<<

NAME

Chum::Result::Interpretter::MatrixScannerInterpretter - Interpret result sets from scanning matrix sets derived from matrix databases (like Jaspar)

SYNOPSIS

DESCRIPTION

Interpret result sets from Chum::Analysis::Local::MatrixScanner and return a score. This score looks at the cumulative difference in two lists of predicted transcription factors using the desired TFBS::MatrixSet. The scores from each list are cumulated. The magnitude of the difference in scores is returned by the interpret function.

i.e. Reference predictions: Factor1 5 Factor2 6 Factor3 7

Variant predictions: Factor1 3 Factor2 8 Factor4 4

The returned score would be for Factor1 abs(5-3) = 2 Factor2 abs(6-8) = 2 Factor3 abs(7-0) = 7 Factor4 abs(0-4) = 4 returned score = 2+2+7+4= 15

For multiple variants, the maximum score is returned.

Example (scan all PWM's in TransFAC):

my $transfac_db = new Chum::DB::Transfac('transfac_dir' => "$transfac_directory"); $transfac_db->matrixset($transfac_db->_find_all_matrixset()); my $matrix_scanner = new Chum::Analysis::Local::MatrixScanner('matrix_db' => $transfac_db); my $resultset = $matrix_scanner->run_variation_analysis($stack); my $msi = new Chum::Result::Interpretter::MatrixScannerInterpretter($resultset); my $score = $msi->interpret();

CONTACT

Stephen Montgomery (smontgom@bcsgc.bc.ca)

METHODS

interpret

 Usage   : $interpretter->interpret();
        Calculate the difference score for a MatrixScanner analysis run between
        two sets of hits.  For an example, see Usage at the beginning of this module.
 Returns : integer, score
 Args    : 

interpret

 Usage   : $interpretter->interpret_with_modifiers($modifiers);
        Calculate the difference score for a MatrixScanner analysis run between
        two sets of hits.  For an example, see Usage at the beginning of this module.
        Results are modified by dividing using the modifier hashref.
 Returns : integer, score
 Args    : Arg1: Hashref (keys are binding site names, values are modifier scores which are divided against raw factor scores)

<<