Pasha (version 0.99.21)

multiread_CSEMDispatch: Multiread scoring CSEM dispatch

Description

This script aims to dispatch the scoring of multi-read aligned reads according to the CSEM algorithm developped by Chung et al. (see "Discovering Transcription Factor Binding Sites of Genomes with Multi-Read Analysis of ChIP-seq Data" (2011) PLoS Computational Biology).

Usage

multiread_CSEMDispatch(alignedFile,  outputFolder,  referenceFile, window_size=101,  iteration_number=200,  incrArtefactThrEvery=NA,  verbosity=0)

Arguments

alignedFile

An atomic character string. The full path to the file containing the reads aligned by bowtie with the --concise option.

outputFolder

An atomic character string. The path to the folder where the file output by the script must be stored.

referenceFile

An atomic character string. Either a full path to a reference file (see details for format specification), or the ID of one reference included in the package (see details for available ones).

window_size

A positive integer. The size of the window used by the algorithm (see algorithm details). Default value is 101.

iteration_number

A positive integer. The number of iteration executed by the algorithm (see algorithm details). Default value is 200.

incrArtefactThrEvery

A complex parameter (see details). A numeric value or NA. A strictly positive numeric value activate the option that allow to remove the 'artifacts', defining a threshold to consider piles like 'artifacts' as 'number of reads in the experiment devided by incrArtefactThrEvery'. A NA will ignore the eventual artifactual piles.

verbosity

An integer. The verbose level : 0 = no message, 1 = trace level

Value

A tab separated value text file formated as below:

• Column 1 : Chromosome name
• Column 2 : Strand
• Column 3 : Position
• Column 4 : Score

Details

The script consider the reads that have been aligned in several location by bowtie (multi-reads). At each read, it assign a score determined by the CSEM algorithm (Chung et al. "Discovering Transcription Factor Binding Sites of Genomes with Multi-Read Analysis of ChIP-seq Data" (2011) PLoS Computational Biology). The script output a tab separated value text file formated as below:

• Column 1 : Chromosome name
• Column 2 : Strand
• Column 3 : Position
• Column 4 : Score

The output file is named according to the name of the input file suffixed with "_csemDispatch.txt".

If the parameter 'incrArtefactThrEvery' is set to a strictly positive value, the input file is first passed to a script removing the 'artefacts'. This script looks at the reads aligned at the exact same position. If the number of such reads exceed a limit, these reads are considered as due to experiment artifact. In that case, only one read is kept and the rest is removed. The limit defining artefact is controled by the parameter 'incrArtefactThrEvery' such as:

limit = Total number of reads / incrArtefactThrEvery

See Also

processPipeline multiread_RemoveArtifact multiread_UniformDispatch

Examples

# Define input aligned file
my_aligned_file <- system.file("extdata",
                               "embededDataTest_MultiSignal.bow",
                               package="Pasha")

# Define the output folder
my_output_folder <- tempdir()

# Define the genome reference file
genome_reference_file <- system.file("resources",
                                     "mm9.ref",
                                     package="Pasha")

# Launch the script
multiread_CSEMDispatch(my_aligned_file, 
                       my_output_folder, 
                       genome_reference_file,
                       incrArtefactThrEvery=7000000, 
                       verbosity=1)