Scott Zuyderduyn >
Bio-SAGE-Comparison-1.00 >
Bio::SAGE::Comparison

Module Version: 1.00
Bio::SAGE::Comparison - Compares data from serial analysis of gene expression (SAGE) libraries.

use Bio::SAGE::Comparison; $sage = Bio::SAGE::Comparison->new();

This module provides several tools for comparing data generated from serial analysis of gene expression (SAGE) libraries.

**BACKGROUND**

Serial analysis of gene expression (SAGE) is a molecular technique for generating a near-global snapshot of a cell population’s transcriptome. Briefly, the technique extracts short sequences at defined positions of transcribed mRNA. These short sequences are then paired to form ditags. The ditags are concatamerized to form long sequences that are then cloned. The cloned DNA is then sequenced. Bioinformatic techniques are then employed to determine the original short tag sequences, and to derive their progenitor mRNA. The number of times a particular tag is observed can be used to quantitate the amount of a particular transcript. The original technique was described by Velculescu et al. (1995) and utilized an ~14bp sequence tag. A modified protocol was introduced by Saha et al. (2002) that produced ~21bp tags.

**PURPOSE**

This module facilitates the comparison of SAGE libraries. Specifically:

1. Calculations for determining the statistical significance of expression differences. 2. Dynamically convert longer-tag libraries to a shorter type for comparison (e.g. comparing a LongSAGE vs. a regular SAGE library).

Both regular SAGE (14mer tag) and LongSAGE (21mer tag) are supported by this module.

Statistical significance in library comparisons is calculated using the method described by Audic and Claverie (1997). Code was generated by directly porting the authors' original C source.

**REFERENCES**

Velculescu V, Zhang L, Vogelstein B, Kinzler KW. (1995) Serial analysis of gene expression. Science. 270:484-487. Saha S, Sparks AB, Rago C, Akmaev V, Wang CJ, Vogelstein B, Kinzler KW, Velculescu V. (2002) Using the transcriptome to annotate the genome. Nat. Biotechnol. 20:508-512. Audic S, Claverie JM. (1997) The significance of digital gene expression profiles. Genome Res. 7:986-995.

Follow the usual steps for installing any Perl module:

perl Makefile.PL make test make install

None.

1.00 2004.05.24 - Initial release.

**Settings**

$DEBUGPrints debugging output if value if >= 1.

Constructor for a new Bio::SAGE::Comparison object.

**Arguments**

None.

**Usage**

my $sage = Bio::SAGE::Comparison->new();

Determines the statistical significance of the difference in tag count (expression) between two libraries. This function uses the method described by Audic and Claverie (1997). This method can be called on an instantiated object, as well as statically.

**Arguments**

*$x,$y*

The number of tags in the x- and y-axis libraries, respectively.

*$Nx,$Ny*

The total number of tags in the x- and y-axis libraries, respectively.

*$signedValue* (optional)

A boolean value (>=1 is FALSE). If this flag is set to TRUE, downregulated comparisons will return a both p-value and either +1, -1, or 0 to indicate up/down/same-regulation (i.e. -1 if the expression ratio of tags in the x-axis library(s) is greater than that of the y-axis library(s)). This flag is FALSE by default.

**Returns**

The p-value for the observation. A lower number is more significant. Typically, observations with p <= 0.05 are considered statistically significant. If $signedValue is set to TRUE, the function also returns a 0, -1 or +1 to indicate same/down/up-regulation.

**Usage**

# the function is static, so it can be accessed directly my $p = Bio::SAGE::Comparison::calculate_significance( 3, 10, 50000, 60000 ); # or: my ( $p, $sign ) = Bio::SAGE::Comparison::calculate_significance( 3, 10, 50000, 60000, 1 ); if( $p <= 0.05 ) { if( $sign == +1 ) { print "Significantly upregulated.\n"; } if( $sign == -1 ) { print "Significantly downregulated.\n"; } if( $sign == 0 ) { die( "Same expression should never be significant!" ); } }

Takes a Perl handle to SAGE library data and prepares a tag data hash (format: [tag]<whitespace>[count]).

**Arguments**

*$handle*

A Perl handle (ie. filehandle, STDIN, etc.) that can be used to read in SAGE library data.

**Returns**

A hashref containing tag sequences as keys and the number of times that tag was observed as the value.

**Usage**

my $sage = Bio::SAGE::Comparison->new(); my %data = %{$sage->load_library( *STDIN )}; # print data in descending order of tag count map { print join( "\t", $_, $data{$_} ) . "\n"; } sort { $data{$b} <=> {$a} } keys %data;

Adds a library to this object.

**Arguments**

*$label*

A unique label for this library data. This label can then be used to refer to the data.

*\%tagData*

A hashref containing the library data. The keys are tag sequences, the values are tag counts. A properly formatted hash can be created using load_library.

**Usage**

my $sage = Bio::SAGE::Comparison->new(); my %data = %{$sage->load_library( *STDIN )}; # or alternatively: my %data = ( 'AACGACTGTT' => 100, 'CAGATACAAG' => 23, 'AGATAAAGAC' => 45 ); $sage->add_library( 'MYLIB', \%data );

Gets the labels that identify the currently added libraries.

**Arguments**

None.

**Returns**

An array of library labels.

**Usage**

my $sage = Bio::SAGE::Comparison->new(); my %data = %{$sage->load_library( *STDIN )}; $sage->add_library( 'MYLIB', \%data ); print "LIBRARY_NAMES: " . join( ", " , $sage->get_library_labels() ) . "\n";

Gets the total number of tags (the sum of all observed tag counts for a library(s)).

**Arguments**

*$label*

This can be: a) a string denoting the library label, b) a reference to a string denoting the library label, or c) an arrayref containing several library labels that are pooled to calculate total size.

**Returns**

The total number of observed tags in the library(s) specified.

**Usage**

my $sage = Bio::SAGE::Comparison->new(); my %data = %{$sage->load_library( *STDIN )}; $sage->add_library( 'MYLIB', \%data ); print "LIBRARY_SIZE: " . $sage->get_library_size( 'MYLIB' ) . "\n";

Gets the number of discrete tag sequences present in the specified library(s).

**Arguments**

*$label*

This can be: a) a string denoting the library label, b) a reference to a string denoting the library label, or c) an arrayref containing several library labels that are pooled to calculate the number of tag sequences.

**Returns**

The number of different tags in the library(s) specified.

**Usage**

my $sage = Bio::SAGE::Comparison->new(); my %data = %{$sage->load_library( *STDIN )}; $sage->add_library( 'MYLIB', \%data ); print "TAG_SEQUENCES: " . $sage->get_number_tag_sequences( 'MYLIB' ) . "\n";

Creates a comparison between two libraries. This method returns a hash reference containing the library data and a p-value for determining statistical signifance.

If the libraries contain tags of different sizes (i.e. comparing a regular SAGE library vs. a LongSAGE library) the data is converted to the shortest tag length of the libraries specified prior to comparison.

**Arguments**

*$x_axis_libraries,$y_axis_libraries*

Library labels for the x- and y-axis, respectively. This can be: a) a string denoting the library label, b) a reference to a string denoting the library label, or c) an arrayref containing several library labels that are pooled in the comparison.

**Returns**

A hashref is returned where the keys are tag sequences, and the values are hashrefs with the keys 'x' (tags in x-axis library(s)), 'y' (tags in y-axis library(s)), 'reg' (0,-1,+1 denoting same/down/up-regulation), and 'p' (statistical significance). i.e. $HASH{$tag}->{'p'} = 0.05;

**Usage**

my $sage = Bio::SAGE::Comparison->new(); open( LIB1, "lib1.tags.txt" ); $sage->add_library( 'LIB1', $sage->load_library( *LIB1 ) ); close( LIB1 ); open( LIB2, "lib2.tags.txt" ); $sage->add_library( 'LIB2', $sage->load_library( *LIB2 ) ); close( LIB2 ); my %comparison = %{$sage->get_library_comparison( 'LIB1', 'LIB2' ); # or alternatively: my %comparison = %{$sage->get_library_comparison( ['LIB1'], ['LIB2'] ); # print results in ascending order of p-value (more significant first) foreach my $tag ( sort { $comparison{$a}->{'p'} <=> $comparison{$b}->{'p'} } keys %comparison ) { print join( "\t", $tag, map { $comparison{$tag}->{$_} } ( 'x','y','reg','p' ) ) . "\n"; }

Prints a report based on the supplied comparison result hash.

**Arguments**

*\%comparison*

A properly formed hashref containing library comparison results. This structure can be created with get_library_comparison. A sample output looks like: Tag N(x) N(y) reg p AGATCAAGAT 3388 50 -1 0 GATAACACTT 11481 186 -1 0 TATAACACCA 4 607 1 4.1136713480649e-306 ... etc.

**Usage**

my $sage = Bio::SAGE::Comparison->new(); # load library data # ... $sage->print_library_comparison( $sage->get_library_comparison( 'LIB1','LIB2' ) );

Copyright(c)2004 Scott Zuyderduyn <scottz@bccrc.ca>. All rights reserved.

This program is free software; you can redistribute it and/or modify it under the same terms as Perl itself.

Scott Zuyderduyn <scottz@bccrc.ca> BC Cancer Research Centre

1.00

perl(1).

Nothing yet.

syntax highlighting: