filtering: Filtering FraserDataSets

In c-mertes/FRASER: Find RAre Splicing Events in RNA-Seq Data

Filtering FraserDataSets

Description

This method can be used to filter out introns that are not reliably detected and to remove introns with no variablity between samples.

Usage

filterVariability(object, ...)

filterExpressionAndVariability(
  object,
  minExpressionInOneSample = 20,
  quantile = 0.75,
  quantileMinExpression = 10,
  minDeltaPsi = 0,
  filter = TRUE,
  delayed = ifelse(ncol(object) <= 300, FALSE, TRUE),
  filterOnJaccard = TRUE,
  BPPARAM = bpparam()
)

## S4 method for signature 'FraserDataSet'
filterExpression(
  object,
  minExpressionInOneSample = 20,
  quantile = 0.75,
  quantileMinExpression = 10,
  filter = TRUE,
  delayed = ifelse(ncol(object) <= 300, FALSE, TRUE),
  filterOnJaccard = TRUE,
  BPPARAM = bpparam()
)

## S4 method for signature 'FraserDataSet'
filterVariability(
  object,
  minDeltaPsi = 0,
  filter = TRUE,
  delayed = ifelse(ncol(object) <= 300, FALSE, TRUE),
  filterOnJaccard = TRUE,
  BPPARAM = bpparam()
)

Arguments

object	A FraserDataSet object
...	Further parameters passed on to Rsubread::featureCounts.
minExpressionInOneSample	The minimal read count in at least one sample that is required for an intron to pass the filter.
quantile	Defines which quantile should be considered for the filter.
quantileMinExpression	The minimum read count an intron needs to have at the specified quantile to pass the filter.
minDeltaPsi	Only introns for which the maximal difference in the psi value of a sample to the mean psi of the intron is larger than this value pass the filter.
filter	If TRUE, a subsetted fds containing only the introns that passed all filters is returned. If FALSE, no subsetting is done and the information of whether an intron passed the filters is only stored in the mcols.
delayed	If FALSE, count matrices will be loaded into memory, otherwise the function works on the delayedMatrix representations. The default value depends on the number of samples in the fds-object.
filterOnJaccard	If TRUE, the Intron Jaccard Metric is used to define express introns during fitlering. Otherwise, the psi5, psi3 and theta metrics are used (default: TRUE).
BPPARAM	the BiocParallel parameters for the parallelization

Value

A FraserDataSet with information about which junctions passed the filters. If filter=TRUE, the filtered FraserDataSet is returned.

Functions

• filterExpressionAndVariability(): This functions filters out both introns with low read support and introns that are not variable across samples.

• filterExpression(FraserDataSet): This function filters out introns and corresponding splice sites that have low read support in all samples.

• filterVariability(FraserDataSet): This function filters out introns and corresponding splice sites that have low read support in all samples.

Examples

fds <- createTestFraserDataSet()
fds <- filterExpressionAndVariability(fds, minDeltaPsi=0.1, filter=FALSE)
mcols(fds, type="jaccard")[, c(
        "maxCount", "passedExpression", "maxDJaccard", "passedVariability")]

plotFilterExpression(fds)
plotFilterVariability(fds)     

c-mertes/FRASER documentation built on June 14, 2024, 7:49 p.m.