R: Determine the optimal pooled threshold for unidirectional tag...

tuneTagClustering {CAGEfightR}

R Documentation

Determine the optimal pooled threshold for unidirectional tag clustering.

Description

This function counts the number of Tag Clusters (TCs) for an series of small incremental pooled cutoffs

Usage

tuneTagClustering(object, ...)

## S4 method for signature 'GRanges'
tuneTagClustering(
  object,
  steps = 10L,
  mergeDist = 20L,
  searchMethod = "minUnique",
  maxExponent = 1
)

## S4 method for signature 'RangedSummarizedExperiment'
tuneTagClustering(object, ...)

## S4 method for signature 'GPos'
tuneTagClustering(object, ...)

Arguments

`object`	GenomicRanges or RangedSummarizedExperiment: Pooled CTSS.
`...`	additional arguments passed to methods.
`steps`	integer: Number of thresholds to analyze (in addition to treshold=0).
`mergeDist`	integer: Merge TCs within this distance.
`searchMethod`	character: For advanced user only, see details.
`maxExponent`	numeric: The maximal threshold to analyse is obtained as min(score)*2^maxExponent (only used if searchMethod='exponential').

Value

data.frame with two columns: threshold and nTCs (number of Tag Clusters)

Examples

## Not run: 
data(exampleCTSSs)

# Calculate pooledTPM, using supplied number of total tags
exampleCTSSs <- calcTPM(exampleCTSSs,
                        inputAssay='counts',
                        outputAssay='TPM',
                        totalTags='totalTags')
exampleCTSSs <- calcPooled(exampleCTSSs, inputAssay='TPM')

# Set backend
library(BiocParallel)
register(SerialParam())

# Find optimal slice-threshold for reduce distance of 20:
tuneTagClustering(object=exampleCTSSs)

## End(Not run)