NearestGenomicNeighbor: Find closest gene for genomic intervals within the same TAD.

In ATpoint/dailytRumpet: Unrelated R functions for daily use in bioinformatics that do not fit into any other repository

Description

Function accepts genomic intervals of regions (e.g. from ChIP-seq), genes and TADs and will return a data frame with the region coordiates, the closest gene, the distance to it and the intersecting TAD. Input format is GRanges. Genes.gr must have a slot <gene_id> with gene names.

Usage

NearestGenomicNeighbor(
  Genes.gr,
  Regions.gr,
  TADs.gr,
  cTADs.gr = NULL,
  max.distance = NULL,
  Find.Genes = "closest",
  cores = detectCores() - 1,
  Sort.output = FALSE
)

Arguments

Genes.gr	GRanges with gene coordinates, will be collapse to start position depending on strand
Regions.gr	GRanges with genomic regions to find nearby genes, will be collapsed to interval center
TADs.gr	GRanges with TAD intervals
cTADs.gr	an optional set of GRanges with the complement of TADs and the genome, see details
max.distance	numerical, a cutoff for maximum istance of region to gene
cores	workers for mclapply
Sort.output	logical, whether to sort output by chromosome and start position
Find.Genes=c("closest", "all")	whether to return only the closest or all genes within the same TAD, limited by max.distance

Details

Currently, in case that one regions has > 1 associated gene with identical distance (ties), it will return all tied genes. Also, when cTADs.gr is provided (the complement of genome and TADs), then these "cTADs" will be included into the GRanges of the normal TADs and treated equally. This has the advantage that regions outside of TADs are not categorically excluded but the disadvantage that the assignment is even more unreliable than the nearest-matching we do anyway, this is why this parameter is optonal.

Author(s)

Alexander Toenges

ATpoint/dailytRumpet documentation built on April 28, 2020, 1:47 a.m.