draw.controls: Select control genomic regions for enrichment analysis

In jmonlong/PopSV: Population-based detection of structural variants from Read-Depth signal

Description

Find random genomic regions with the same properties as a set of input regions. The properties are defined using annotation tracks. For example, we can simulate random regions that overlap genes as much as the input regions. It can be used to perform enrichment analysis while controling for specific properties.

Usage

draw.controls(cnv.gr, feat.grl, nb.class = 100, nb.cores = 3,
  redo.duplicates = TRUE, seed.nb.max = 1e+05, min.nb.gr = NULL,
  chr.prefix = "", dist.gr = NULL)

Arguments

cnv.gr	the input regions
feat.grl	a list of the GRanges defining the features to fit. E.g. gene annotation, centromere, etc
nb.class	the number of size class to speed up the computation. Default is 100. Inf will ensure exactly the same overlap proportion but might take some time if the size distribution is diverse.
nb.cores	the number of cores to use. Default is 3.
redo.duplicates	should duplicate regions be reselected. Default is TRUE.
seed.nb.max	the maximum number of genomic position to use as seeds. Default is 1e5.
min.nb.gr	the minimum number of control regions. If NULL (default), as many controls as input are simulated.
chr.prefix	the chromosome name prefix. Default is "" (no prefix). Other value could be "chr" if chromosome are defined as 'chr1', 'chr2', etc.
dist.gr	a GRanges defining the feature for which we want to control the distance to. Default is NULL, i.e. no control.

Value

a GRanges object defining the control regions

Author(s)

Jean Monlong

jmonlong/PopSV documentation built on Sept. 15, 2019, 9:29 p.m.