Bio::Tools::dpAlign(3) Perl extension to do pairwise dynamic programming sequence alignment

SYNOPSIS


use Bio::Tools::dpAlign;
use Bio::SeqIO;
use Bio::SimpleAlign;
use Bio::AlignIO;
use Bio::Matrix::IO;
$seq1 = Bio::SeqIO->new(-file => $ARGV[0], -format => 'fasta');
$seq2 = Bio::SeqIO->new(-file => $ARGV[1], -format => 'fasta');
# create a dpAlign object
# to do global alignment, specify DPALIGN_GLOBAL_MILLER_MYERS
# to do ends-free alignment, specify DPALIGN_ENDSFREE_MILLER_MYERS
$factory = new dpAlign(-match => 3,
-mismatch => -1,
-gap => 3,
-ext => 1,
-alg => Bio::Tools::dpAlign::DPALIGN_LOCAL_MILLER_MYERS);
# actually do the alignment
$out = $factory->pairwise_alignment($seq1->next_seq, $seq2->next_seq);
$alnout = Bio::AlignIO->new(-format => 'pfam', -fh => \*STDOUT);
$alnout->write_aln($out);
# To do protein alignment, set the sequence type to protein
# By default all protein alignments are using BLOSUM62 matrix
# the gap opening cost is 7 and gap extension is 1. These
# values are from ssearch. To use your own custom substitution
# matrix, you can create a Bio::Matrix::MatrixI object.
$parser = Bio::Matrix::IO->new(-format => 'scoring', -file => 'blosum50.mat');
$matrix = $parser->next_matrix;
$factory = Bio::Tools::dpAlign->new(-matrix => $matrix, -alg => Bio::Tools::dpAlign::DPALIGN_LOCAL_MILLERMYERS);
$seq1->alphabet('protein');
$seq2->alphabet('protein');
$out = $factory->pairwise_alignment($seq1->next_seq, $seq2->next_seq);
$alnout->write_aln($out);
# use the factory to make some output
$factory->align_and_show($seq1, $seq2, STDOUT);
# use Phil Green's algorithm to calculate the optimal local
# alignment score between two sequences quickly. It is very
# useful when you are searching a query sequence in a database
# of sequences. Since finding a alignment is more costly
# than just calculating scores, you can save time if you only
# align sequences that have a high alignment score.
# To use this feature, first you call the sequence_profile function
# to obtain the profile of the query sequence.
$profile = $factory->sequence_profile($query);
%scores = ();
# Then use a loop to run a database of sequences against the
# profile to obtain a table of alignment scores
$dbseq = Bio::SeqIO(-file => 'dbseq.fa', -format => 'fasta');
while (defined($seq = $dbseq->next_seq)) {
$scores{$seq->id} = $factory->pairwise_alignment_score($profile, $seq);
}

DESCRIPTION

Dynamic Programming approach is considered to be the most sensitive way to align two biological sequences. There are currently three major types of dynamic programming algorithms: Global Alignment, Local Alignment and Ends-free Alignment.

Global Alignment compares two sequences in their entirety. By inserting gaps in the two sequences, it aligns two sequences to minimize the edit distance as defined by the gap cost function and the substitution matrix. Global Alignment is generally applied to two sequences that are very similar in length and content.

Local Alignment instead attempts to find out the subsequences that has the minimal edit distance among all possible subsequences. It is good for sequences that has a stretch of subsequences that are similar to each other.

Ends-free Alignment is a special case of Global Alignment. There are no gap penalty imposed for the gaps that extended from the end points of two sequences. Therefore it will be a good application when you think one sequence is contained by the other or when you think two sequences overlap each other.

Dynamic Programming was first introduced by Needleman-Wunsch (1970) to globally align two sequences. The idea of local alignment was later introduced by Smith-Waterman (1981). Gotoh (1982) improved both algorithms by introducing auxillary arrays that reduced the time complexity of the algorithms to O(m*n). Miller-Myers (1988) exploits the divide-and-conquer idea introduced by Hirschberg (1975) to solve the affine gap cost dynamic programming using only linear space. At the time of this writing, it is accepted that Miller-Myers is the fastest single CPU implementation and using the least memory that is truly equivalent to original algorithm introduced by Needleman-Wunsch. According to Aaron Mackey, Phil Green's SWAT implementation introduced a heuristic that does not consider paths through the matrix where the score would be less than the gap opening penalty, yielding a 1.5-2X speedup on most comparisons. to skip the calculation of some cells. However, his approach is only good for calculating the minimum edit distance and find out the corresponding subsequences (aka search phase). Bill Pearson's popular dynamic programming alignment program SSEARCH uses Phil Green's algorithm to find the subsequences and then Miller-Myers's algorithm to find the actual alignment. (aka alignment phase)

The current implementation supports local alignment of either DNA sequences or protein sequences. It allows you to specify either the Miller-Myers Global Alignment (DPALIGN_GLOBAL_MILLER_MYERS) or Miller-Myers Local Alignment (DPALIGN_LOCAL_MILLER_MYERS). For DNA alignment, you can specify the scores for match, mismatch, gap opening cost and gap extension cost. For protein alignment, it is using BLOSUM62 by default. Currently the substitution matrix is not configurable.

Note: If you supply LocatableSeq objects to pairwise_alignment, pairwise_alignment_score, align_and_show or sequence_profile and the sequence supplied contains gaps, these functions will treat these sequences as if they are without gaps.

DEPENDENCIES

This package comes with the main bioperl distribution. You also need to install the lastest bioperl-ext package which contains the XS code that implements the algorithms. This package won't work if you haven't compiled the bioperl-ext package.

TO-DO

1.
Basic support for IUPAC code for DNA sequence is now implemented. X will mismatch any character. T will match U. For others, whenever there is a possibility for match, it is considered a full match, for example, W will match B.
2.
Allow custom substitution matrix for DNA. Note that for proteins, you can now use your own subsitution matirx.

FEEDBACK

Mailing Lists

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.

  [email protected]                  - General discussion
  http://bioperl.org/wiki/Mailing_lists  - About the mailing lists

Support

Please direct usage questions or support issues to the mailing list:

[email protected]

rather than to the module maintainer directly. Many experienced and reponsive experts will be able look at the problem and quickly address it. Please include a thorough description of the problem with code and data examples if at all possible.

Reporting Bugs

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

  https://github.com/bioperl/bioperl-live/issues

AUTHOR

        This implementation was written by Yee Man Chan ([email protected]).
        Copyright (c) 2003 Yee Man Chan. All rights reserved. This program
        is free software; you can redistribute it and/or modify it under
        the same terms as Perl itself. Special thanks to Aaron Mackey
        and WIlliam Pearson for the helpful discussions. [The portion
        of code inside pgreen subdirectory was borrowed from ssearch. It
        should be distributed in the same terms as ssearch.]

sequence_profile

 Title   : sequence_profile
 Usage   : $prof = $factory->sequence_profile($seq1)
 Function: Makes a dpAlign_SequenceProfile object from one sequence
 Returns : A dpAlign_SequenceProfile object
 Args    : The lone argument is a Bio::PrimarySeqI that we want to 
           build a profile for. Usually, this would be the Query sequence

pairwise_alignment_score

 Title   : pairwise_alignment_score
 Usage   : $score = $factory->pairwise_alignment_score($prof,$seq2)
 Function: Makes a SimpleAlign object from two sequences
 Returns : An integer that is the score of the optimal alignment.
 Args    : The first argument is the sequence profile obtained from a
           call to the sequence_profile function. The second argument 
           is a Bio::PrimarySeqI object to be aligned. The second argument
           is usually a sequence in the database sequence. Note
           that this function only uses Phil Green's algorithm and 
           therefore theoretically may not always give you the optimal
           score.

pairwise_alignment

 Title   : pairwise_alignment
 Usage   : $aln = $factory->pairwise_alignment($seq1,$seq2)
 Function: Makes a SimpleAlign object from two sequences
 Returns : A SimpleAlign object if there is an alignment with positive
           score. Otherwise, return undef.
 Args    : The first and second arguments are both Bio::PrimarySeqI
           objects that are to be aligned.

align_and_show

 Title   : align_and_show
 Usage   : $factory->align_and_show($seq1,$seq2,STDOUT)

match

 Title     : match 
 Usage     : $match = $factory->match() #get
           : $factory->match($value) #set
 Function  : the set get for the match score
 Example   :
 Returns   : match value
 Arguments : new value

mismatch

 Title     : mismatch 
 Usage     : $mismatch = $factory->mismatch() #get
           : $factory->mismatch($value) #set
 Function  : the set get for the mismatch penalty
 Example   :
 Returns   : mismatch value
 Arguments : new value

gap

 Title     : gap
 Usage     : $gap = $factory->gap() #get
           : $factory->gap($value) #set
 Function  : the set get for the gap penalty
 Example   :
 Returns   : gap value
 Arguments : new value

ext

 Title     : ext
 Usage     : $ext = $factory->ext() #get
           : $factory->ext($value) #set
 Function  : the set get for the ext penalty
 Example   :
 Returns   : ext value
 Arguments : new value

alg

 Title     : alg
 Usage     : $alg = $factory->alg() #get
           : $factory->alg($value) #set
 Function  : the set get for the algorithm
 Example   :
 Returns   : alg value
 Arguments : new value