NAME
wiggle2gff3.pl - Converts UCSC WIG format files into gff3 files
SYNOPSIS
wiggle2gff3.pl [options] WIG_FILE > load_data.gff3
Converts UCSC WIG format files into gff3 files suitable for loading into GBrowse databases. This is used for high-density quantitative data such as CNV, SNP and expression arrays.
DESCRIPTION
Use this converter when you have dense quantitative data to display using the xyplot, density, or heatmap glyphs, and too many data items (thousands) to load into GBrowse. It creates one or more space- efficient binary files containing the quantitative data, as well as a small GFF3 file that can be loaded into Chado or other GBrowse databases.
Typical usage is as follows:
% wiggle2gff3.pl --method=microarray_oligo my_data.wig > my_data.gff3
Options
The following options are accepted:
--method=<method name> Set the method for the GFF3 lines representingeach quantitative data point in the track.The default is "microarray_oligo."--source=<source> Set the source field for the GFF3 file. The default isnone.--gff3 Create a GFF3-format file (the default)--featurefile Create a "featurefile" format file -- this is thesimplified format used for GBrowse uploads. Thisoption is incompatible with the --gff3 option.--sample If true, then very large files (>5 MB) will be sampledto obtain minimum, maximum and standard deviation; otherwisethe entire file will be scanned to obtain these statistics.This will process the files faster but may miss outliervalues.--path=<path> Specify the directory in which to place the binary wigglefiles. The default is the current temporary directory(/tmp or whatever is appropriate for your operating system).--base=<path> Same as "--path".--trackname specify the trackname base for the wigfile creation--help This documentation.
This script will accept a variety of option styles, including abbreviated options ("--meth=foo"), single character options ("-m foo"), and other common variants.
Binary wiggle files
The binary "wiggle" files created by this utility are readable using the Bio::Graphics::Wiggle module. The quantitative data is scaled to the range of 1-255 (losing lots of precision, but still more than enough for data visualization), and stored in a packed format in which each file corresponds to the length of a single chromosome or contig.
Once created, the binary files should not be moved or renamed, unless you are careful to make corresponding changes to the pathnames given by the "wigfile" attribute in the GFF3 file feature lines. You should also be careful about using the cp command to copy the binary files; they are formatted with "holes" in such a way that missing data does not take up any space on disk. If you cp them, the holes will fill up with zeroes and the space savings will be lost. Better to use the "tar" command with its --sparse option to move the files from one place to another.
Example WIG File
This example is from <http://genome.ucsc.edu/goldenPath/help/wiggle.html>:
# filename: example.wig## 300 base wide bar graph, autoScale is on by default == graphing# limits will dynamically change to always show full range of data# in viewing window, priority = 20 positions this as the second graph# Note, zero-relative, half-open coordinate system in use for bed formattrack type=wiggle_0 name="Bed Format" description="BED format" \visibility=full color=200,100,0 altColor=0,100,200 priority=20chr19 59302000 59302300 -1.0chr19 59302300 59302600 -0.75chr19 59302600 59302900 -0.50chr19 59302900 59303200 -0.25chr19 59303200 59303500 0.0chr19 59303500 59303800 0.25chr19 59303800 59304100 0.50chr19 59304100 59304400 0.75chr19 59304400 59304700 1.00# 150 base wide bar graph at arbitrarily spaced positions,# threshold line drawn at y=11.76# autoScale off viewing range set to [0:25]# priority = 10 positions this as the first graph# Note, one-relative coordinate system in use for this formattrack type=wiggle_0 name="variableStep" description="variableStep format" \visibility=full autoScale=off viewLimits=0.0:25.0 color=255,200,0 \yLineMark=11.76 yLineOnOff=on priority=10variableStep chrom=chr19 span=15059304701 10.059304901 12.559305401 15.059305601 17.559305901 20.059306081 17.559306301 15.059306691 12.559307871 10.0# 200 base wide points graph at every 300 bases, 50 pixel high graph# autoScale off and viewing range set to [0:1000]# priority = 30 positions this as the third graph# Note, one-relative coordinate system in use for this formattrack type=wiggle_0 name="fixedStep" description="fixed step" visibility=full \autoScale=off viewLimits=0:1000 color=0,200,100 maxHeightPixels=100:50:20 \graphType=points priority=30fixedStep chrom=chr19 start=59307401 step=300 span=2001000900800700600500400300200100
You can convert this into a loadable GFF3 file with the following command:
wiggle2gff3.pl --meth=example --so=example --path=/var/gbrowse/db example.wig \> example.gff3
The output will look like this:
##gff-version 3chr19 example example 59302001 59304700 . . . Name=Bed Format;wigfile=/var/gbrowse/db/track001.chr19.1199828298.wigchr19 example example 59304701 59308020 . . . Name=variableStep;wigfile=/var/gbrowse/db/track002.chr19.1199828298.wigchr19 example example 59307401 59310400 . . . Name=fixedStep;wigfile=/var/gbrowse/db/track003.chr19.1199828298.wig
PROBLEMS
This script has trouble with wig files from very fragmented genomes (>100K scaffolds). In this case, you may wish to run split_wig.pl, which splits the original wig file into a series of smaller files with a maximum of 900 scaffolds each. It then runs wiggle2gff3.pl for each subfile and stores the results in separate folders.
SEE ALSO
Bio::DB::GFF, bp_bulk_load_gff.pl, bp_fast_load_gff.pl, bp_load_gff.pl, bp_seqfeature_load.pl
AUTHOR
Lincoln Stein <[email protected]>.
Copyright (c) 2008 Cold Spring Harbor Laboratory
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. Refer to LICENSE for the full license text. See DISCLAIMER.txt for disclaimers of warranty.