
dists.only	flag to return distances only. If this is TRUE, then 'feature.colnam' is not required and only distance to the nearest feature will be returned. By default this is FALSE.
parallel	use parallel backend to perform calculation with foreach. Defaults to FALSE. If no parallel backend is registered, then a serial version of foreach is ran using registerDoSEQ.
relativeTo	calculate distance relative to query or subject. Default is 'subject'. This essentially means whether to use query or subject as the anchor point to get distance from!

Value

a GRanges object with new annotation columns appended at the end of sites.rd.

Note

• When side='midpoint', the distance to nearest feature is calculated by (start+stop)/2.

• If strand information doesn't exist, then everything is defaulted to '+' orientation (5' -> 3')

• If parallel=TRUE, then be sure to have a parallel backend registered before running the function. One can use any of the following libraries compatible with foreach: doMC, doSMP, doSNOW, doMPI, doParallel. For example: library(doMC); registerDoMC(2)

• When relativeTo="subject", the biological distance is relative to subject, meaning, the function reports the distance to query from subject (i.e. an integration site is upstream or downstream from a gene). When relativeTo="query", the distance is from the point of view of query or an integration site (i.e. gene is upstream or downstream from an integration site).

See Also

makeGRanges, getFeatureCounts, getSitesInFeature, get2NearestFeature.

Examples

# Convert a dataframe to GRanges object
data(sites)
alldata.rd <- makeGRanges(sites, soloStart = TRUE)

data(genes)
genes.rd <- makeGRanges(genes)

nearestGenes <- getNearestFeature(alldata.rd, genes.rd, "NearestGene")
nearestGenes
nearestGenes <- getNearestFeature(alldata.rd, genes.rd, "NearestGene",
side = "5p")
nearestGenes
## Not run: 
nearestGenes <- getNearestFeature(alldata.rd, genes.rd, "NearestGene",
side = "3p")
nearestGenes
nearestGenes <- getNearestFeature(alldata.rd, genes.rd, "NearestGene",
side = "midpoint")
## Parallel version of getNearestFeature
nearestGenes <- getNearestFeature(alldata.rd, genes.rd, "NearestGene",
parallel = TRUE)
nearestGenes

## End(Not run)

hiAnnotator documentation built on Nov. 8, 2020, 7:27 p.m.