add bin files for search routines

[function2gene.git] / bin / combine_results

diff --git a/bin/combine_results b/bin/combine_results
new file mode 100755 (executable)
index 0000000..8125c11
--- /dev/null
+++ b/bin/combine_results
@@ -0,0 +1,172 @@
+#! /usr/bin/perl
+
+# parse_ncbi_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_ncbi_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_ncbi_results -D ./ncbi_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 XML::Parser::Expat;
+use IO::File;
+
+# XXX parse config file
+
+my %options = (debug    => 0,
+              help     => 0,
+              man      => 0,
+              dir      => '.',
+              keyword  => undef,
+             );
+
+GetOptions(\%options,'keyword|k=s','debug|d+','help|h|?','man|m');
+
+
+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,
+            };
+
+my @csv_fields = qw(name hits rzscore refseq location alias database terms description function);
+
+my %genes;
+
+for my $file_name (@ARGV) {
+     my $file = new IO::File $file_name, 'r' or die "Unable to open file $file_name $!";
+     while (<$file>) {
+         next if /^"Name"/;
+         my @gene = map {s/^\"//; s/\"$//; $_;} split /(?<=\")\,(?=\")/, $_;
+         $genes{$gene[NAME]}{name} = $gene[NAME];
+         $genes{$gene[NAME]}{database}{$gene[DBNAME]}++;
+         $genes{$gene[NAME]}{hits}++;
+         $genes{$gene[NAME]}{terms}{$gene[KEYWORD]}++;
+         add_unique_parts($genes{$gene[NAME]},'refseq',$gene[REFSEQ]);
+         add_if_better($genes{$gene[NAME]},'description',$gene[DESCRIPTION]);
+         add_if_better($genes{$gene[NAME]},'location',$gene[LOCATION]);
+         add_unique_parts($genes{$gene[NAME]},'function',split(/; /, $gene[FUNCTION]));
+         add_unique_parts($genes{$gene[NAME]},'alias', split(/; /, $gene[ALIAS]));
+     }
+}
+
+print join(',',map {qq("$_")} @csv_fields),qq(\n);
+for my $gene (keys %genes) {
+     $genes{$gene}{rzscore} = scalar keys %{$genes{$gene}{terms}};
+     next if $genes{$gene}{rzscore} == 1 and exists $genes{$gene}{terms}{antigen};
+     $genes{$gene}{rzscore} -= 1 if exists $genes{$gene}{terms}{antigen};
+     print STDOUT join (',',
+                       map {s/"//g; qq("$_")}
+                       map {
+                            my $value = $_;
+                            if (ref $value eq 'HASH') {
+                                 join('; ',map {qq($_:$$value{$_})} keys %$value);
+                            }
+                            elsif (ref $value eq 'ARRAY') {
+                                 join('; ', @$value);
+                            }
+                            else {
+                                 $value;
+                            }
+                       } @{$genes{$gene}}{@csv_fields}
+                      ), qq(\n);
+}
+
+
+sub add_unique_parts{
+     my ($hr,$key,@values) = @_;
+     if (not defined $$hr{key}) {
+         $$hr{$key} = [@values];
+     }
+     else {
+         my %temp_hash;
+         @temp_hash{@{$$hr{$key}}} = (1) x scalar @{$$hr{$key}};
+         $temp_hash{@values} = (1) x scalar @values;
+         $$hr{$key} = [keys %temp_hash];
+     }
+}
+
+sub add_if_better{
+     my ($hash, $key, $value) = @_;
+     if (not defined $$hash{$key}) {
+         $$hash{$key} = $value;
+     }
+     elsif (length $$hash{$key} < length $value and $value !~ /^NO\s+(LOCATION|DESCRIPTION)+$/) {
+         $$hash{$key} = $value;
+     }
+}
+
+
+
+
+__END__