miscLocal: Detect local shape changes
In hyochoi/SCISSOR: Shape changes in selecting sample outliers in RNA-seq

miscLocal

R Documentation

Detect local shape changes

Description

This function discovers outlying subjects whose RNA-seq have "local" abnormal shapes and provides the most outlying window-direction for each outlier.

Usage

miscLocal(
  miscGlobalResult,
  pileupData,
  Ranges,
  siglev = 1e-04,
  cutoff = NULL,
  ADcutoff = 3,
  windowSize = 100,
  reducedReturn = TRUE
)

Arguments

`miscGlobalResult`	Result from miscGlobal
`pileupData`	raw coverage matrix, dataI from process_data
`siglev`	the significance level for a robust outlier detection. Default is 1e-5. IF cutoff is specified, siglev is not used.
`cutoff`	the cutoff value for outlying statistics. If NULL, the cutoff value is computed based on the specified siglev.
`windowSize`	a window length. Default is 100.
`reducedReturn`
`residualData`	a residual matrix from PCA, a data matrix subtracted by a low-rank matrix.
`exonset`	data annotation matrix, exonset from process_data