line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
# |
2
|
|
|
|
|
|
|
# BioPerl module for Bio::Tree::DistanceFactory |
3
|
|
|
|
|
|
|
# |
4
|
|
|
|
|
|
|
# Please direct questions and support issues to |
5
|
|
|
|
|
|
|
# |
6
|
|
|
|
|
|
|
# Cared for by Jason Stajich |
7
|
|
|
|
|
|
|
# |
8
|
|
|
|
|
|
|
# Copyright Jason Stajich |
9
|
|
|
|
|
|
|
# |
10
|
|
|
|
|
|
|
# You may distribute this module under the same terms as perl itself |
11
|
|
|
|
|
|
|
|
12
|
|
|
|
|
|
|
# POD documentation - main docs before the code |
13
|
|
|
|
|
|
|
|
14
|
|
|
|
|
|
|
=head1 NAME |
15
|
|
|
|
|
|
|
|
16
|
|
|
|
|
|
|
Bio::Tree::DistanceFactory - Construct a tree using distance based methods |
17
|
|
|
|
|
|
|
|
18
|
|
|
|
|
|
|
=head1 SYNOPSIS |
19
|
|
|
|
|
|
|
|
20
|
|
|
|
|
|
|
use Bio::Tree::DistanceFactory; |
21
|
|
|
|
|
|
|
use Bio::AlignIO; |
22
|
|
|
|
|
|
|
use Bio::Align::DNAStatistics; |
23
|
|
|
|
|
|
|
my $tfactory = Bio::Tree::DistanceFactory->new(-method => "NJ"); |
24
|
|
|
|
|
|
|
my $stats = Bio::Align::DNAStatistics->new(); |
25
|
|
|
|
|
|
|
|
26
|
|
|
|
|
|
|
my $alnin = Bio::AlignIO->new(-format => 'clustalw', |
27
|
|
|
|
|
|
|
-file => 'file.aln'); |
28
|
|
|
|
|
|
|
my $aln = $alnin->next_aln; |
29
|
|
|
|
|
|
|
# Of course matrix can come from a different place |
30
|
|
|
|
|
|
|
# like PHYLIP if you prefer, Bio::Matrix::IO should be able |
31
|
|
|
|
|
|
|
# to parse many things |
32
|
|
|
|
|
|
|
my $jcmatrix = $stats->distance(-align => $aln, |
33
|
|
|
|
|
|
|
-method => 'Jukes-Cantor'); |
34
|
|
|
|
|
|
|
my $tree = $tfactory->make_tree($jcmatrix); |
35
|
|
|
|
|
|
|
|
36
|
|
|
|
|
|
|
|
37
|
|
|
|
|
|
|
=head1 DESCRIPTION |
38
|
|
|
|
|
|
|
|
39
|
|
|
|
|
|
|
This is a factory which will construct a phylogenetic tree based on |
40
|
|
|
|
|
|
|
the pairwise sequence distances for a set of sequences. Currently |
41
|
|
|
|
|
|
|
UPGMA (Sokal and Michener 1958) and NJ (Saitou and Nei 1987) tree |
42
|
|
|
|
|
|
|
construction methods are implemented. |
43
|
|
|
|
|
|
|
|
44
|
|
|
|
|
|
|
=head1 REFERENCES |
45
|
|
|
|
|
|
|
|
46
|
|
|
|
|
|
|
Eddy SR, Durbin R, Krogh A, Mitchison G, (1998) "Biological Sequence Analysis", |
47
|
|
|
|
|
|
|
Cambridge Univ Press, Cambridge, UK. |
48
|
|
|
|
|
|
|
|
49
|
|
|
|
|
|
|
Howe K, Bateman A, Durbin R, (2002) "QuickTree: building huge |
50
|
|
|
|
|
|
|
Neighbour-Joining trees of protein sequences." Bioinformatics |
51
|
|
|
|
|
|
|
18(11):1546-1547. |
52
|
|
|
|
|
|
|
|
53
|
|
|
|
|
|
|
Saitou N and Nei M, (1987) "The neighbor-joining method: a new method |
54
|
|
|
|
|
|
|
for reconstructing phylogenetic trees." Mol Biol Evol 4(4):406-25. |
55
|
|
|
|
|
|
|
|
56
|
|
|
|
|
|
|
=head1 FEEDBACK |
57
|
|
|
|
|
|
|
|
58
|
|
|
|
|
|
|
=head2 Mailing Lists |
59
|
|
|
|
|
|
|
|
60
|
|
|
|
|
|
|
User feedback is an integral part of the evolution of this and other |
61
|
|
|
|
|
|
|
Bioperl modules. Send your comments and suggestions preferably to |
62
|
|
|
|
|
|
|
the Bioperl mailing list. Your participation is much appreciated. |
63
|
|
|
|
|
|
|
|
64
|
|
|
|
|
|
|
bioperl-l@bioperl.org - General discussion |
65
|
|
|
|
|
|
|
http://bioperl.org/wiki/Mailing_lists - About the mailing lists |
66
|
|
|
|
|
|
|
|
67
|
|
|
|
|
|
|
=head2 Support |
68
|
|
|
|
|
|
|
|
69
|
|
|
|
|
|
|
Please direct usage questions or support issues to the mailing list: |
70
|
|
|
|
|
|
|
|
71
|
|
|
|
|
|
|
I |
72
|
|
|
|
|
|
|
|
73
|
|
|
|
|
|
|
rather than to the module maintainer directly. Many experienced and |
74
|
|
|
|
|
|
|
reponsive experts will be able look at the problem and quickly |
75
|
|
|
|
|
|
|
address it. Please include a thorough description of the problem |
76
|
|
|
|
|
|
|
with code and data examples if at all possible. |
77
|
|
|
|
|
|
|
|
78
|
|
|
|
|
|
|
=head2 Reporting Bugs |
79
|
|
|
|
|
|
|
|
80
|
|
|
|
|
|
|
Report bugs to the Bioperl bug tracking system to help us keep track |
81
|
|
|
|
|
|
|
of the bugs and their resolution. Bug reports can be submitted the web: |
82
|
|
|
|
|
|
|
|
83
|
|
|
|
|
|
|
https://github.com/bioperl/bioperl-live/issues |
84
|
|
|
|
|
|
|
|
85
|
|
|
|
|
|
|
=head1 AUTHOR - Jason Stajich |
86
|
|
|
|
|
|
|
|
87
|
|
|
|
|
|
|
Email jason-at-bioperl.org |
88
|
|
|
|
|
|
|
|
89
|
|
|
|
|
|
|
=head1 APPENDIX |
90
|
|
|
|
|
|
|
|
91
|
|
|
|
|
|
|
The rest of the documentation details each of the object methods. |
92
|
|
|
|
|
|
|
Internal methods are usually preceded with a _ |
93
|
|
|
|
|
|
|
|
94
|
|
|
|
|
|
|
=cut |
95
|
|
|
|
|
|
|
|
96
|
|
|
|
|
|
|
package Bio::Tree::DistanceFactory; |
97
|
1
|
|
|
1
|
|
825
|
use vars qw($DefaultMethod $Precision); |
|
1
|
|
|
|
|
2
|
|
|
1
|
|
|
|
|
48
|
|
98
|
1
|
|
|
1
|
|
4
|
use strict; |
|
1
|
|
|
|
|
1
|
|
|
1
|
|
|
|
|
27
|
|
99
|
|
|
|
|
|
|
|
100
|
|
|
|
|
|
|
# some defaults |
101
|
|
|
|
|
|
|
$DefaultMethod = 'UPGMA'; |
102
|
|
|
|
|
|
|
$Precision = 5; |
103
|
|
|
|
|
|
|
|
104
|
1
|
|
|
1
|
|
313
|
use Bio::Tree::Node; |
|
1
|
|
|
|
|
1
|
|
|
1
|
|
|
|
|
24
|
|
105
|
1
|
|
|
1
|
|
279
|
use Bio::Tree::Tree; |
|
1
|
|
|
|
|
3
|
|
|
1
|
|
|
|
|
26
|
|
106
|
|
|
|
|
|
|
|
107
|
1
|
|
|
1
|
|
6
|
use base qw(Bio::Root::Root); |
|
1
|
|
|
|
|
2
|
|
|
1
|
|
|
|
|
1553
|
|
108
|
|
|
|
|
|
|
|
109
|
|
|
|
|
|
|
=head2 new |
110
|
|
|
|
|
|
|
|
111
|
|
|
|
|
|
|
Title : new |
112
|
|
|
|
|
|
|
Usage : my $obj = Bio::Tree::DistanceFactory->new(); |
113
|
|
|
|
|
|
|
Function: Builds a new Bio::Tree::DistanceFactory object |
114
|
|
|
|
|
|
|
Returns : an instance of Bio::Tree::DistanceFactory |
115
|
|
|
|
|
|
|
Args : -method => 'NJ' or 'UPGMA' |
116
|
|
|
|
|
|
|
|
117
|
|
|
|
|
|
|
|
118
|
|
|
|
|
|
|
=cut |
119
|
|
|
|
|
|
|
|
120
|
|
|
|
|
|
|
sub new { |
121
|
1
|
|
|
1
|
1
|
443
|
my($class,@args) = @_; |
122
|
1
|
|
|
|
|
7
|
my $self = $class->SUPER::new(@args); |
123
|
|
|
|
|
|
|
|
124
|
1
|
|
|
|
|
7
|
my ($method) = $self->_rearrange([qw(METHOD)], |
125
|
|
|
|
|
|
|
@args); |
126
|
1
|
|
33
|
|
|
8
|
$self->method($method || $DefaultMethod); |
127
|
1
|
|
|
|
|
2
|
return $self; |
128
|
|
|
|
|
|
|
} |
129
|
|
|
|
|
|
|
|
130
|
|
|
|
|
|
|
=head2 make_tree |
131
|
|
|
|
|
|
|
|
132
|
|
|
|
|
|
|
Title : make_tree |
133
|
|
|
|
|
|
|
Usage : my $tree = $disttreefact->make_tree($matrix); |
134
|
|
|
|
|
|
|
Function: Build a Tree based on a distance matrix |
135
|
|
|
|
|
|
|
Returns : L |
136
|
|
|
|
|
|
|
Args : L object |
137
|
|
|
|
|
|
|
|
138
|
|
|
|
|
|
|
|
139
|
|
|
|
|
|
|
=cut |
140
|
|
|
|
|
|
|
|
141
|
|
|
|
|
|
|
sub make_tree{ |
142
|
1
|
|
|
1
|
1
|
5
|
my ($self,$matrix) = @_; |
143
|
1
|
50
|
33
|
|
|
10
|
if( ! defined $matrix || !ref($matrix) || |
|
|
|
33
|
|
|
|
|
144
|
|
|
|
|
|
|
! $matrix->isa('Bio::Matrix::MatrixI') ) { |
145
|
0
|
|
|
|
|
0
|
$self->warn("Need to provide a valid Bio::Matrix::MatrixI object to make_tree"); |
146
|
0
|
|
|
|
|
0
|
return; |
147
|
|
|
|
|
|
|
} |
148
|
|
|
|
|
|
|
|
149
|
1
|
|
|
|
|
3
|
my $method = uc ($self->method); |
150
|
1
|
50
|
|
|
|
6
|
if( $method =~ /NJ/i ) { |
|
|
0
|
|
|
|
|
|
151
|
1
|
|
|
|
|
4
|
return $self->_nj($matrix); |
152
|
|
|
|
|
|
|
} elsif( $method =~ /UPGMA/i ) { |
153
|
0
|
|
|
|
|
0
|
return $self->_upgma($matrix); |
154
|
|
|
|
|
|
|
} else { |
155
|
0
|
|
|
|
|
0
|
$self->warn("Unknown tree construction method '$method'. Cannot run."); |
156
|
0
|
|
|
|
|
0
|
return; |
157
|
|
|
|
|
|
|
} |
158
|
|
|
|
|
|
|
|
159
|
|
|
|
|
|
|
} |
160
|
|
|
|
|
|
|
|
161
|
|
|
|
|
|
|
|
162
|
|
|
|
|
|
|
=head2 _nj |
163
|
|
|
|
|
|
|
|
164
|
|
|
|
|
|
|
Title : _nj |
165
|
|
|
|
|
|
|
Usage : my $tree = $disttreefact->_nj($matrix); |
166
|
|
|
|
|
|
|
Function: Construct a tree based on distance matrix using the |
167
|
|
|
|
|
|
|
Neighbor Joining algorithm (Saitou and Nei, 1987) |
168
|
|
|
|
|
|
|
Implementation based on Kevin Howe's Quicktree implementation |
169
|
|
|
|
|
|
|
and uses his tricks (some based on Bill Bruno's work) to eliminate |
170
|
|
|
|
|
|
|
negative branch lengths |
171
|
|
|
|
|
|
|
Returns : L |
172
|
|
|
|
|
|
|
Args : L object |
173
|
|
|
|
|
|
|
|
174
|
|
|
|
|
|
|
=cut |
175
|
|
|
|
|
|
|
|
176
|
|
|
|
|
|
|
sub _nj { |
177
|
1
|
|
|
1
|
|
2
|
my ($self,$distmat) = @_; |
178
|
|
|
|
|
|
|
|
179
|
|
|
|
|
|
|
# we assume type checking of $aln has already been done |
180
|
|
|
|
|
|
|
# client shouldn't be calling this directly anyways, using the |
181
|
|
|
|
|
|
|
# make_tree method is preferred |
182
|
|
|
|
|
|
|
|
183
|
|
|
|
|
|
|
# so that we can trim the number of digits shown as the branch length |
184
|
1
|
|
|
|
|
4
|
my $precisionstr = "%.$Precision"."f"; |
185
|
|
|
|
|
|
|
|
186
|
1
|
|
|
|
|
3
|
my @names = $distmat->column_names; |
187
|
1
|
|
|
|
|
3
|
my $N = scalar @names; |
188
|
1
|
|
|
|
|
1
|
my ($i,$j,$m,@nodes,$mat,@r); |
189
|
1
|
|
|
|
|
2
|
my $L = $N; |
190
|
|
|
|
|
|
|
|
191
|
1
|
50
|
|
|
|
5
|
if( $N < 2 ) { |
|
|
50
|
|
|
|
|
|
192
|
0
|
|
|
|
|
0
|
$self->warn("Can only perform NJ treebuilding on sets of 2 or more species\n"); |
193
|
0
|
|
|
|
|
0
|
return; |
194
|
|
|
|
|
|
|
} elsif( $N == 2 ) { |
195
|
0
|
|
|
|
|
0
|
$i = 0; |
196
|
0
|
|
|
|
|
0
|
my $d = sprintf($precisionstr, |
197
|
|
|
|
|
|
|
$distmat->get_entry($names[0],$names[1]) / 2); |
198
|
0
|
|
|
|
|
0
|
my $root = Bio::Tree::Node->new(); |
199
|
0
|
|
|
|
|
0
|
for my $nm ( @names ) { |
200
|
0
|
|
|
|
|
0
|
$root->add_Descendents( Bio::Tree::Node->new(-id => $nm, |
201
|
|
|
|
|
|
|
-branch_length => $d)); |
202
|
|
|
|
|
|
|
} |
203
|
0
|
|
|
|
|
0
|
return Bio::Tree::Tree(-root => $root); |
204
|
|
|
|
|
|
|
} |
205
|
1
|
|
|
|
|
2
|
my $c = 0; |
206
|
|
|
|
|
|
|
|
207
|
1
|
|
|
|
|
4
|
for ( $i = 0; $i < $N; $i++ ) { |
208
|
14
|
|
|
|
|
38
|
push @nodes, Bio::Tree::Node->new(-id => $names[$i]); |
209
|
14
|
|
|
|
|
12
|
my $ri = 0; |
210
|
14
|
|
|
|
|
22
|
for( $j = 0; $j < $N; $j++ ) { |
211
|
196
|
|
|
|
|
263
|
$mat->[$i][$j] = $distmat->get_entry($names[$i],$names[$j]); |
212
|
196
|
|
|
|
|
359
|
$ri += $mat->[$i][$j]; |
213
|
|
|
|
|
|
|
} |
214
|
14
|
|
|
|
|
34
|
$r[$i] = $ri / ($L -2); |
215
|
|
|
|
|
|
|
} |
216
|
|
|
|
|
|
|
|
217
|
1
|
|
|
|
|
5
|
for( my $nodecount = 0; $nodecount < $N-3; $nodecount++) { |
218
|
11
|
|
|
|
|
9
|
my ($mini,$minj,$min); |
219
|
11
|
|
|
|
|
17
|
for($i = 0; $i < $N; $i++ ) { |
220
|
154
|
100
|
|
|
|
198
|
next unless defined $nodes[$i]; |
221
|
99
|
|
|
|
|
126
|
for( $j = 0; $j < $i; $j++ ) { |
222
|
516
|
100
|
|
|
|
581
|
next unless defined $nodes[$j]; |
223
|
451
|
|
|
|
|
319
|
my $dist = $mat->[$i][$j] - ($r[$i] + $r[$j]); |
224
|
451
|
100
|
100
|
|
|
1294
|
if( ! defined $min || |
225
|
|
|
|
|
|
|
$dist <= $min) { |
226
|
32
|
|
|
|
|
60
|
($mini,$minj,$min) = ($i,$j,$dist); |
227
|
|
|
|
|
|
|
} |
228
|
|
|
|
|
|
|
} |
229
|
|
|
|
|
|
|
} |
230
|
11
|
|
|
|
|
9
|
my $dij = $mat->[$mini][$minj]; |
231
|
11
|
|
|
|
|
15
|
my $dist_i = ($dij + $r[$mini] - $r[$minj]) / 2; |
232
|
11
|
|
|
|
|
11
|
my $dist_j = $dij - $dist_i; |
233
|
|
|
|
|
|
|
|
234
|
|
|
|
|
|
|
# deal with negative branch lengths |
235
|
|
|
|
|
|
|
# per code in K.Howe's quicktree |
236
|
11
|
50
|
|
|
|
25
|
if( $dist_i < 0 ) { |
|
|
50
|
|
|
|
|
|
237
|
0
|
|
|
|
|
0
|
$dist_i = 0; |
238
|
0
|
|
|
|
|
0
|
$dist_j = $dij; |
239
|
0
|
0
|
|
|
|
0
|
$dist_j = 0 if( $dist_j < 0 ); |
240
|
|
|
|
|
|
|
} elsif( $dist_j < 0 ) { |
241
|
0
|
|
|
|
|
0
|
$dist_j = 0; |
242
|
0
|
|
|
|
|
0
|
$dist_i = $dij; |
243
|
0
|
0
|
|
|
|
0
|
$dist_i = 0 if( $dist_i < 0 ); |
244
|
|
|
|
|
|
|
} |
245
|
|
|
|
|
|
|
|
246
|
11
|
|
|
|
|
80
|
$nodes[$mini]->branch_length(sprintf($precisionstr,$dist_i)); |
247
|
11
|
|
|
|
|
45
|
$nodes[$minj]->branch_length(sprintf($precisionstr,$dist_j)); |
248
|
|
|
|
|
|
|
|
249
|
11
|
|
|
|
|
32
|
my $newnode = Bio::Tree::Node->new(-descendents => [ $nodes[$mini], |
250
|
|
|
|
|
|
|
$nodes[$minj] ]); |
251
|
|
|
|
|
|
|
|
252
|
11
|
|
|
|
|
12
|
$nodes[$mini] = $newnode; |
253
|
11
|
|
|
|
|
11
|
delete $nodes[$minj]; |
254
|
|
|
|
|
|
|
|
255
|
|
|
|
|
|
|
# update the distance matrix |
256
|
11
|
|
|
|
|
9
|
$r[$mini] = 0; |
257
|
11
|
|
|
|
|
9
|
my ($dmi,$dmj); |
258
|
11
|
|
|
|
|
19
|
for( $m = 0; $m < $N; $m++ ) { |
259
|
154
|
100
|
|
|
|
203
|
next unless defined $nodes[$m]; |
260
|
88
|
100
|
|
|
|
107
|
if( $m != $mini ) { |
261
|
77
|
|
|
|
|
63
|
$dmj = $mat->[$m][$minj]; |
262
|
|
|
|
|
|
|
|
263
|
77
|
|
|
|
|
42
|
my ($row,$col); |
264
|
77
|
|
|
|
|
47
|
($row,$col) = ($m,$mini); |
265
|
77
|
|
|
|
|
59
|
$dmi = $mat->[$row][$col]; |
266
|
|
|
|
|
|
|
|
267
|
|
|
|
|
|
|
# from K.Howe's notes in quicktree |
268
|
|
|
|
|
|
|
# we can actually adjust r[m] here, by using the form: |
269
|
|
|
|
|
|
|
# rm = ((rm * numseqs) - dmi - dmj + dmk) / (numseqs-1) |
270
|
|
|
|
|
|
|
|
271
|
|
|
|
|
|
|
# Note: in Bill Bruno's method for negative branch |
272
|
|
|
|
|
|
|
# elimination, then if either dist_i is positive and |
273
|
|
|
|
|
|
|
# dist_j is 0, or dist_i is zero and dist_j is positive |
274
|
|
|
|
|
|
|
# (after adjustment) then the matrix entry is formed |
275
|
|
|
|
|
|
|
# from the distance to the node in question (m) to the |
276
|
|
|
|
|
|
|
# node with the zero branch length (whichever it was). |
277
|
|
|
|
|
|
|
# I think my code already has the same effect; this is |
278
|
|
|
|
|
|
|
# certainly true if dij is equal to dist_i + dist_j, |
279
|
|
|
|
|
|
|
# which it should have been fixed to |
280
|
|
|
|
|
|
|
|
281
|
77
|
|
|
|
|
67
|
my $dmk = $mat->[$row][$col] = $mat->[$col][$row] = |
282
|
|
|
|
|
|
|
($dmi + $dmj - $dij) / 2; |
283
|
|
|
|
|
|
|
|
284
|
|
|
|
|
|
|
# If we don't want to try and correct negative brlens |
285
|
|
|
|
|
|
|
# this is essentially what is in Edddy et al, BSA book. |
286
|
|
|
|
|
|
|
# $r[$m] = (($r[$m] * $L) - $dmi - $dmj + $dmk) / ($L-1); |
287
|
|
|
|
|
|
|
# |
288
|
77
|
|
|
|
|
80
|
$r[$m] = (($r[$m] * ($L - 2)) - $dmi - $dmj + |
289
|
|
|
|
|
|
|
$mat->[$row][$col]) / ( $L - 3); |
290
|
77
|
|
|
|
|
101
|
$r[$mini] += $dmk; |
291
|
|
|
|
|
|
|
} |
292
|
|
|
|
|
|
|
} |
293
|
11
|
|
|
|
|
8
|
$L--; |
294
|
11
|
|
|
|
|
20
|
$r[$mini] /= $L - 2; |
295
|
|
|
|
|
|
|
} |
296
|
|
|
|
|
|
|
|
297
|
|
|
|
|
|
|
# should be 3 nodes left |
298
|
1
|
|
|
|
|
2
|
my (@leftovernodes,@leftovers); |
299
|
1
|
|
|
|
|
20
|
for( my $k = 0; $k < $N; $k++ ) { |
300
|
14
|
100
|
|
|
|
25
|
if( defined $nodes[$k] ) { |
301
|
3
|
|
|
|
|
2
|
push @leftovers, $k; |
302
|
3
|
|
|
|
|
8
|
push @leftovernodes, $nodes[$k]; |
303
|
|
|
|
|
|
|
} |
304
|
|
|
|
|
|
|
} |
305
|
1
|
|
|
|
|
2
|
my ($l_0,$l_1,$l_2) = @leftovers; |
306
|
|
|
|
|
|
|
|
307
|
1
|
|
|
|
|
4
|
my $dist_i = ( $mat->[$l_1][$l_0] + $mat->[$l_2][$l_0] - |
308
|
|
|
|
|
|
|
$mat->[$l_2][$l_1] ) / 2; |
309
|
|
|
|
|
|
|
|
310
|
1
|
|
|
|
|
2
|
my $dist_j = ( $mat->[$l_1][$l_0] - $dist_i); |
311
|
1
|
|
|
|
|
3
|
my $dist_k = ( $mat->[$l_2][$l_0] - $dist_i); |
312
|
|
|
|
|
|
|
|
313
|
|
|
|
|
|
|
# This is Kev's code to get rid of negative branch lengths |
314
|
1
|
50
|
|
|
|
18
|
if( $dist_i < 0 ) { |
|
|
50
|
|
|
|
|
|
|
|
50
|
|
|
|
|
|
315
|
0
|
|
|
|
|
0
|
$dist_i = 0; |
316
|
0
|
|
|
|
|
0
|
$dist_j = $mat->[$l_1][$l_0]; |
317
|
0
|
|
|
|
|
0
|
$dist_k = $mat->[$l_2][$l_0]; |
318
|
0
|
0
|
|
|
|
0
|
if( $dist_j < 0 ) { |
|
|
0
|
|
|
|
|
|
319
|
0
|
|
|
|
|
0
|
$dist_j = 0; |
320
|
0
|
|
|
|
|
0
|
$dist_k = ( $mat->[$l_2][$l_0] + $mat->[$l_2][$l_1] ) / 2; |
321
|
0
|
0
|
|
|
|
0
|
$dist_k = 0 if( $dist_k < 0 ); |
322
|
|
|
|
|
|
|
} elsif( $dist_k < 0 ) { |
323
|
0
|
|
|
|
|
0
|
$dist_k = 0; |
324
|
0
|
|
|
|
|
0
|
$dist_j = ($mat->[$l_1][$l_0] + $mat->[$l_2][$l_1]) / 2; |
325
|
0
|
0
|
|
|
|
0
|
$dist_j = 0 if( $dist_j < 0 ); |
326
|
|
|
|
|
|
|
} |
327
|
|
|
|
|
|
|
} elsif( $dist_j < 0 ) { |
328
|
0
|
|
|
|
|
0
|
$dist_j = 0; |
329
|
0
|
|
|
|
|
0
|
$dist_i = $mat->[$l_1][$l_0]; |
330
|
0
|
|
|
|
|
0
|
$dist_k = $mat->[$l_2][$l_1]; |
331
|
0
|
0
|
|
|
|
0
|
if( $dist_i < 0 ) { |
|
|
0
|
|
|
|
|
|
332
|
0
|
|
|
|
|
0
|
$dist_i = 0; |
333
|
0
|
|
|
|
|
0
|
$dist_k = ( $mat->[$l_2][$l_0] + $mat->[$l_2][$l_1]) / 2; |
334
|
0
|
0
|
|
|
|
0
|
$dist_k = 0 if( $dist_k < 0 ); |
335
|
|
|
|
|
|
|
} elsif( $dist_k < 0 ) { |
336
|
0
|
|
|
|
|
0
|
$dist_k = 0; |
337
|
0
|
|
|
|
|
0
|
$dist_i = ( $mat->[$l_1][$l_0] + $mat->[$l_2][$l_0]) / 2; |
338
|
0
|
0
|
|
|
|
0
|
$dist_i = 0 if( $dist_i < 0 ); |
339
|
|
|
|
|
|
|
} |
340
|
|
|
|
|
|
|
} elsif( $dist_k < 0 ) { |
341
|
0
|
|
|
|
|
0
|
$dist_k = 0; |
342
|
0
|
|
|
|
|
0
|
$dist_i = $mat->[$l_2][$l_0]; |
343
|
0
|
|
|
|
|
0
|
$dist_j = $mat->[$l_2][$l_1]; |
344
|
0
|
0
|
|
|
|
0
|
if( $dist_i < 0 ) { |
|
|
0
|
|
|
|
|
|
345
|
0
|
|
|
|
|
0
|
$dist_i = 0; |
346
|
0
|
|
|
|
|
0
|
$dist_j = ( $mat->[$l_1][$l_0] + $mat->[$l_2][$l_1] ) / 2; |
347
|
0
|
0
|
|
|
|
0
|
$dist_j = 0 if $dist_j < 0; |
348
|
|
|
|
|
|
|
} elsif( $dist_j < 0 ) { |
349
|
0
|
|
|
|
|
0
|
$dist_j = 0; |
350
|
0
|
|
|
|
|
0
|
$dist_i = ($mat->[$l_1][$l_0] + $mat->[$l_2][$l_0]) / 2; |
351
|
0
|
0
|
|
|
|
0
|
$dist_i = 0 if $dist_i < 0; |
352
|
|
|
|
|
|
|
} |
353
|
|
|
|
|
|
|
} |
354
|
1
|
|
|
|
|
10
|
$leftovernodes[0]->branch_length(sprintf($precisionstr,$dist_i)); |
355
|
1
|
|
|
|
|
6
|
$leftovernodes[1]->branch_length(sprintf($precisionstr,$dist_j)); |
356
|
1
|
|
|
|
|
18
|
$leftovernodes[2]->branch_length(sprintf($precisionstr,$dist_k)); |
357
|
|
|
|
|
|
|
|
358
|
1
|
|
|
|
|
7
|
Bio::Tree::Tree->new(-root => Bio::Tree::Node->new |
359
|
|
|
|
|
|
|
(-descendents => \@leftovernodes)); |
360
|
|
|
|
|
|
|
} |
361
|
|
|
|
|
|
|
|
362
|
|
|
|
|
|
|
=head2 _upgma |
363
|
|
|
|
|
|
|
|
364
|
|
|
|
|
|
|
Title : _upgma |
365
|
|
|
|
|
|
|
Usage : my $tree = $disttreefact->_upgma($matrix); |
366
|
|
|
|
|
|
|
Function: Construct a tree based on alignment using UPGMA |
367
|
|
|
|
|
|
|
Returns : L |
368
|
|
|
|
|
|
|
Args : L object |
369
|
|
|
|
|
|
|
|
370
|
|
|
|
|
|
|
|
371
|
|
|
|
|
|
|
=cut |
372
|
|
|
|
|
|
|
|
373
|
|
|
|
|
|
|
sub _upgma{ |
374
|
0
|
|
|
0
|
|
0
|
my ($self,$distmat) = @_; |
375
|
|
|
|
|
|
|
# we assume type checking of $matrix has already been done |
376
|
|
|
|
|
|
|
# client shouldn't be calling this directly anyways, using the |
377
|
|
|
|
|
|
|
# make_tree method is preferred |
378
|
|
|
|
|
|
|
|
379
|
|
|
|
|
|
|
# algorithm, from Eddy, Durbin, Krogh, Mitchison, 1998 |
380
|
|
|
|
|
|
|
# originally by Sokal and Michener 1956 |
381
|
|
|
|
|
|
|
|
382
|
0
|
|
|
|
|
0
|
my $precisionstr = "%.$Precision"."f"; |
383
|
|
|
|
|
|
|
|
384
|
0
|
|
|
|
|
0
|
my ($i,$j,$x,$y,@dmat,@orig,@nodes); |
385
|
|
|
|
|
|
|
|
386
|
0
|
|
|
|
|
0
|
my @names = $distmat->column_names; |
387
|
0
|
|
|
|
|
0
|
my $c = 0; |
388
|
|
|
|
|
|
|
my @clusters = map { |
389
|
0
|
|
|
|
|
0
|
my $r = { 'id' => $c, |
|
0
|
|
|
|
|
0
|
|
390
|
|
|
|
|
|
|
'height' => 0, |
391
|
|
|
|
|
|
|
'contains' => [$c], |
392
|
|
|
|
|
|
|
}; |
393
|
0
|
|
|
|
|
0
|
$c++; |
394
|
0
|
|
|
|
|
0
|
$r; |
395
|
|
|
|
|
|
|
} @names; |
396
|
|
|
|
|
|
|
|
397
|
0
|
|
|
|
|
0
|
my $K = scalar @clusters; |
398
|
0
|
|
|
|
|
0
|
my (@mins,$min); |
399
|
0
|
|
|
|
|
0
|
for ( $i = 0; $i < $K; $i++ ) { |
400
|
0
|
|
|
|
|
0
|
for( $j = $i+1; $j < $K; $j++ ) { |
401
|
0
|
|
|
|
|
0
|
my $d = $distmat->get_entry($names[$i],$names[$j]); |
402
|
|
|
|
|
|
|
# get Min here on first time around, save 1 cycle |
403
|
0
|
|
|
|
|
0
|
$dmat[$j][$i] = $dmat[$i][$j] = $d; |
404
|
0
|
|
|
|
|
0
|
$orig[$i][$j] = $orig[$j][$i] = $d; |
405
|
0
|
0
|
0
|
|
|
0
|
if ( ! defined $min || $d <= $min ) { |
406
|
0
|
0
|
0
|
|
|
0
|
if( defined $min && $min == $d ) { |
407
|
0
|
|
|
|
|
0
|
push @mins, [$i,$j]; |
408
|
|
|
|
|
|
|
} else { |
409
|
0
|
|
|
|
|
0
|
@mins = [$i,$j]; |
410
|
0
|
|
|
|
|
0
|
$min = $d; |
411
|
|
|
|
|
|
|
} |
412
|
|
|
|
|
|
|
} |
413
|
|
|
|
|
|
|
} |
414
|
|
|
|
|
|
|
} |
415
|
|
|
|
|
|
|
# distance between each cluster is avg distance |
416
|
|
|
|
|
|
|
# between pairs of sequences from each cluster |
417
|
0
|
|
|
|
|
0
|
while( $K > 1 ) { |
418
|
|
|
|
|
|
|
# fencepost - we already have found the $min |
419
|
|
|
|
|
|
|
# so very first time loop is executed we can skip checking |
420
|
0
|
0
|
|
|
|
0
|
unless( defined $min ) { |
421
|
0
|
|
|
|
|
0
|
for($i = 0; $i < $K; $i++ ) { |
422
|
0
|
|
|
|
|
0
|
for( $j = $i+1; $j < $K; $j++ ) { |
423
|
0
|
|
|
|
|
0
|
my $dij = $dmat[$i][$j]; |
424
|
0
|
0
|
0
|
|
|
0
|
if( ! defined $min || |
425
|
|
|
|
|
|
|
$dij <= $min) { |
426
|
0
|
0
|
0
|
|
|
0
|
if( defined $min && |
427
|
|
|
|
|
|
|
$min == $dij ) { |
428
|
0
|
|
|
|
|
0
|
push @mins, [$i,$j]; |
429
|
|
|
|
|
|
|
} else { |
430
|
0
|
|
|
|
|
0
|
@mins = [ $i,$j ]; |
431
|
0
|
|
|
|
|
0
|
$min = $dij; |
432
|
|
|
|
|
|
|
} |
433
|
|
|
|
|
|
|
} |
434
|
|
|
|
|
|
|
} |
435
|
|
|
|
|
|
|
} |
436
|
|
|
|
|
|
|
} |
437
|
|
|
|
|
|
|
# randomly break ties |
438
|
0
|
|
|
|
|
0
|
($x,$y) = @{ $mins[int(rand(scalar @mins))] }; |
|
0
|
|
|
|
|
0
|
|
439
|
|
|
|
|
|
|
|
440
|
|
|
|
|
|
|
# now we are going to join clusters x and y, make a new cluster |
441
|
|
|
|
|
|
|
|
442
|
0
|
|
|
|
|
0
|
my $node = Bio::Tree::Node->new(); |
443
|
0
|
|
|
|
|
0
|
my @subids; |
444
|
0
|
|
|
|
|
0
|
for my $cid ( $x,$y ) { |
445
|
0
|
|
|
|
|
0
|
my $nid = $clusters[$cid]->{'id'}; |
446
|
0
|
0
|
|
|
|
0
|
if( ! defined $nodes[$nid] ) { |
447
|
0
|
|
|
|
|
0
|
$nodes[$nid] = Bio::Tree::Node->new(-id => $names[$nid]); |
448
|
|
|
|
|
|
|
} |
449
|
|
|
|
|
|
|
$nodes[$nid]->branch_length |
450
|
0
|
|
|
|
|
0
|
(sprintf($precisionstr,$min/2 - $clusters[$cid]->{'height'})); |
451
|
0
|
|
|
|
|
0
|
$node->add_Descendent($nodes[$nid]); |
452
|
0
|
|
|
|
|
0
|
push @subids, @{ $clusters[$cid]->{'contains'} }; |
|
0
|
|
|
|
|
0
|
|
453
|
|
|
|
|
|
|
} |
454
|
0
|
|
|
|
|
0
|
my $cluster = { 'id' => $c++, |
455
|
|
|
|
|
|
|
'height' => $min / 2, |
456
|
|
|
|
|
|
|
'contains' => [@subids], |
457
|
|
|
|
|
|
|
}; |
458
|
|
|
|
|
|
|
|
459
|
0
|
|
|
|
|
0
|
$K--; # we are going to drop the last node so go ahead and decrement K |
460
|
0
|
|
|
|
|
0
|
$nodes[$cluster->{'id'}] = $node; |
461
|
0
|
0
|
|
|
|
0
|
if ( $y != $K ) { |
462
|
0
|
|
|
|
|
0
|
$clusters[$y] = $clusters[$K]; |
463
|
0
|
|
|
|
|
0
|
$dmat[$y] = $dmat[$K]; |
464
|
0
|
|
|
|
|
0
|
for ( $i = 0; $i < $K; $i++ ) { |
465
|
0
|
|
|
|
|
0
|
$dmat[$i][$y] = $dmat[$y][$i]; |
466
|
|
|
|
|
|
|
} |
467
|
|
|
|
|
|
|
} |
468
|
0
|
|
|
|
|
0
|
delete $clusters[$K]; |
469
|
0
|
|
|
|
|
0
|
$clusters[$x] = $cluster; |
470
|
|
|
|
|
|
|
# now recalculate @dmat |
471
|
0
|
|
|
|
|
0
|
for( $i = 0; $i < $K; $i++ ) { |
472
|
0
|
0
|
|
|
|
0
|
if( $i != $x) { |
473
|
0
|
|
|
|
|
0
|
$dmat[$i][$x] = $dmat[$x][$i] = |
474
|
|
|
|
|
|
|
&_upgma_distance($clusters[$i],$clusters[$x],\@orig); |
475
|
|
|
|
|
|
|
} else { |
476
|
0
|
|
|
|
|
0
|
$dmat[$i][$i] = 0; |
477
|
|
|
|
|
|
|
} |
478
|
|
|
|
|
|
|
} |
479
|
|
|
|
|
|
|
# reset so next loop iteration |
480
|
|
|
|
|
|
|
# we will find minimum distance |
481
|
0
|
|
|
|
|
0
|
@mins = (); |
482
|
0
|
|
|
|
|
0
|
$min = undef; |
483
|
|
|
|
|
|
|
} |
484
|
0
|
|
|
|
|
0
|
Bio::Tree::Tree->new(-root => $nodes[-1]); |
485
|
|
|
|
|
|
|
} |
486
|
|
|
|
|
|
|
|
487
|
|
|
|
|
|
|
# calculate avg distance between clusters - be they |
488
|
|
|
|
|
|
|
# single sequences or the combination of multiple seqences |
489
|
|
|
|
|
|
|
# $cluster_i and $cluster_j are the clusters to operate on |
490
|
|
|
|
|
|
|
# and $distances is a matrix (arrayref of arrayrefs) of pairwise |
491
|
|
|
|
|
|
|
# differences indexed on the sequence ids - |
492
|
|
|
|
|
|
|
# so $distances->[0][1] is the distance between sequences 0 and 1 |
493
|
|
|
|
|
|
|
|
494
|
|
|
|
|
|
|
sub _upgma_distance { |
495
|
0
|
|
|
0
|
|
0
|
my ($cluster_i, $cluster_j, $distances) = @_; |
496
|
0
|
|
|
|
|
0
|
my $ilen = scalar @{ $cluster_i->{'contains'} }; |
|
0
|
|
|
|
|
0
|
|
497
|
0
|
|
|
|
|
0
|
my $jlen = scalar @{ $cluster_j->{'contains'} }; |
|
0
|
|
|
|
|
0
|
|
498
|
0
|
|
|
|
|
0
|
my ($d,$count); |
499
|
0
|
|
|
|
|
0
|
for( my $i = 0; $i < $ilen; $i++ ) { |
500
|
0
|
|
|
|
|
0
|
my $i_id = $cluster_i->{'contains'}->[$i]; |
501
|
0
|
|
|
|
|
0
|
for( my $j = 0; $j < $jlen; $j++) { |
502
|
0
|
|
|
|
|
0
|
my $j_id = $cluster_j->{'contains'}->[$j]; |
503
|
0
|
0
|
|
|
|
0
|
if( ! defined $distances->[$i_id][$j_id] ) { |
504
|
0
|
|
|
|
|
0
|
warn("no value for $i_id $j_id\n"); |
505
|
|
|
|
|
|
|
} else { |
506
|
0
|
|
|
|
|
0
|
$d += $distances->[$i_id][$j_id]; |
507
|
|
|
|
|
|
|
} |
508
|
0
|
|
|
|
|
0
|
$count++; |
509
|
|
|
|
|
|
|
} |
510
|
|
|
|
|
|
|
} |
511
|
0
|
|
|
|
|
0
|
return $d / $count; |
512
|
|
|
|
|
|
|
} |
513
|
|
|
|
|
|
|
|
514
|
|
|
|
|
|
|
=head2 method |
515
|
|
|
|
|
|
|
|
516
|
|
|
|
|
|
|
Title : method |
517
|
|
|
|
|
|
|
Usage : $obj->method($newval) |
518
|
|
|
|
|
|
|
Function: |
519
|
|
|
|
|
|
|
Example : |
520
|
|
|
|
|
|
|
Returns : value of method (a scalar) |
521
|
|
|
|
|
|
|
Args : on set, new value (a scalar or undef, optional) |
522
|
|
|
|
|
|
|
|
523
|
|
|
|
|
|
|
|
524
|
|
|
|
|
|
|
=cut |
525
|
|
|
|
|
|
|
|
526
|
|
|
|
|
|
|
sub method{ |
527
|
2
|
|
|
2
|
1
|
4
|
my $self = shift; |
528
|
2
|
100
|
|
|
|
6
|
return $self->{'_method'} = shift if @_; |
529
|
1
|
|
|
|
|
3
|
return $self->{'_method'}; |
530
|
|
|
|
|
|
|
} |
531
|
|
|
|
|
|
|
|
532
|
|
|
|
|
|
|
|
533
|
|
|
|
|
|
|
=head2 check_additivity |
534
|
|
|
|
|
|
|
|
535
|
|
|
|
|
|
|
Title : check_additivity |
536
|
|
|
|
|
|
|
Usage : if( $distance->check_additivity($matrix) ) { |
537
|
|
|
|
|
|
|
} |
538
|
|
|
|
|
|
|
Function : See if matrix obeys additivity principal |
539
|
|
|
|
|
|
|
Returns : boolean |
540
|
|
|
|
|
|
|
Args : Bio::Matrix::MatrixI |
541
|
|
|
|
|
|
|
References: Based on a Java implementation by |
542
|
|
|
|
|
|
|
Peter Sestoft, sestoft@dina.kvl.dk 1999-12-07 version 0.3 |
543
|
|
|
|
|
|
|
http://www.dina.kvl.dk/~sestoft/bsa.html |
544
|
|
|
|
|
|
|
which in turn is based on algorithms described in |
545
|
|
|
|
|
|
|
R. Durbin, S. Eddy, A. Krogh, G. Mitchison. |
546
|
|
|
|
|
|
|
Biological Sequence Analysis CUP 1998, Chapter 7. |
547
|
|
|
|
|
|
|
|
548
|
|
|
|
|
|
|
=cut |
549
|
|
|
|
|
|
|
|
550
|
|
|
|
|
|
|
sub check_additivity{ |
551
|
0
|
|
|
0
|
1
|
|
my ($self,$matrix) = @_; |
552
|
0
|
|
|
|
|
|
my @names = $matrix->column_names; |
553
|
0
|
|
|
|
|
|
my $len = scalar @names; |
554
|
0
|
0
|
|
|
|
|
return unless $len >= 4; |
555
|
|
|
|
|
|
|
# look at all sets of 4 |
556
|
0
|
|
|
|
|
|
for( my $i = 0; $i < $len; $i++ ) { |
557
|
0
|
|
|
|
|
|
for( my $j = $i+1; $j< $len; $j++) { |
558
|
0
|
|
|
|
|
|
for( my $k = $j+1; $k < $len; $k ++ ) { |
559
|
0
|
|
|
|
|
|
for( my $m = $k +1; $m < $len; $m++ ) { |
560
|
0
|
|
|
|
|
|
my $DijDkm = $matrix->get_entry($names[$i],$names[$j]) + |
561
|
|
|
|
|
|
|
$matrix->get_entry($names[$k],$names[$m]); |
562
|
0
|
|
|
|
|
|
my $DikDjm = $matrix->get_entry($names[$i],$names[$k]) + |
563
|
|
|
|
|
|
|
$matrix->get_entry($names[$j],$names[$m]); |
564
|
0
|
|
|
|
|
|
my $DimDjk = $matrix->get_entry($names[$i],$names[$m]) + |
565
|
|
|
|
|
|
|
$matrix->get_entry($names[$j],$names[$k]); |
566
|
0
|
0
|
0
|
|
|
|
if( !( ( $DijDkm == $DikDjm && $DijDkm >= $DimDjk) |
|
|
|
0
|
|
|
|
|
|
|
|
0
|
|
|
|
|
|
|
|
0
|
|
|
|
|
|
|
|
0
|
|
|
|
|
567
|
|
|
|
|
|
|
|| ( $DijDkm == $DimDjk && $DijDkm >= $DikDjm) |
568
|
|
|
|
|
|
|
|| ( $DikDjm == $DimDjk && $DikDjm >= $DijDkm) )) { |
569
|
0
|
|
|
|
|
|
return 0; |
570
|
|
|
|
|
|
|
} |
571
|
|
|
|
|
|
|
} |
572
|
|
|
|
|
|
|
} |
573
|
|
|
|
|
|
|
} |
574
|
|
|
|
|
|
|
} |
575
|
0
|
|
|
|
|
|
return 1; |
576
|
|
|
|
|
|
|
} |
577
|
|
|
|
|
|
|
|
578
|
|
|
|
|
|
|
1; |