Toolbar

Summary

Bio::SearchDist - A perl wrapper around Sean Eddy's histogram object

Package variables

Included modules

Inherit

Bio::Root::Root

Synopsis

  $dis = Bio::SearchDist->new();
  foreach $score ( @scores ) {
     $dis->add_score($score);
  }

  if( $dis->fit_evd() ) {
    foreach $score ( @scores ) {
      $evalue = $dis->evalue($score);
      print "Score $score had an evalue of $evalue\n";
    }
  } else {
    warn("Could not fit histogram to an EVD!");
  }

Description

The Bio::SearchDist object is a wrapper around Sean Eddy's excellent
histogram object. The histogram object can bascially take in a number
of scores which are sensibly distributed somewhere around 0 that come
from a supposed Extreme Value Distribution. Having add all the scores
from a database search via the add_score method you can then fit a
extreme value distribution using fit_evd(). Once fitted you can then
get out the evalue for each score (or a new score) using
evalue($score).
The fitting procedure is better described in Sean Eddy's own code
(available from http://hmmer.wustl.edu, or in the histogram.h header
file in Compile/SW). Bascially it fits a EVD via a maximum likelhood
method with pruning of the top end of the distribution so that real
positives are discarded in the fitting procedure. This comes from
an orginally idea of Richard Mott's and the likelhood fitting
is from a book by Lawless [should ref here].
The object relies on the fact that the scores are sensibly distributed
around about 0 and that integer bins are sensible for the
histogram. Scores based on bits are often ideal for this (bits based
scoring mechanisms is what this histogram object was originally
designed for).

Methods

Methods description

 Title   : _engine
 Usage   : $obj->_engine($newval)
 Function: underlyine bp_sw:: histogram engine
 Returns : value of _engine
 Args    : newvalue (optional)

 Title   : add_score
 Usage   : $dis->add_score(300);
 Function: Adds a single score to the distribution
 Returns : nothing
 Args    :

 Title   : evalue
 Usage   : $eval = $dis->evalue($score)
 Function: Returns the evalue of this score 
 Returns : float 
 Args    :

 Title   : fit_Gaussian
 Usage   :
 Function:
 Example :
 Returns : 
 Args    :

 Title   : fit_evd
 Usage   : $dis->fit_evd();
 Function: fits an evd to the current distribution
 Returns : 1 if it fits successfully, 0 if not
 Args    :

Methods code

BEGIN {

    eval {
	require Bio::Ext::Align;
    };
    if ( $@ ) {
print $@;
	print STDERR ("\nThe C-compiled engine for histogram object (Bio::Ext::Align) has not been installed.\n Please install the bioperl-ext package\n\n");
	exit(1);

}

sub _engine {

    my ($self,$value) = @_;
    if( defined $value) {
	$self->{'_engine'} = $value;
    }
    return $self->{'_engine'};
}


## End of Package

1;
__END__

}

sub add_score {

   my ($self,$score) = @_;
   my ($eng);
   $eng = $self->_engine();
   #$eng->AddToHistogram($score);
   $eng->add($score);

}

sub evalue {

   my ($self,$score) = @_;

   return $self->_engine()->evalue($score);

}

sub fit_Gaussian {

   my ($self,$high) = @_;

   if( ! defined $high ) {
       $high = 10000;
   }

   return $self->_engine()->fit_Gaussian($high);

}

sub fit_evd {

   my ($self,@args) = @_;

   return $self->_engine()->fit_EVD(10000,1);

}

sub new {

  my($class,@args) = @_;
  my $self = $class->SUPER::new(@args);
  my($min, $max, $lump) = 
	$self->_rearrange([qw(MIN MAX LUMP)], @args);

  if( !  $min ) {
    $min = -100;
  }

  if( !  $max ) {
    $max = +100;
  }

  if( ! $lump ) {
    $lump = 50;
  }

  $self->_engine(&Bio::Ext::Align::new_Histogram($min,$max,$lump));

  return $self;

}

General documentation

The original code this was based on comes from the histogram module as
part of the HMMer2 package. Look at http://hmmer.wustl.edu/
Its use in Bioperl is via the Compiled XS extension which is cared for
by Ewan Birney (birney@sanger.ac.uk). Please contact Ewan first about
the use of this module

User feedback is an integral part of the evolution of this and other
Bioperl modules. Send your comments and suggestions preferably to one
of the Bioperl mailing lists. Your participation is much appreciated.

  bioperl-l@bioperl.org          - General discussion
  http://bio.perl.org/MailList.html             - About the mailing lists

Report bugs to the Bioperl bug tracking system to help us keep track
the bugs and their resolution. Bug reports can be submitted via email
or the web:

  bioperl-bugs@bio.perl.org
  http://bugzilla.bioperl.org/

The rest of the documentation details each of the object
methods. Internal methods are usually preceded with a _