| 1 |
#!/usr/bin/perl |
| 2 |
|
| 3 |
use strict; |
| 4 |
use warnings; |
| 5 |
|
| 6 |
## Used to output the 'usage' message |
| 7 |
use Pod::Usage; |
| 8 |
|
| 9 |
## Used to parse command line options |
| 10 |
use Getopt::Long; |
| 11 |
|
| 12 |
## Used to create temporary files, if necessary |
| 13 |
use File::Spec; |
| 14 |
|
| 15 |
## BioPerl! |
| 16 |
use Bio::DB::SeqFeature::Store; |
| 17 |
use Bio::DB::SeqFeature::Store::GFF3Loader; |
| 18 |
|
| 19 |
|
| 20 |
|
| 21 |
## The available options. Note, these defaults are 'hard coded' into |
| 22 |
## the USAGE POD, so if you change one of the defaults (you shouldn't), |
| 23 |
## you should update the USAGE. |
| 24 |
|
| 25 |
my $DSN = 'dbi:mysql:test'; |
| 26 |
my $SFCLASS = 'Bio::DB::SeqFeature'; |
| 27 |
my $ADAPTOR = 'DBI::mysql'; |
| 28 |
my $NAMESPACE; |
| 29 |
my $VERBOSE = 1; |
| 30 |
my $FAST = 0; |
| 31 |
my $TMP = File::Spec->tmpdir(); |
| 32 |
my $IGNORE_SEQREGION = 0; |
| 33 |
my $CREATE = 0; |
| 34 |
my $USER = ''; |
| 35 |
my $PASS = ''; |
| 36 |
my $COMPRESS = 0; |
| 37 |
my $INDEX_SUB = 1; |
| 38 |
my $NOALIAS_TARGET = 0; |
| 39 |
my $SUMMARY_STATS = 0; |
| 40 |
my $NOSUMMARY_STATS = 0; |
| 41 |
my $FTS = 0; |
| 42 |
|
| 43 |
## Two flags based on http://stackoverflow.com/questions/1232116 |
| 44 |
## how-to-create-pod-and-use-pod2usage-in-perl |
| 45 |
my $opt_help; |
| 46 |
my $opt_man; |
| 47 |
|
| 48 |
GetOptions( 'd|dsn=s' => \$DSN, |
| 49 |
's|seqfeature=s' => \$SFCLASS, |
| 50 |
'n|namespace=s' => \$NAMESPACE, |
| 51 |
'a|adaptor=s' => \$ADAPTOR, |
| 52 |
'v|verbose!' => \$VERBOSE, |
| 53 |
'f|fast' => \$FAST, |
| 54 |
'T|temporary-directory=s' => \$TMP, |
| 55 |
'i|ignore-seqregion' => \$IGNORE_SEQREGION, |
| 56 |
'c|create' => \$CREATE, |
| 57 |
'u|user=s' => \$USER, |
| 58 |
'p|password=s' => \$PASS, |
| 59 |
'z|zip' => \$COMPRESS, |
| 60 |
'S|subfeatures!' => \$INDEX_SUB, |
| 61 |
|
| 62 |
## Any good single letter choices here? |
| 63 |
'noalias-target' => \$NOALIAS_TARGET, |
| 64 |
'summary' => \$SUMMARY_STATS, |
| 65 |
'N|nosummary' => \$NOSUMMARY_STATS, |
| 66 |
'fts' => \$FTS, |
| 67 |
|
| 68 |
## I miss '--help' when it isn't there! |
| 69 |
'h|help!' => \$opt_help, |
| 70 |
'm|man!' => \$opt_man, |
| 71 |
) |
| 72 |
or pod2usage( -message => |
| 73 |
"\nTry 'bp_seqfeature_load.pl --help' for more information\n", |
| 74 |
-verbose => 0, |
| 75 |
-exitval => 2, |
| 76 |
); |
| 77 |
|
| 78 |
## Should we output usage information? |
| 79 |
pod2usage( -verbose => 1 ) if $opt_help; |
| 80 |
pod2usage( -verbose => 2 ) if $opt_man; |
| 81 |
|
| 82 |
## Did we get any files to process? |
| 83 |
@ARGV |
| 84 |
or pod2usage( -message => |
| 85 |
"\nYou need to pass some GFF or fasta files to load\n", |
| 86 |
-verbose => 0, |
| 87 |
-exitval => 2, |
| 88 |
); |
| 89 |
|
| 90 |
pod2usage( -message => "\n--fts requires --create\n", |
| 91 |
-verbose => 0, |
| 92 |
-exitval => 2, |
| 93 |
) if ($FTS and not $CREATE); |
| 94 |
|
| 95 |
|
| 96 |
## POD |
| 97 |
|
| 98 |
=head1 NAME |
| 99 |
|
| 100 |
bp_seqfeature_load.pl - Load GFF into a SeqFeature database |
| 101 |
|
| 102 |
=head1 DESCRIPTION |
| 103 |
|
| 104 |
Pass any number of GFF or fasta format files (or GFF with embedded |
| 105 |
fasta) to load the features and sequences into a SeqFeature |
| 106 |
database. The database (and adaptor) to use is specified on the |
| 107 |
command line. Use the --create flag to create a new SeqFeature |
| 108 |
database. |
| 109 |
|
| 110 |
=head1 SYNOPSIS |
| 111 |
|
| 112 |
bp_seqfeature_load.pl [options] gff_or_fasta_file1 [gff_or_fasta_file2 [...]] |
| 113 |
|
| 114 |
Try 'bp_seqfeature_load.pl --help' or '--man' for more information. |
| 115 |
|
| 116 |
=head1 OPTIONS |
| 117 |
|
| 118 |
=over 4 |
| 119 |
|
| 120 |
=item -d, --dsn |
| 121 |
|
| 122 |
DBI data source (default dbi:mysql:test) |
| 123 |
|
| 124 |
=item -n, --namespace |
| 125 |
|
| 126 |
The table prefix to use (default undef) Allows several independent |
| 127 |
sequence feature databases to be stored in a single database |
| 128 |
|
| 129 |
=item -s, --seqfeature |
| 130 |
|
| 131 |
The type of SeqFeature to create... RTSC (default Bio::DB::SeqFeature) |
| 132 |
|
| 133 |
=item -a, --adaptor |
| 134 |
|
| 135 |
The storage adaptor (class) to use (default DBI::mysql) |
| 136 |
|
| 137 |
=item -v, --verbose |
| 138 |
|
| 139 |
Turn on verbose progress reporting (default true) Use --noverbose to |
| 140 |
switch this off. |
| 141 |
|
| 142 |
=item -f, --fast |
| 143 |
|
| 144 |
Activate fast loading. (default 0) Only available for some adaptors. |
| 145 |
|
| 146 |
=item -T, --temporary-directory |
| 147 |
|
| 148 |
Specify temporary directory for fast loading (default |
| 149 |
File::Spec->tmpdir()) |
| 150 |
|
| 151 |
=item -i, --ignore-seqregion |
| 152 |
|
| 153 |
If true, then ignore ##sequence-region directives in the GFF3 file |
| 154 |
(default, create a feature for each region) |
| 155 |
|
| 156 |
=item -c, --create |
| 157 |
|
| 158 |
Create the database and reinitialize it (default false) Note, this |
| 159 |
will erase previous database contents, if any. |
| 160 |
|
| 161 |
=item -u, --user |
| 162 |
|
| 163 |
User to connect to database as |
| 164 |
|
| 165 |
=item -p, --password |
| 166 |
|
| 167 |
Password to use to connect to database |
| 168 |
|
| 169 |
=item -z, --zip |
| 170 |
|
| 171 |
Compress database tables to save space (default false) |
| 172 |
|
| 173 |
=item -S, --subfeatures |
| 174 |
|
| 175 |
Turn on indexing of subfeatures (default true) Use --nosubfeatures to |
| 176 |
switch this off. |
| 177 |
|
| 178 |
=item --fts |
| 179 |
|
| 180 |
Index the attribute table for full-text search (default false). Applicable |
| 181 |
only when --create is specified. Currently applicable to the DBI::SQLite |
| 182 |
storage adaptor only (using the most recent supported FTS indexing method, |
| 183 |
which may not be portable to older DBI::SQLite versions). |
| 184 |
|
| 185 |
=item --summary |
| 186 |
|
| 187 |
Generate summary statistics for coverage graphs (default false) This |
| 188 |
can be run on a previously loaded database or during the load. It will |
| 189 |
default to true if --create is used. |
| 190 |
|
| 191 |
=item -N, --nosummary |
| 192 |
|
| 193 |
Do not generate summary statistics to save some space and load time (default if |
| 194 |
--create is not specified, use this option to explicitly turn off summary |
| 195 |
statistics when --create is specified) |
| 196 |
|
| 197 |
=item --noalias-target |
| 198 |
|
| 199 |
Don't create an Alias attribute whose value is the target_id in a |
| 200 |
Target attribute (if the feature contains a Target attribute, the |
| 201 |
default is to create an Alias attribute whose value is the target_id |
| 202 |
in the Target attribute) |
| 203 |
|
| 204 |
=back |
| 205 |
|
| 206 |
Please see http://www.sequenceontology.org/gff3.shtml for information |
| 207 |
about the GFF3 format. BioPerl extends the format slightly by adding a |
| 208 |
##index-subfeatures directive. Set this to a true value if you wish |
| 209 |
the database to be able to retrieve a feature's individual parts (such |
| 210 |
as the exons of a transcript) independently of the top level feature: |
| 211 |
|
| 212 |
##index-subfeatures 1 |
| 213 |
|
| 214 |
It is also possible to control the indexing of subfeatures on a |
| 215 |
case-by-case basis by adding "index=1" or "index=0" to the feature's |
| 216 |
attribute list. This should only be used for subfeatures. |
| 217 |
|
| 218 |
Subfeature indexing is true by default. Set to false (0) to save lots |
| 219 |
of database space and speed performance. You may use --nosubfeatures |
| 220 |
to force this. |
| 221 |
|
| 222 |
=cut |
| 223 |
|
| 224 |
|
| 225 |
|
| 226 |
|
| 227 |
|
| 228 |
if ($FAST) { |
| 229 |
-d $TMP && -w $TMP |
| 230 |
or die "Fast loading is requested, but I cannot write into the directory $TMP"; |
| 231 |
$DSN .= ";mysql_local_infile=1" if $ADAPTOR =~ /mysql/i && $DSN !~ /mysql_local_infile/; |
| 232 |
} |
| 233 |
|
| 234 |
my @options; |
| 235 |
@options = ($USER,$PASS) if $USER || $PASS; |
| 236 |
|
| 237 |
my $store = Bio::DB::SeqFeature::Store->new |
| 238 |
( |
| 239 |
-dsn => $DSN, |
| 240 |
-namespace => $NAMESPACE, |
| 241 |
-adaptor => $ADAPTOR, |
| 242 |
-tmpdir => $TMP, |
| 243 |
-user => $USER, |
| 244 |
-pass => $PASS, |
| 245 |
-write => 1, |
| 246 |
-create => $CREATE, |
| 247 |
-compress => $COMPRESS, |
| 248 |
-fts => $FTS, |
| 249 |
) |
| 250 |
or die "Couldn't create connection to the database"; |
| 251 |
|
| 252 |
$store->init_database('erase') if $CREATE; |
| 253 |
$SUMMARY_STATS++ if $CREATE; # this is a good thing |
| 254 |
|
| 255 |
my $loader = Bio::DB::SeqFeature::Store::GFF3Loader->new |
| 256 |
( |
| 257 |
-store => $store, |
| 258 |
-sf_class => $SFCLASS, |
| 259 |
-verbose => $VERBOSE, |
| 260 |
-tmpdir => $TMP, |
| 261 |
-fast => $FAST, |
| 262 |
-ignore_seqregion => $IGNORE_SEQREGION, |
| 263 |
-index_subfeatures => $INDEX_SUB, |
| 264 |
-noalias_target => $NOALIAS_TARGET, |
| 265 |
-summary_stats => $NOSUMMARY_STATS ? 0 : $SUMMARY_STATS, |
| 266 |
) |
| 267 |
or die "Couldn't create GFF3 loader"; |
| 268 |
|
| 269 |
# on signals, give objects a chance to call their DESTROY methods |
| 270 |
$SIG{TERM} = $SIG{INT} = sub { undef $loader; undef $store; die "Aborted..."; }; |
| 271 |
|
| 272 |
$loader->load(@ARGV); |
| 273 |
|
| 274 |
exit 0; |