Skip to contents

Predict copy number aberrations using WisecondorX

rwisecondorx_predict.Rd

Native R implementation of the WisecondorX predict pipeline. Takes a binned sample and a reference (from rwisecondorx_newref()) and detects copy number aberrations using within-sample normalization, PCA correction, and circular binary segmentation (CBS).

Usage

rwisecondorx_predict(
  sample,
  reference,
  sample_binsize = NULL,
  outprefix = NULL,
  minrefbins = 150L,
  maskrepeats = 5L,
  alpha = 1e-04,
  zscore = 5,
  beta = NULL,
  blacklist = NULL,
  gender = NULL,
  seed = NULL,
  parallel = TRUE,
  cpus = 1L
)

Arguments

sample

Named list of integer vectors keyed by chromosome ("1"–"24"), as returned by bam_convert().

reference

A WisecondorXReference object from rwisecondorx_newref(), or a list with equivalent structure.

sample_binsize

Integer; bin size of the input sample. Used for rescaling to the reference bin size.

outprefix

Character; path prefix for output files. BED and statistics files will be written as <outprefix>_bins.bed, <outprefix>_segments.bed, <outprefix>_aberrations.bed, <outprefix>_statistics.txt. If NULL (default), no files are written and results are returned as a list.

minrefbins

Integer; minimum reference bins per target. Bins with fewer references are zeroed out. Default 150L.

maskrepeats

Integer; number of iterative distance-masking cycles. Default 5L.

alpha

Numeric; CBS breakpoint p-value threshold. Default 1e-4.

zscore

Numeric; Z-score cutoff for aberration calling. Default 5.

beta

Optional numeric; if given, ratio-based cutoff is used instead of Z-score. Should approximate purity (0, 1]. Default NULL.

blacklist

Optional character; path to a headerless BED file of regions to mask.

gender

Optional character; force gender ("F" or "M").

seed

Optional integer; RNG seed for CBS reproducibility.

parallel

Logical; use ParDNAcopy::parSegment() for CBS when available. Default TRUE. Falls back to DNAcopy::segment() with a message if ParDNAcopy is not installed.

cpus

Integer; number of threads for parallel CBS (parSegment) and any other OpenMP-accelerated steps. Default 1L.

Value

A list with class "WisecondorXPrediction" containing:

results_r

List of per-chromosome log2-ratio vectors.

results_z

List of per-chromosome Z-score vectors.

results_w

List of per-chromosome weight vectors.

results_c

Data frame of CBS segments (chr, start, end, zscore, ratio).

aberrations

Data frame of called aberrations.

statistics

Data frame of per-chromosome statistics.

gender

Predicted (or forced) gender.

n_reads

Total read count.

binsize

Reference bin size.

Details

CBS is performed using the DNAcopy Bioconductor package (or optionally ParDNAcopy for parallel segmentation). Both must be available in Suggests.

See also

rwisecondorx_newref(), scale_sample()