mapCorrect: Correct read depth for mapability

In chrisamiller/copyCat: Find regions of genomic copy number loss and gain from short reads

Description

takes a readDepth object containing read counts and updates the rd counts, accounting for repetitive and unmappable genomic regions. Also adds a "map" column containing that bin's percentage mapability

Usage

  mapCorrect(rdo, outlierPercentage=0.01, minMapability=0.60,
resolution=0.01, skipCorrection=FALSE)

Arguments

rdo	a readDepth object created with new("rdObject") and filled with read counts using the getReadDepth() command
minMapability	the minimum amount of mapability required. If the mapability of a region is below this level, the read depth is set to NA.
outlierPercentage	Removes the highest and lowest N This prevents overfitting near the extremes, to which loess is susceptible. Default value is 0.01
skipCorrection	Set this to true to just remove low-mapability sites without doing loess correction on any of the readcounts. Helps prevent spurious hits in low-coverage regions (2 reads vs 1 read should not be interpreted as a CN doubling).
resolution	The amount of division between bins, expressed in percent. A value of 1 would bin regions with mapability percentage of [1,2,...,99,100].

Details

Assumes that appropriate annotation files are in place.

Value

returns a readDepth object with the read depth values altered to remove bias due to mapability

See Also

'getReadDepth'

Examples

  ##create a rdObject, set bin params
  #  x <- new("rdObject")

  ##fill the bins with reads
  #  x <- getReadDepth(x)

  ##do mapping correction on the read counts
  #  x.mapCorr = mapCorrect(x)

chrisamiller/copyCat documentation built on July 20, 2021, 12:59 a.m.