add parse ensembl results

git-svn-id: file:///srv/svn/function2gene/trunk@30 a0738b58-4706-0410-8799-fb830574a030

diff --git a/bin/parse_ensembl_results b/bin/parse_ensembl_results
new file mode 100755 (executable)
index 0000000..cdd1fba
--- /dev/null
+++ b/bin/parse_ensembl_results
@@ -0,0 +1,199 @@
+#! /usr/bin/perl
+
+# parse_genecard_results retreives files of search results from ncbi,
+# and is released under the terms of the GPL version 2, or any later
+# version, at your option. See the file README and COPYING for more
+# information.
+
+# Copyright 2004 by Don Armstrong <don@donarmstrong.com>.
+
+# $Id: ss,v 1.1 2004/06/29 05:26:35 don Exp $
+
+
+use warnings;
+use strict;
+
+
+use Getopt::Long;
+use Pod::Usage;
+
+=head1 NAME
+
+  parse_genecard_results [options]
+
+=head1 SYNOPSIS
+
+
+ Options:
+  --dir, -D directory to stick results into [default .]
+  --name, -n file naming scheme [default ${search}_results.$format]
+  --terms, -t file of search terms [default -]
+  --debug, -d debugging level [default 0]
+  --help, -h display this help
+  --man, -m display manual
+
+=head1 OPTIONS
+
+=over
+
+=item B<--debug, -d>
+
+Debug verbosity. (Default 0)
+
+=item B<--help, -h>
+
+Display brief useage information.
+
+=item B<--man, -m>
+
+Display this manual.
+
+=back
+
+=head1 EXAMPLES
+
+  parse_harvester_results -D ./harvester_results/ -n '${search}_name.html' < search_parameters
+
+Will pretty much do what you want
+
+=cut
+
+
+
+use vars qw($DEBUG $REVISION);
+
+BEGIN{
+     ($REVISION) = q$LastChangedRevision: 1$ =~ /LastChangedRevision:\s+([^\s]+)/;
+     $DEBUG = 0 unless defined $DEBUG;
+}
+
+use IO::File;
+use IO::Dir;
+
+use HTML::TreeBuilder;
+use HTML::ElementTable;
+
+my %options = (debug    => 0,
+              help     => 0,
+              man      => 0,
+              dir      => '.',
+              keyword  => undef,
+              keywords => 0,
+             );
+
+GetOptions(\%options,'keyword|k=s','dir|D=s','debug|d+','help|h|?','man|m',
+          'keywords',
+         );
+
+
+pod2usage() if $options{help};
+pod2usage({verbose=>2}) if $options{man};
+
+$DEBUG = $options{debug};
+
+# CSV columns
+use constant {NAME        => 0,
+             REFSEQ      => 1,
+             LOCATION    => 2,
+             ALIAS       => 3,
+             FUNCTION    => 4,
+             DESCRIPTION => 5,
+             KEYWORD     => 6,
+             DBNAME      => 7,
+             FILENAME    => 8,
+            };
+
+if ($options{keywords}) {
+     if (@ARGV != 1) {
+         pod2usage("If the --keywords option is used, exactly one argument (the keyword) must be passed");
+     }
+     $options{dir} = "$ARGV[0]_results_genecard";
+}
+
+if (not -d $options{dir}) {
+     die "$options{dir} does not exist or is not a directory";
+}
+
+my $dir = new IO::Dir $options{dir} or die "Unable to open dir $options{dir}: $!";
+
+print join(",", map {qq("$_");} qw(Name RefSeq Location Alias Function Description Keyword DBName Filename)),qq(\n);
+
+my ($keyword) = $options{keyword} || $options{dir} =~ m#(?:^|/)([^\/]+)_results_genecard#;
+
+while ($_ = $dir->read) {
+     my $file_name = $_;
+     next if $file_name =~ /^\./;
+     next unless -f "$options{dir}/$file_name" and -r "$options{dir}/$file_name";
+
+     my $file = new IO::File "$options{dir}/$file_name", 'r' or die "Unable to open file $file_name";
+
+     local $/;
+     my $result = <$file>;
+     next if $result =~ m/is not present in the current release/;
+
+     my @results;
+
+     # Find gene name
+     ($results[NAME]) = map {s/^[^:]+://; $_;}$result =~ m{a\s+href=\"[^"]+genenames.org[^"]+">\s*([^<]+?)\s*</a>}xis;
+
+     $results[NAME] ||= 'NO NAME';
+     # Find REF SEQ number
+     ($results[REFSEQ]) = $result =~ m{for\s*(ENSG\d+)}xis;
+
+     $results[REFSEQ] ||= 'NO REFSEQ';
+
+     # Find Gene Location
+     my @location = $result =~ m{on\s+Chromosome\s+([\dX]+)\s+at\s+location\s+<[^>]+>([\d,]+)}xis;
+     if (@location) {
+         $results[LOCATION] = "$location[0] $location[1]";
+     }
+     $results[LOCATION] ||= 'NO LOCATION';
+
+     # Find gene aliases
+     my @gene_aliases = map {split/,\s+/;} $result =~ m{<b>Synonyms:\s+</b>\s*([^<]+)\s*</td>}gxis;
+
+     $results[ALIAS] = join('; ', @gene_aliases);
+     $results[ALIAS] ||= 'NO ALIASES';
+
+     # Find gene function(s)
+#
+#      my @functions;
+#      # GO Functions
+#      push @functions, (map {s/\n//g; $_;}
+#                     map {s#\s*</a>(?:</td><td>\s*)?\s*# #g; $_;}
+#                     $result =~ m{(GO:\d+\s*</a>(?:</td><td>\s*)?.+?)(?:</font>|</td>|<dd>|<p>)}gis
+#                    );
+#      $results[FUNCTION] = join('; ', map {(defined $_)?($_):()} @functions);
+     $results[FUNCTION] ||= 'NO FUNCTION';
+
+     # Figure out the keyword used
+     $results[KEYWORD] ||= $keyword || 'NO KEYWORD';
+
+     my @description = (map {s/\n/ /g;
+                            s/\s+/ /g;
+                            $_;
+                       }
+                       $result =~ m{<th\s*class="two-column">\s*
+                                    Description\s*
+                                    </th>\s*
+                                    <td\s*class="two-column">\s*
+                                    <p>\s*([^<]+)}xgis
+                      );
+     # Figure out what the description is
+     $results[DESCRIPTION] = join('; ',
+                                 map {(defined $_)?($_):()}
+                                 @description);
+
+     # Database searched
+     $results[DBNAME] = 'ensembl';
+     $results[FILENAME] = $file_name;
+
+     print join(',',map {qq("$_")} @results),qq(\n);
+}
+
+
+
+
+
+
+__END__