NAME

bin_genomic_data.pl

A script to bin genomic data into windows.

SYNOPSIS

bin_genomic_data.pl --data <filename> --in <filename> --method [method]

bin_genomic_data.pl --data <filename> --new --win <integer> --db <db_name>
    --out <filename> --method [method]

 Options:
 --data <filename>
 --in <filename>
 --new
 --method [count | mean | median |sum]
 --out <filename>
 --db <database_name>
 --win <integer>
 --step <integer>
 --paired
 --span
 --midpoint
 --shift <integer>
 --interpolate
 --log
 --gff
 --gz
 --version
 --help

OPTIONS

The command line flags and descriptions:

--data <filename>

Specify the source genomic data file containing the datapoints to be binned. Supported files include .gff, .bed, , .wig, .sgr, and .bam files. Text files may be gzipped.

--in <filename>

Specify the filename of an existing data table of genomic bins. It should be in the tim data file format with metadata and have the feature 'genome'. It may be generated with this program or 'get_datasets.pl'. Required unless --new is used.

--new

Indicate that a new data file should be generated. Required unless --in is used.

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

Indicate which method should be used to collect the binned data. Acceptable values include:

- count         Count the number of unique features in the bin.
- mean          Take the mean score of features in the bin.
- median        Take the median score of features in the bin.
- sum           Sum the scores of features in the bin

--out <filename>

Specify the output file name. Required if generating a new binned data file, otherwise if an input file was specified it will be overwritten.

--db <database_name>

Specify a database. This is used only to retrieve chromosome information to generate genomic bins. Required if --new is used.

--win <integer>

Specify the genomic bin window size in bp. Required if --new is used.

--step <integer>

Specify the step size for moving the window when generating bins. The default is to equal the window size.

--paired

Indicate that the data bam file consists of paired-end alignments rather than single-end alignments.

--span

Assign feature value (or count) at each genomic bin across the feature. This is dependent on the metadata step value when the genomic bins were generated.

--midpoint

Assign the the feature value (or count) at the feature's midpoint rather than the default feature's start point.

--shift <integer>

For stranded single-end alignments from a bam data file, the start position may be shifted in the 3' direction by the indicated number of bp. This is to account for ChIP-Seq data where the peak of tag counts is offset from the actual center of the sequenced fragments. Use a shift value of 1/2 the mean fragment length of the sequencing library.

--interpolate

Optionally interpolate missing bin values from flanking bins.

--log

Flag to indicate that source data is log2, and to calculate accordingly and report as log2 data.

--gff

Write a new gff output file instead of a normal tim data file.

--gz

Compress the output file through gzip.

--version

Print the version number.

--help

Display the POD documentation.

DESCRIPTION

This program will convert genomic data into genomic binned data. The genomic bins are generated from the indicated database as segments of the genome of the specified window size. The step size may be optionally specified, or defaults to the window size. The source data may be collected from multiple sources, including GFF, BED, BAM, WIG, or SGR files. The data may be collected in one of two ways.

First, the program can enumerate (or count) features which overlap with a genomic bin. Sequence tag (BAM file) alignments (both single- and paired-end), BED, and GFF features may be counted. The feature start point, or optionally the midpoint, is used to record the feature. Alternatively, the counts may be enumerated across the entire span of the feature when using small (single bp) bins.

Second, the program can statistically combine the scores (values) of features that overlap each genomic bin. Feature scores may be collected from GFF, BED, WIG, or SGR files. Either the mean, median, or sum value can be recorded for the genomic bin.

Genomic bins lacking a score may be interpolated from neighboring bins. Up to four consecutive non-value bins may interpolated from neighboring bins that contain values.

The program reads/writes a data file compatable with the 'get_datasets.pl' program. It can optionally write a new GFF file based on the genomic binned data.

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.