| 1 |
# TFBS module for TFBS::Word::Consensus |
| 2 |
# |
| 3 |
# Copyright Boris Lenhard |
| 4 |
# |
| 5 |
# You may distribute this module under the same terms as perl itself |
| 6 |
# |
| 7 |
|
| 8 |
# POD |
| 9 |
|
| 10 |
=head1 NAME |
| 11 |
|
| 12 |
TFBS::Word - IUPAC DNA consensus word-based pattern class |
| 13 |
=head1 DESCRIPTION |
| 14 |
|
| 15 |
TFBS::Word is a base class consisting of universal constructor called by |
| 16 |
its subclasses (TFBS::Matrix::*), and word pattern manipulation methods that |
| 17 |
are independent of the word type. It is not meant to be instantiated itself. |
| 18 |
|
| 19 |
=head1 FEEDBACK |
| 20 |
|
| 21 |
Please send bug reports and other comments to the author. |
| 22 |
|
| 23 |
=head1 AUTHOR - Boris Lenhard |
| 24 |
|
| 25 |
Boris Lenhard E<lt>[email protected]<gt> |
| 26 |
|
| 27 |
=head1 APPENDIX |
| 28 |
|
| 29 |
The rest of the documentation details each of the object |
| 30 |
methods. Internal methods are preceded with an underscore. |
| 31 |
|
| 32 |
=cut |
| 33 |
|
| 34 |
package TFBS::Word::Consensus; |
| 35 |
|
| 36 |
use vars '@ISA'; |
| 37 |
|
| 38 |
use TFBS::Word; |
| 39 |
use TFBS::Matrix::PWM; |
| 40 |
|
| 41 |
use strict; |
| 42 |
|
| 43 |
@ISA = qw(TFBS::Word); |
| 44 |
|
| 45 |
|
| 46 |
=head2 new |
| 47 |
|
| 48 |
Title : new |
| 49 |
Usage : my $pwm = TFBS::Matrix::PWM->new(%args) |
| 50 |
Function: constructor for the TFBS::Matrix::PWM object |
| 51 |
Returns : a new TFBS::Matrix::PWM object |
| 52 |
Args : # you must specify the -word argument: |
| 53 |
-word, # a strig consisting of letters in |
| 54 |
# IUPAC degenerate DNA alphabet |
| 55 |
# (any of ACGTSWKMPYBDHVN) |
| 56 |
|
| 57 |
####### |
| 58 |
|
| 59 |
-name, # string, OPTIONAL |
| 60 |
-ID, # string, OPTIONAL |
| 61 |
-class, # string, OPTIONAL |
| 62 |
-tags # a hash reference reference, OPTIONAL |
| 63 |
|
| 64 |
=cut |
| 65 |
|
| 66 |
# "new" is inherited from TFBS::Word |
| 67 |
|
| 68 |
=head2 search_seq |
| 69 |
|
| 70 |
Title : search_seq |
| 71 |
Usage : my $siteset = $pwm->search_seq(%args) |
| 72 |
Function: scans a nucleotide sequence with the pattern represented |
| 73 |
by the PWM |
| 74 |
Returns : a TFBS::SiteSet object |
| 75 |
Args : # you must specify either one of the following three: |
| 76 |
|
| 77 |
-file, # the name od a fasta file (single sequence) |
| 78 |
#or |
| 79 |
-seqobj # a Bio::Seq object |
| 80 |
# (more accurately, a Bio::PrimarySeqobject or a |
| 81 |
# subclass thereof) |
| 82 |
#or |
| 83 |
-seqstring # a string containing the sequence |
| 84 |
|
| 85 |
-max_mismatches, # number of allowed positions in the site that do |
| 86 |
# not match the consensus |
| 87 |
# OPTIONAL: default 0 |
| 88 |
|
| 89 |
=cut |
| 90 |
|
| 91 |
|
| 92 |
sub search_seq { |
| 93 |
my ($self, @args) = @_; |
| 94 |
my ($max_mismatch) = $self->_rearrange([qw(MAX_MISMATCHES)], @args) or 0; |
| 95 |
$max_mismatch = 0 unless defined $max_mismatch; |
| 96 |
my $pwm = $self->to_PWM; |
| 97 |
my $siteset = $pwm->search_seq(@args, |
| 98 |
-threshold => $self->length - $max_mismatch); |
| 99 |
$self->_replace_patterns_in_siteset($siteset); |
| 100 |
return $siteset; |
| 101 |
} |
| 102 |
|
| 103 |
|
| 104 |
=head2 search_aln |
| 105 |
|
| 106 |
Title : search_aln |
| 107 |
Usage : my $site_pair_set = $pwm->search_aln(%args) |
| 108 |
Function: Scans a pairwise alignment of nucleotide sequences |
| 109 |
with the pattern represented by the word: it reports only |
| 110 |
those hits that are present in equivalent positions of both |
| 111 |
sequences and exceed a specified threshold score in both, AND |
| 112 |
are found in regions of the alignment above the specified |
| 113 |
conservation cutoff value. |
| 114 |
Returns : a TFBS::SitePairSet object |
| 115 |
Args : # you must specify either one of the following three: |
| 116 |
|
| 117 |
-file, # the name of the alignment file in Clustal |
| 118 |
format |
| 119 |
#or |
| 120 |
-alignobj # a Bio::SimpleAlign object |
| 121 |
# (more accurately, a Bio::PrimarySeqobject or a |
| 122 |
# subclass thereof) |
| 123 |
#or |
| 124 |
-alignstring # a multi-line string containing the alignment |
| 125 |
# in clustal format |
| 126 |
############# |
| 127 |
|
| 128 |
-max_mismatches, # number of allowed positions in the site that do |
| 129 |
# not match the consensus |
| 130 |
# OPTIONAL: default 0 |
| 131 |
|
| 132 |
-window, # size of the sliding window (inn nucleotides) |
| 133 |
# for calculating local conservation in the |
| 134 |
# alignment |
| 135 |
# OPTIONAL: default 50 |
| 136 |
|
| 137 |
-cutoff # conservation cutoff (%) for including the |
| 138 |
# region in the results of the pattern search |
| 139 |
# OPTIONAL: default "70%" |
| 140 |
|
| 141 |
=cut |
| 142 |
|
| 143 |
|
| 144 |
|
| 145 |
|
| 146 |
sub search_aln { |
| 147 |
my ($self, @args) = @_; |
| 148 |
my ($max_mismatch) = $self->_rearrange([qw(MAX_MISMATCHES)], @args) or 0; |
| 149 |
$max_mismatch = 0 unless defined $max_mismatch; |
| 150 |
my $pwm = $self->to_PWM; |
| 151 |
my $sitepairset = $pwm->search_aln(@args, |
| 152 |
-threshold => $self->length - $max_mismatch); |
| 153 |
$self->_replace_patterns_in_sitepairset($sitepairset); |
| 154 |
return $sitepairset; |
| 155 |
|
| 156 |
} |
| 157 |
|
| 158 |
=head2 to_PWM |
| 159 |
|
| 160 |
=cut |
| 161 |
|
| 162 |
sub to_PWM { |
| 163 |
my ($self, @args) = @_; |
| 164 |
my $pwm = TFBS::Matrix::PWM->new(-ID => $self->ID, |
| 165 |
-name => $self->name, |
| 166 |
-class => $self->class, |
| 167 |
-matrix => _consensus2matrixref($self->word), |
| 168 |
-tags => {$self->all_tags} |
| 169 |
); |
| 170 |
return $pwm; |
| 171 |
} |
| 172 |
|
| 173 |
=head2 validate_word |
| 174 |
|
| 175 |
=cut |
| 176 |
|
| 177 |
|
| 178 |
sub validate_word { |
| 179 |
my ($self, $word) = @_; |
| 180 |
$word =~ s/[ACGTSWKMRYBDHVN]//gi; |
| 181 |
return ($word eq ""); |
| 182 |
} |
| 183 |
|
| 184 |
=head2 length |
| 185 |
|
| 186 |
=cut |
| 187 |
|
| 188 |
|
| 189 |
sub length { |
| 190 |
return length $_[0]->word; |
| 191 |
} |
| 192 |
|
| 193 |
|
| 194 |
|
| 195 |
# private methods |
| 196 |
|
| 197 |
|
| 198 |
sub _replace_patterns_in_siteset { |
| 199 |
my ($self, $siteset) = @_; |
| 200 |
my $iter = $siteset->Iterator; |
| 201 |
while (my $site = $iter->next) { |
| 202 |
$site->pattern($self); |
| 203 |
} |
| 204 |
} |
| 205 |
|
| 206 |
|
| 207 |
|
| 208 |
sub _replace_patterns_in_sitepairset { |
| 209 |
my ($self, $sitepairset) = @_; |
| 210 |
my $iter = $sitepairset->Iterator; |
| 211 |
while (my $sitepair = $iter->next) { |
| 212 |
$sitepair->feature1->pattern($self); |
| 213 |
$sitepair->feature2->pattern($self); |
| 214 |
} |
| 215 |
|
| 216 |
} |
| 217 |
|
| 218 |
# utility functions |
| 219 |
|
| 220 |
sub _consensus2matrixref { |
| 221 |
my ($word) = @_; |
| 222 |
my %iupac = ( T => [0,0,0,1], |
| 223 |
G => [0,0,1,0], |
| 224 |
K => [0,0,1,1], |
| 225 |
C => [0,1,0,0], |
| 226 |
Y => [0,1,0,1], |
| 227 |
S => [0,1,1,0], |
| 228 |
B => [0,1,1,1], |
| 229 |
A => [1,0,0,0], |
| 230 |
W => [1,0,0,1], |
| 231 |
R => [1,0,1,0], |
| 232 |
D => [1,0,1,1], |
| 233 |
M => [1,1,0,0], |
| 234 |
H => [1,1,0,1], |
| 235 |
V => [1,1,1,0], |
| 236 |
N => [1,1,1,1] |
| 237 |
); |
| 238 |
my @vert_array; |
| 239 |
foreach my $letter (split '', $word) { |
| 240 |
push @vert_array, |
| 241 |
($iupac{uc($letter)} |
| 242 |
or croak ("$letter is not a legal IUPAC DNA character")); |
| 243 |
} |
| 244 |
return _transpose_arrayref(\@vert_array); |
| 245 |
|
| 246 |
} |
| 247 |
|
| 248 |
|
| 249 |
|
| 250 |
sub _transpose_arrayref { |
| 251 |
my $vert_arrayref = shift; |
| 252 |
my $maxcol = scalar(@$vert_arrayref) - 1; |
| 253 |
my @horiz_array; |
| 254 |
foreach my $row (0..3) { |
| 255 |
push @horiz_array, [ map { $vert_arrayref->[$_][$row] } 0..$maxcol ]; |
| 256 |
|
| 257 |
} |
| 258 |
return \@horiz_array; |
| 259 |
} |
| 260 |
|
| 261 |
|
| 262 |
|
| 263 |
1; |