| line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
|
1
|
|
|
|
|
|
|
package FAST; |
|
2
|
1
|
|
|
1
|
|
14317
|
use 5.10.1; |
|
|
1
|
|
|
|
|
3
|
|
|
|
1
|
|
|
|
|
47
|
|
|
3
|
1
|
|
|
1
|
|
4
|
use strict; |
|
|
1
|
|
|
|
|
2
|
|
|
|
1
|
|
|
|
|
28
|
|
|
4
|
1
|
|
|
1
|
|
3
|
use Carp; |
|
|
1
|
|
|
|
|
6
|
|
|
|
1
|
|
|
|
|
67
|
|
|
5
|
1
|
|
|
1
|
|
4
|
use warnings FATAL => 'all'; |
|
|
1
|
|
|
|
|
1
|
|
|
|
1
|
|
|
|
|
234
|
|
|
6
|
|
|
|
|
|
|
|
|
7
|
|
|
|
|
|
|
|
|
8
|
|
|
|
|
|
|
our $VERSION = '1.06'; |
|
9
|
|
|
|
|
|
|
|
|
10
|
|
|
|
|
|
|
our $DEF_FORMAT = "fasta"; |
|
11
|
|
|
|
|
|
|
our $DEF_LOGNAME = "FAST.log.txt"; |
|
12
|
|
|
|
|
|
|
our $DEF_JOIN_STRING = " "; |
|
13
|
|
|
|
|
|
|
our $DEF_SPLIT_REGEX = ' '; |
|
14
|
|
|
|
|
|
|
|
|
15
|
|
|
|
|
|
|
sub log { |
|
16
|
0
|
|
|
0
|
0
|
|
my ($logname, $DATE, $COMMAND, $comment, $fromSTDIN) = @_; |
|
17
|
0
|
0
|
0
|
|
|
|
unless ($logname and $DATE and $COMMAND) { |
|
|
|
|
0
|
|
|
|
|
|
18
|
0
|
|
|
|
|
|
carp join "","FAST::log: logname, DATE, or COMMAND undefined; logging not possible.\n"; |
|
19
|
0
|
|
|
|
|
|
return; |
|
20
|
|
|
|
|
|
|
} |
|
21
|
0
|
0
|
|
|
|
|
open (LOG, ">>$logname") or carp "FAST::log: cannot open logfile $logname\n"; |
|
22
|
0
|
0
|
|
|
|
|
if (*LOG) { |
|
23
|
0
|
0
|
|
|
|
|
if (!$fromSTDIN) { |
|
24
|
0
|
|
|
|
|
|
print LOG "\n# FAST $VERSION $DATE "; |
|
25
|
0
|
0
|
|
|
|
|
print LOG "$comment" if ($comment); |
|
26
|
0
|
|
|
|
|
|
print LOG "\n"; |
|
27
|
0
|
|
|
|
|
|
print LOG "$COMMAND" |
|
28
|
|
|
|
|
|
|
} |
|
29
|
|
|
|
|
|
|
else { |
|
30
|
0
|
|
|
|
|
|
print LOG " | $COMMAND" |
|
31
|
|
|
|
|
|
|
} |
|
32
|
0
|
|
|
|
|
|
close LOG; |
|
33
|
|
|
|
|
|
|
} |
|
34
|
|
|
|
|
|
|
} |
|
35
|
|
|
|
|
|
|
|
|
36
|
|
|
|
|
|
|
|
|
37
|
|
|
|
|
|
|
=head1 NAME |
|
38
|
|
|
|
|
|
|
|
|
39
|
|
|
|
|
|
|
FAST - FAST Analysis of Sequences Toolbox |
|
40
|
|
|
|
|
|
|
|
|
41
|
|
|
|
|
|
|
=head1 VERSION |
|
42
|
|
|
|
|
|
|
|
|
43
|
|
|
|
|
|
|
Version 1.03 |
|
44
|
|
|
|
|
|
|
|
|
45
|
|
|
|
|
|
|
=head1 SYNOPSIS |
|
46
|
|
|
|
|
|
|
|
|
47
|
|
|
|
|
|
|
The Fast Analysis of Sequences Toolbox (FAST) is a set of UNIX |
|
48
|
|
|
|
|
|
|
utilities (for example fasgrep, fascut, fashead and fastr) that |
|
49
|
|
|
|
|
|
|
extends the UNIX toolbox paradigm to bioinformatic sequence |
|
50
|
|
|
|
|
|
|
records. Modeled after the UNIX textutils (such as grep, cut, head, |
|
51
|
|
|
|
|
|
|
tr, etc), FAST workflows are designed for "inline" (serial) processing |
|
52
|
|
|
|
|
|
|
of flatfile biological sequence record databases per-sequence, rather |
|
53
|
|
|
|
|
|
|
than per-line, through UNIX pipelines. The default data exchange |
|
54
|
|
|
|
|
|
|
format is multifasta (specifically, a restriction of BioPerl FastA |
|
55
|
|
|
|
|
|
|
format). FASTQ format is supported. FAST is designed for learnability, |
|
56
|
|
|
|
|
|
|
interoperability, interface consistency, rapid prototyping, fine-tuned |
|
57
|
|
|
|
|
|
|
control, and reproducibility. FAST tools expose the power of Perl and |
|
58
|
|
|
|
|
|
|
BioPerl to users in an easy-to-learn command-line paradigm. |
|
59
|
|
|
|
|
|
|
|
|
60
|
|
|
|
|
|
|
=head1 UTILITIES |
|
61
|
|
|
|
|
|
|
|
|
62
|
|
|
|
|
|
|
FAST 1.0x contains the following utilities. Each has its own man page. |
|
63
|
|
|
|
|
|
|
|
|
64
|
|
|
|
|
|
|
=over 4 |
|
65
|
|
|
|
|
|
|
|
|
66
|
|
|
|
|
|
|
=item B |
|
67
|
|
|
|
|
|
|
select sites based on site variation and gap-content |
|
68
|
|
|
|
|
|
|
|
|
69
|
|
|
|
|
|
|
=item B |
|
70
|
|
|
|
|
|
|
calculate molecular population genetic statistics |
|
71
|
|
|
|
|
|
|
|
|
72
|
|
|
|
|
|
|
=item B |
|
73
|
|
|
|
|
|
|
tally/annotate codon usage |
|
74
|
|
|
|
|
|
|
|
|
75
|
|
|
|
|
|
|
=item B |
|
76
|
|
|
|
|
|
|
tally/annotate monomer frequencies |
|
77
|
|
|
|
|
|
|
|
|
78
|
|
|
|
|
|
|
=item B |
|
79
|
|
|
|
|
|
|
convert sequences to or from from fasta format |
|
80
|
|
|
|
|
|
|
|
|
81
|
|
|
|
|
|
|
=item B |
|
82
|
|
|
|
|
|
|
select/reorder sequence records based on index lists/ranges |
|
83
|
|
|
|
|
|
|
|
|
84
|
|
|
|
|
|
|
=item B |
|
85
|
|
|
|
|
|
|
filter sequence records based on numerical values |
|
86
|
|
|
|
|
|
|
|
|
87
|
|
|
|
|
|
|
=item B |
|
88
|
|
|
|
|
|
|
filter sequence records based on perl regular expressions |
|
89
|
|
|
|
|
|
|
|
|
90
|
|
|
|
|
|
|
=item B |
|
91
|
|
|
|
|
|
|
filter leading sequence records |
|
92
|
|
|
|
|
|
|
|
|
93
|
|
|
|
|
|
|
=item B |
|
94
|
|
|
|
|
|
|
annotate sequence lengths |
|
95
|
|
|
|
|
|
|
|
|
96
|
|
|
|
|
|
|
=item B |
|
97
|
|
|
|
|
|
|
concatenate sequence records |
|
98
|
|
|
|
|
|
|
|
|
99
|
|
|
|
|
|
|
=item B |
|
100
|
|
|
|
|
|
|
reverse complement nucleotide sequences and alignments |
|
101
|
|
|
|
|
|
|
|
|
102
|
|
|
|
|
|
|
=item B |
|
103
|
|
|
|
|
|
|
sort sequence records |
|
104
|
|
|
|
|
|
|
|
|
105
|
|
|
|
|
|
|
=item B |
|
106
|
|
|
|
|
|
|
transform sequence records using regex-based substitutions |
|
107
|
|
|
|
|
|
|
|
|
108
|
|
|
|
|
|
|
=item B |
|
109
|
|
|
|
|
|
|
filter trailing sequence records |
|
110
|
|
|
|
|
|
|
|
|
111
|
|
|
|
|
|
|
=item B |
|
112
|
|
|
|
|
|
|
filter sequence records based on NCBI taxonomy IDs or names |
|
113
|
|
|
|
|
|
|
|
|
114
|
|
|
|
|
|
|
=item B |
|
115
|
|
|
|
|
|
|
sort sequence records based on NCBI taxonomy IDs or names |
|
116
|
|
|
|
|
|
|
|
|
117
|
|
|
|
|
|
|
=item B |
|
118
|
|
|
|
|
|
|
transform sequence records by character, degap, strict |
|
119
|
|
|
|
|
|
|
|
|
120
|
|
|
|
|
|
|
=item B |
|
121
|
|
|
|
|
|
|
remove duplicate sequence records from sorted data |
|
122
|
|
|
|
|
|
|
|
|
123
|
|
|
|
|
|
|
=item B |
|
124
|
|
|
|
|
|
|
tally sequences and characters |
|
125
|
|
|
|
|
|
|
|
|
126
|
|
|
|
|
|
|
=item B |
|
127
|
|
|
|
|
|
|
translate gapped and ungapped sequences and alignments |
|
128
|
|
|
|
|
|
|
|
|
129
|
|
|
|
|
|
|
=item B |
|
130
|
|
|
|
|
|
|
select sites by regex matching of features in a GenBank file |
|
131
|
|
|
|
|
|
|
|
|
132
|
|
|
|
|
|
|
=item B |
|
133
|
|
|
|
|
|
|
print sequences by regex matching to features in a GenBank file |
|
134
|
|
|
|
|
|
|
|
|
135
|
|
|
|
|
|
|
=back |
|
136
|
|
|
|
|
|
|
|
|
137
|
|
|
|
|
|
|
=head1 AUTHORS |
|
138
|
|
|
|
|
|
|
|
|
139
|
|
|
|
|
|
|
David H. Ardell, C<< >> |
|
140
|
|
|
|
|
|
|
and members of the Ardell Laboratory |
|
141
|
|
|
|
|
|
|
and other contributors including: |
|
142
|
|
|
|
|
|
|
Travis Lawrence |
|
143
|
|
|
|
|
|
|
Dana Carper |
|
144
|
|
|
|
|
|
|
Katherine Amrine |
|
145
|
|
|
|
|
|
|
Kyle Kauffman |
|
146
|
|
|
|
|
|
|
Claudia Canales |
|
147
|
|
|
|
|
|
|
|
|
148
|
|
|
|
|
|
|
=head1 BUGS |
|
149
|
|
|
|
|
|
|
|
|
150
|
|
|
|
|
|
|
Please report any bugs or feature requests to C, or through |
|
151
|
|
|
|
|
|
|
the web interface at L. I will be notified, and then you'll |
|
152
|
|
|
|
|
|
|
automatically be notified of progress on your bug as I make changes. |
|
153
|
|
|
|
|
|
|
|
|
154
|
|
|
|
|
|
|
=head1 ACKNOWLEDGEMENTS |
|
155
|
|
|
|
|
|
|
|
|
156
|
|
|
|
|
|
|
|
|
157
|
|
|
|
|
|
|
=head1 LICENSE AND COPYRIGHT |
|
158
|
|
|
|
|
|
|
|
|
159
|
|
|
|
|
|
|
Copyright 2015 David H. Ardell. |
|
160
|
|
|
|
|
|
|
|
|
161
|
|
|
|
|
|
|
This program is free software; you can redistribute it and/or modify it |
|
162
|
|
|
|
|
|
|
under the terms of the the Artistic License (2.0). You may obtain a |
|
163
|
|
|
|
|
|
|
copy of the full license at: |
|
164
|
|
|
|
|
|
|
|
|
165
|
|
|
|
|
|
|
L |
|
166
|
|
|
|
|
|
|
|
|
167
|
|
|
|
|
|
|
Any use, modification, and distribution of the Standard or Modified |
|
168
|
|
|
|
|
|
|
Versions is governed by this Artistic License. By using, modifying or |
|
169
|
|
|
|
|
|
|
distributing the Package, you accept this license. Do not use, modify, |
|
170
|
|
|
|
|
|
|
or distribute the Package, if you do not accept this license. |
|
171
|
|
|
|
|
|
|
|
|
172
|
|
|
|
|
|
|
If your Modified Version has been derived from a Modified Version made |
|
173
|
|
|
|
|
|
|
by someone other than you, you are nevertheless required to ensure that |
|
174
|
|
|
|
|
|
|
your Modified Version complies with the requirements of this license. |
|
175
|
|
|
|
|
|
|
|
|
176
|
|
|
|
|
|
|
This license does not grant you the right to use any trademark, service |
|
177
|
|
|
|
|
|
|
mark, tradename, or logo of the Copyright Holder. |
|
178
|
|
|
|
|
|
|
|
|
179
|
|
|
|
|
|
|
This license includes the non-exclusive, worldwide, free-of-charge |
|
180
|
|
|
|
|
|
|
patent license to make, have made, use, offer to sell, sell, import and |
|
181
|
|
|
|
|
|
|
otherwise transfer the Package with respect to any patent claims |
|
182
|
|
|
|
|
|
|
licensable by the Copyright Holder that are necessarily infringed by the |
|
183
|
|
|
|
|
|
|
Package. If you institute patent litigation (including a cross-claim or |
|
184
|
|
|
|
|
|
|
counterclaim) against any party alleging that the Package constitutes |
|
185
|
|
|
|
|
|
|
direct or contributory patent infringement, then this Artistic License |
|
186
|
|
|
|
|
|
|
to you shall terminate on the date that such litigation is filed. |
|
187
|
|
|
|
|
|
|
|
|
188
|
|
|
|
|
|
|
Disclaimer of Warranty: THE PACKAGE IS PROVIDED BY THE COPYRIGHT HOLDER |
|
189
|
|
|
|
|
|
|
AND CONTRIBUTORS "AS IS' AND WITHOUT ANY EXPRESS OR IMPLIED WARRANTIES. |
|
190
|
|
|
|
|
|
|
THE IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR |
|
191
|
|
|
|
|
|
|
PURPOSE, OR NON-INFRINGEMENT ARE DISCLAIMED TO THE EXTENT PERMITTED BY |
|
192
|
|
|
|
|
|
|
YOUR LOCAL LAW. UNLESS REQUIRED BY LAW, NO COPYRIGHT HOLDER OR |
|
193
|
|
|
|
|
|
|
CONTRIBUTOR WILL BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, OR |
|
194
|
|
|
|
|
|
|
CONSEQUENTIAL DAMAGES ARISING IN ANY WAY OUT OF THE USE OF THE PACKAGE, |
|
195
|
|
|
|
|
|
|
EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. |
|
196
|
|
|
|
|
|
|
|
|
197
|
|
|
|
|
|
|
|
|
198
|
|
|
|
|
|
|
=cut |
|
199
|
|
|
|
|
|
|
|
|
200
|
|
|
|
|
|
|
1; # End of FAST |