pileupXenograft: test harness for simultaneous mouse/human MT variant calling...

In trichelab/MTseeker: Bioconductor Tools for Human Mitochondrial Variant Analysis

Description

Note that the default pileup params here do NOT distinguish strands in PDXs.

Usage

pileupXenograft(
  bam,
  sbp = NULL,
  pup = NULL,
  callIndels = FALSE,
  minVAF = 0.03,
  minDepth = 20,
  verbose = FALSE,
  ...
)

Arguments

bam	the BAM or object with a @bam slot
sbp	optional ScanBamParam object (autogenerated if missing)
pup	optional PileupParam object (autogenerated if missing)
callIndels	call indels? (This can reduce compute and RAM if FALSE.)
minVAF	minimum VAF to accept (default is 0.03, usual NuMT cutoff)
minDepth	minimum read depth to accept (default is 20 for sanity)
verbose	be verbose, as for profiling? (default is FALSE / quiet)
...	additional args to pass on to the pileup() function

Details

FIXME: add an iterator to create (human, mouse) MVRangesLists when running pileupXenograft on thousands of individual cells (as in PDX studies).

Value

       a list of MVRanges (one human, one mouse)

Examples

  BAMdir <- system.file("extdata", "BAMs", package="MTseekerData")
  addPath <- function(x) file.path(BAMdir, x)
  pdxBAMs <- addPath(list.files(path=BAMdir, pattern="^PDX.*.bam$"))
  names(pdxBAMs) <- sapply(strsplit(basename(pdxBAMs), "_"), `[`, 1)
  mvrls <- lapply(pdxBAMs, pileupXenograft, verbose=TRUE)
  vapply(mvrls, names, character(2))

trichelab/MTseeker documentation built on March 8, 2021, 6:20 p.m.