NAME

bar2wig.pl

A script to convert bar files to wig files.

SYNOPSIS

bar2wig.pl [--options...] <filename>

Options:
--in <filename>
--out <filename> 
--method [mean|median|sum|max]
--log
--track
--bw
--barapp </path/to/Bar2Gr>
--bwapp </path/to/wigToBigWig>
--db <database>
--chromof </path/to/chromosomes>
--gz
--version
--help

OPTIONS

The command line flags and descriptions:

--in <filename>

Specify the input bar file or (more likely) a directory of bar files. The bar files may be zipped (*.bar.zip).

--out <filename>

Specify the output filename. By default it uses the base name of the input file or the input directory name. The output file will either have a .wig or .bw file extension.

--method [mean | median | sum | max]

Define the method used to combine multiple data values at a single position. Bar files can frequently have one or more values at a single position, usually due to either multiple identical oligo probes (microarray data) or multiple sequence tags aligning to the same position (next generation sequencing data). Typically, with FDR or microarray data, the mean or max value should be taken, while bar files representing sequence tag PointData should be summed.

--log

If multiple data values need to be combined at a single identical position, indicate whether the data is in log2 space or not. This affects the mathematics behind the combination method.

--track

Wig files typically include a track line at the beginning of the file which defines the appearance. However, conversion of a wig file to bigWig requires that the track line is absent. Do not include the track line when generating bigWig files manually. The default is to include for wig files (true), and exclude for bigWig files (false).

--bw

Flag to indicate that a binary bigWig file should be generated rather than a text wig file.

--barapp </path/to/Bar2Gr>

Specify the full path to David Nix's USeq or T2 application Bar2Gr (it is included in both software packages). By default it uses the path defined in the biotoolbox configuration file, biotoolbox.cfg.

--bwapp </path/to/wigToBigWig>

Specify the full path to Jim Kent's wigToBigWig conversion utility. By default it uses the path defined in the biotoolbox configuration file, biotoolbox.cfg. If it is not defined here or in the config file, then the system path is searched for the executable.

--db <database>

Specify the name of a Bio::DB::SeqFeature::Store annotation database or other indexed data file, e.g. Bam or bigWig file, from which chromosome length information may be obtained. For more information about using databases, see https://code.google.com/p/biotoolbox/wiki/WorkingWithDatabases. It is only required when generating a bigWig file.

--chromof </path/to/chromosomes>

Alternative to the --db argument, a pre-generated chromosome sizes text file may be specified. This text file should consist of two columns, delimited by whitespace, consisting of the chromosome name and size.

--gz

Specify whether (or not) the output wig file should be compressed with gzip. Default is false.

--version

Print the version number.

--help

Display this POD documentation.

DESCRIPTION

This program will convert binary .bar files to a conventional text-based wig file, with the option to further convert to a binary bigWig format.

The bar file is a Java serialized, floating-point, binary file used natively by David Nix's TiMAT2 (T2) and USeq software packages. It allows for effective visualization by the Integrated Genome Browser (IGB), but very few (any?) other genome browsers.

The wig file is a commonly used text-based file format for exchanging and representing dense genomic data (microarray and sequencing) that is read by nearly all genome browsers. In most genome browsers, the data values are downsampled to 8-bit precision (0-255), sufficient for visualization but limited for data analysis.

The bigWig format is an indexed, compressed, binary extension of the wig format. It maintains numeric precision, allows for rapid statistical summaries, and can be rapidly and randomly accessed either locally or remotely. It is supported by the UCSC and GBrowse genome browsers as well as biotoolbox scripts.

This program first uses David Nix's Bar2Gr application to convert the bar file to a very simplified text format, a .gr file. This is then further processed into one or more wig (or bigWig) files. Stranded data (denoted by _+_ and _-_ in the bar file names) is written to two stranded output files, appended with either '_f' or '_r' to the basename. The wig files are in variableStep format.

You may find the latest version of the USeq package, which contains the Bar2Gr Java application, at http://sourceforge.net/projects/useq/. Specify the path to the Bar2Gr file in the USeq/Apps directory. You may record this information in the BioToolBox configuration file biotoolbox.cfg.

Conversion from wig to bigWig requires Jim Kent's wigToBigWig utility or Lincoln Stein's Bio::DB::BigFile support.

AUTHOR

Timothy J. Parnell, PhD
Howard Hughes Medical Institute
Dept of Oncological Sciences
Huntsman Cancer Institute
University of Utah
Salt Lake City, UT, 84112

This package is free software; you can redistribute it and/or modify it under the terms of the GPL (either version 1, or at your option, any later version) or the Artistic License 2.0.