[Bio] / Sprout / SproutLoad.pm Repository:
ViewVC logotype

Diff of /Sprout/SproutLoad.pm

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 1.9, Wed Sep 14 11:21:24 2005 UTC revision 1.93, Tue Sep 9 21:02:10 2008 UTC
# Line 7  Line 7 
7      use PageBuilder;      use PageBuilder;
8      use ERDBLoad;      use ERDBLoad;
9      use FIG;      use FIG;
10        use FIGRules;
11      use Sprout;      use Sprout;
12      use Stats;      use Stats;
13      use BasicLocation;      use BasicLocation;
14        use HTML;
15        use AliasAnalysis;
16        use BioWords;
17    
18  =head1 Sprout Load Methods  =head1 Sprout Load Methods
19    
# Line 29  Line 33 
33      $stats->Accumulate($spl->LoadFeatureData());      $stats->Accumulate($spl->LoadFeatureData());
34      print $stats->Show();      print $stats->Show();
35    
 This module makes use of the internal Sprout property C<_erdb>.  
   
36  It is worth noting that the FIG object does not need to be a real one. Any object  It is worth noting that the FIG object does not need to be a real one. Any object
37  that implements the FIG methods for data retrieval could be used. So, for example,  that implements the FIG methods for data retrieval could be used. So, for example,
38  this object could be used to copy data from one Sprout database to another, or  this object could be used to copy data from one Sprout database to another, or
# Line 51  Line 53 
53    
54  =head3 new  =head3 new
55    
56  C<< my $spl = SproutLoad->new($sprout, $fig, $genomeFile, $subsysFile, $options); >>      my $spl = SproutLoad->new($sprout, $fig, $genomeFile, $subsysFile, $options);
57    
58  Construct a new Sprout Loader object, specifying the two participating databases and  Construct a new Sprout Loader object, specifying the two participating databases and
59  the name of the files containing the list of genomes and subsystems to use.  the name of the files containing the list of genomes and subsystems to use.
# Line 79  Line 81 
81  =item subsysFile  =item subsysFile
82    
83  Either the name of the file containing the list of trusted subsystems or a reference  Either the name of the file containing the list of trusted subsystems or a reference
84  to a list of subsystem names. If nothing is specified, all known subsystems will be  to a list of subsystem names. If nothing is specified, all NMPDR subsystems will be
85  considered trusted. Only subsystem data related to the trusted subsystems is loaded.  considered trusted. (A subsystem is considered NMPDR if it has a file named C<NMPDR>
86    in its data directory.) Only subsystem data related to the NMPDR subsystems is loaded.
87    
88  =item options  =item options
89    
# Line 93  Line 96 
96  sub new {  sub new {
97      # Get the parameters.      # Get the parameters.
98      my ($class, $sprout, $fig, $genomeFile, $subsysFile, $options) = @_;      my ($class, $sprout, $fig, $genomeFile, $subsysFile, $options) = @_;
99      # Load the list of genomes into a hash.      # Create the genome hash.
100      my %genomes;      my %genomes = ();
101        # We only need it if load-only is NOT specified.
102        if (! $options->{loadOnly}) {
103      if (! defined($genomeFile) || $genomeFile eq '') {      if (! defined($genomeFile) || $genomeFile eq '') {
104          # Here we want all the complete genomes and an access code of 1.          # Here we want all the complete genomes and an access code of 1.
105          my @genomeList = $fig->genomes(1);          my @genomeList = $fig->genomes(1);
106          %genomes = map { $_ => 1 } @genomeList;          %genomes = map { $_ => 1 } @genomeList;
107                Trace(scalar(keys %genomes) . " genomes found.") if T(3);
108      } else {      } else {
109          my $type = ref $genomeFile;          my $type = ref $genomeFile;
110          Trace("Genome file parameter type is \"$type\".") if T(3);          Trace("Genome file parameter type is \"$type\".") if T(3);
# Line 118  Line 124 
124                  # an omitted access code can be defaulted to 1.                  # an omitted access code can be defaulted to 1.
125                  for my $genomeLine (@genomeList) {                  for my $genomeLine (@genomeList) {
126                      my ($genomeID, $accessCode) = split("\t", $genomeLine);                      my ($genomeID, $accessCode) = split("\t", $genomeLine);
127                      if (undef $accessCode) {                          if (! defined($accessCode)) {
128                          $accessCode = 1;                          $accessCode = 1;
129                      }                      }
130                      $genomes{$genomeID} = $accessCode;                      $genomes{$genomeID} = $accessCode;
# Line 128  Line 134 
134              Confess("Invalid genome parameter ($type) in SproutLoad constructor.");              Confess("Invalid genome parameter ($type) in SproutLoad constructor.");
135          }          }
136      }      }
137        }
138      # Load the list of trusted subsystems.      # Load the list of trusted subsystems.
139      my %subsystems = ();      my %subsystems = ();
140        # We only need it if load-only is NOT specified.
141        if (! $options->{loadOnly}) {
142      if (! defined $subsysFile || $subsysFile eq '') {      if (! defined $subsysFile || $subsysFile eq '') {
143          # Here we want all the subsystems.              # Here we want all the usable subsystems. First we get the whole list.
144          %subsystems = map { $_ => 1 } $fig->all_subsystems();              my @subs = $fig->all_subsystems();
145                # Loop through, checking for the NMPDR file.
146                for my $sub (@subs) {
147                    if ($fig->nmpdr_subsystem($sub)) {
148                        $subsystems{$sub} = 1;
149                    }
150                }
151      } else {      } else {
152          my $type = ref $subsysFile;          my $type = ref $subsysFile;
153          if ($type eq 'ARRAY') {          if ($type eq 'ARRAY') {
# Line 152  Line 167 
167              Confess("Invalid subsystem parameter in SproutLoad constructor.");              Confess("Invalid subsystem parameter in SproutLoad constructor.");
168          }          }
169      }      }
170            # Go through the subsys hash again, creating the keyword list for each subsystem.
171            for my $subsystem (keys %subsystems) {
172                my $name = $subsystem;
173                $name =~ s/_/ /g;
174                $subsystems{$subsystem} = $name;
175            }
176        }
177        # Get the list of NMPDR-oriented attribute keys.
178        my @propKeys = $fig->get_group_keys("NMPDR");
179      # Get the data directory from the Sprout object.      # Get the data directory from the Sprout object.
180      my ($directory) = $sprout->LoadInfo();      my ($directory) = $sprout->LoadInfo();
181      # Create the Sprout load object.      # Create the Sprout load object.
# Line 161  Line 185 
185                    subsystems => \%subsystems,                    subsystems => \%subsystems,
186                    sprout => $sprout,                    sprout => $sprout,
187                    loadDirectory => $directory,                    loadDirectory => $directory,
188                    erdb => $sprout->{_erdb},                    erdb => $sprout,
189                    loaders => [],                    loaders => [],
190                    options => $options                    options => $options,
191                      propKeys => \@propKeys,
192                   };                   };
193      # Bless and return it.      # Bless and return it.
194      bless $retVal, $class;      bless $retVal, $class;
195      return $retVal;      return $retVal;
196  }  }
197    
198    =head3 LoadOnly
199    
200        my $flag = $spl->LoadOnly;
201    
202    Return TRUE if we are in load-only mode, else FALSE.
203    
204    =cut
205    
206    sub LoadOnly {
207        my ($self) = @_;
208        return $self->{options}->{loadOnly};
209    }
210    
211    
212  =head3 LoadGenomeData  =head3 LoadGenomeData
213    
214  C<< my $stats = $spl->LoadGenomeData(); >>      my $stats = $spl->LoadGenomeData();
215    
216  Load the Genome, Contig, and Sequence data from FIG into Sprout.  Load the Genome, Contig, and Sequence data from FIG into Sprout.
217    
# Line 197  Line 236 
236    
237  =back  =back
238    
 B<TO DO>  
   
 Real quality vectors instead of C<unknown> for everything.  
   
 GenomeGroup relation. (The original script took group information from the C<NMPDR> file  
 in each genome's main directory, but no such file exists anywhere in my version of the  
 data store.)  
   
239  =cut  =cut
240  #: Return Type $%;  #: Return Type $%;
241  sub LoadGenomeData {  sub LoadGenomeData {
# Line 215  Line 246 
246      # Get the genome count.      # Get the genome count.
247      my $genomeHash = $self->{genomes};      my $genomeHash = $self->{genomes};
248      my $genomeCount = (keys %{$genomeHash});      my $genomeCount = (keys %{$genomeHash});
     Trace("Beginning genome data load.") if T(2);  
249      # Create load objects for each of the tables we're loading.      # Create load objects for each of the tables we're loading.
250      my $loadGenome = $self->_TableLoader('Genome', $genomeCount);      my $loadGenome = $self->_TableLoader('Genome');
251      my $loadHasContig = $self->_TableLoader('HasContig', $genomeCount * 300);      my $loadHasContig = $self->_TableLoader('HasContig');
252      my $loadContig = $self->_TableLoader('Contig', $genomeCount * 300);      my $loadContig = $self->_TableLoader('Contig');
253      my $loadIsMadeUpOf = $self->_TableLoader('IsMadeUpOf', $genomeCount * 60000);      my $loadIsMadeUpOf = $self->_TableLoader('IsMadeUpOf');
254      my $loadSequence = $self->_TableLoader('Sequence', $genomeCount * 60000);      my $loadSequence = $self->_TableLoader('Sequence');
255        if ($self->{options}->{loadOnly}) {
256            Trace("Loading from existing files.") if T(2);
257        } else {
258            Trace("Generating genome data.") if T(2);
259            # Get the full info for the FIG genomes.
260            my %genomeInfo = map { $_->[0] => { gname => $_->[1], szdna => $_->[2], maindomain => $_->[3],
261                                                pegs => $_->[4], rnas => $_->[5], complete => $_->[6] } } @{$fig->genome_info()};
262      # Now we loop through the genomes, generating the data for each one.      # Now we loop through the genomes, generating the data for each one.
263      for my $genomeID (sort keys %{$genomeHash}) {      for my $genomeID (sort keys %{$genomeHash}) {
264          Trace("Loading data for genome $genomeID.") if T(3);              Trace("Generating data for genome $genomeID.") if T(3);
265          $loadGenome->Add("genomeIn");          $loadGenome->Add("genomeIn");
266          # The access code comes in via the genome hash.          # The access code comes in via the genome hash.
267          my $accessCode = $genomeHash->{$genomeID};          my $accessCode = $genomeHash->{$genomeID};
268          # Get the genus, species, and strain from the scientific name. Note that we append              # Get the genus, species, and strain from the scientific name.
         # the genome ID to the strain. In some cases this is the totality of the strain name.  
269          my ($genus, $species, @extraData) = split / /, $self->{fig}->genus_species($genomeID);          my ($genus, $species, @extraData) = split / /, $self->{fig}->genus_species($genomeID);
270          my $extra = join " ", @extraData, "[$genomeID]";              my $extra = join " ", @extraData;
271          # Get the full taxonomy.          # Get the full taxonomy.
272          my $taxonomy = $fig->taxonomy_of($genomeID);          my $taxonomy = $fig->taxonomy_of($genomeID);
273                # Get the version. If no version is specified, we default to the genome ID by itself.
274                my $version = $fig->genome_version($genomeID);
275                if (! defined($version)) {
276                    $version = $genomeID;
277                }
278                # Get the DNA size.
279                my $dnaSize = $fig->genome_szdna($genomeID);
280                # Open the NMPDR group file for this genome.
281                my $group;
282                if (open(TMP, "<$FIG_Config::organisms/$genomeID/NMPDR") &&
283                    defined($group = <TMP>)) {
284                    # Clean the line ending.
285                    chomp $group;
286                } else {
287                    # No group, so use the default.
288                    $group = $FIG_Config::otherGroup;
289                }
290                close TMP;
291                # Get the contigs.
292                my @contigs = $fig->all_contigs($genomeID);
293                # Get this genome's info array.
294                my $info = $genomeInfo{$genomeID};
295          # Output the genome record.          # Output the genome record.
296          $loadGenome->Put($genomeID, $accessCode, $fig->is_complete($genomeID), $genus,              $loadGenome->Put($genomeID, $accessCode, $info->{complete}, scalar(@contigs),
297                           $species, $extra, $taxonomy);                               $dnaSize, $genus, $info->{pegs}, $group, $info->{rnas}, $species, $extra, $version, $taxonomy);
298          # Now we loop through each of the genome's contigs.          # Now we loop through each of the genome's contigs.
         my @contigs = $fig->all_contigs($genomeID);  
299          for my $contigID (@contigs) {          for my $contigID (@contigs) {
300              Trace("Processing contig $contigID for $genomeID.") if T(4);              Trace("Processing contig $contigID for $genomeID.") if T(4);
301              $loadContig->Add("contigIn");              $loadContig->Add("contigIn");
# Line 267  Line 324 
324              }              }
325          }          }
326      }      }
327        }
328      # Finish the loads.      # Finish the loads.
329      my $retVal = $self->_FinishAll();      my $retVal = $self->_FinishAll();
330      # Return the result.      # Return the result.
331      return $retVal;      return $retVal;
332  }  }
333    
 =head3 LoadCouplingData  
   
 C<< my $stats = $spl->LoadCouplingData(); >>  
   
 Load the coupling and evidence data from FIG into Sprout.  
   
 The coupling data specifies which genome features are functionally coupled. The  
 evidence data explains why the coupling is functional.  
   
 The following relations are loaded by this method.  
   
     Coupling  
     IsEvidencedBy  
     PCH  
     ParticipatesInCoupling  
     UsesAsEvidence  
   
 =over 4  
   
 =item RETURNS  
   
 Returns a statistics object for the loads.  
   
 =back  
   
 =cut  
 #: Return Type $%;  
 sub LoadCouplingData {  
     # Get this object instance.  
     my ($self) = @_;  
     # Get the FIG object.  
     my $fig = $self->{fig};  
     # Get the genome hash.  
     my $genomeFilter = $self->{genomes};  
     my $genomeCount = (keys %{$genomeFilter});  
     my $featureCount = $genomeCount * 4000;  
     # Start the loads.  
     my $loadCoupling = $self->_TableLoader('Coupling', $featureCount * $genomeCount);  
     my $loadIsEvidencedBy = $self->_TableLoader('IsEvidencedBy', $featureCount * 8000);  
     my $loadPCH = $self->_TableLoader('PCH', $featureCount * 2000);  
     my $loadParticipatesInCoupling = $self->_TableLoader('ParticipatesInCoupling', $featureCount * 2000);  
     my $loadUsesAsEvidence = $self->_TableLoader('UsesAsEvidence', $featureCount * 8000);  
     Trace("Beginning coupling data load.") if T(2);  
     # Loop through the genomes found.  
     for my $genome (sort keys %{$genomeFilter}) {  
         Trace("Generating coupling data for $genome.") if T(3);  
         $loadCoupling->Add("genomeIn");  
         # Create a hash table for holding coupled pairs. We use this to prevent  
         # duplicates. For example, if A is coupled to B, we don't want to also  
         # assert that B is coupled to A, because we already know it. Fortunately,  
         # all couplings occur within a genome, so we can keep the hash table  
         # size reasonably small.  
         my %dupHash = ();  
         # Get all of the genome's PEGs.  
         my @pegs = $fig->pegs_of($genome);  
         # Loop through the PEGs.  
         for my $peg1 (@pegs) {  
             $loadCoupling->Add("pegIn");  
             Trace("Processing PEG $peg1 for $genome.") if T(4);  
             # Get a list of the coupled PEGs.  
             my @couplings = $fig->coupled_to($peg1);  
             # For each coupled PEG, we need to verify that a coupling already  
             # exists. If not, we have to create one.  
             for my $coupleData (@couplings) {  
                 my ($peg2, $score) = @{$coupleData};  
                 # Compute the coupling ID.  
                 my $coupleID = Sprout::CouplingID($peg1, $peg2);  
                 if (! exists $dupHash{$coupleID}) {  
                     $loadCoupling->Add("couplingIn");  
                     # Here we have a new coupling to store in the load files.  
                     Trace("Storing coupling ($coupleID) with score $score.") if T(4);  
                     # Ensure we don't do this again.  
                     $dupHash{$coupleID} = $score;  
                     # Write the coupling record.  
                     $loadCoupling->Put($coupleID, $score);  
                     # Connect it to the coupled PEGs.  
                     $loadParticipatesInCoupling->Put($peg1, $coupleID, 1);  
                     $loadParticipatesInCoupling->Put($peg2, $coupleID, 2);  
                     # Get the evidence for this coupling.  
                     my @evidence = $fig->coupling_evidence($peg1, $peg2);  
                     # Organize the evidence into a hash table.  
                     my %evidenceMap = ();  
                     # Process each evidence item.  
                     for my $evidenceData (@evidence) {  
                         $loadPCH->Add("evidenceIn");  
                         my ($peg3, $peg4, $usage) = @{$evidenceData};  
                         # Only proceed if the evidence is from a Sprout  
                         # genome.  
                         if ($genomeFilter->{$fig->genome_of($peg3)}) {  
                             $loadUsesAsEvidence->Add("evidenceChosen");  
                             my $evidenceKey = "$coupleID $peg3 $peg4";  
                             # We store this evidence in the hash if the usage  
                             # is nonzero or no prior evidence has been found. This  
                             # insures that if there is duplicate evidence, we  
                             # at least keep the meaningful ones. Only evidence is  
                             # the hash makes it to the output.  
                             if ($usage || ! exists $evidenceMap{$evidenceKey}) {  
                                 $evidenceMap{$evidenceKey} = $evidenceData;  
                             }  
                         }  
                     }  
                     for my $evidenceID (keys %evidenceMap) {  
                         # Create the evidence record.  
                         my ($peg3, $peg4, $usage) = @{$evidenceMap{$evidenceID}};  
                         $loadPCH->Put($evidenceID, $usage);  
                         # Connect it to the coupling.  
                         $loadIsEvidencedBy->Put($coupleID, $evidenceID);  
                         # Connect it to the features.  
                         $loadUsesAsEvidence->Put($evidenceID, $peg3, 1);  
                         $loadUsesAsEvidence->Put($evidenceID, $peg4, 1);  
                     }  
                 }  
             }  
         }  
     }  
     # All done. Finish the load.  
     my $retVal = $self->_FinishAll();  
     return $retVal;  
 }  
   
334  =head3 LoadFeatureData  =head3 LoadFeatureData
335    
336  C<< my $stats = $spl->LoadFeatureData(); >>      my $stats = $spl->LoadFeatureData();
337    
338  Load the feature data from FIG into Sprout.  Load the feature data from FIG into Sprout.
339    
# Line 405  Line 343 
343    
344      Feature      Feature
345      FeatureAlias      FeatureAlias
346        IsAliasOf
347      FeatureLink      FeatureLink
348      FeatureTranslation      FeatureTranslation
349      FeatureUpstream      FeatureUpstream
350      IsLocatedIn      IsLocatedIn
351        HasFeature
352        HasRoleInSubsystem
353        FeatureEssential
354        FeatureVirulent
355        FeatureIEDB
356        CDD
357        IsPresentOnProteinOf
358        CellLocation
359        IsPossiblePlaceFor
360        ExternalDatabase
361        IsAlsoFoundIn
362        Keyword
363    
364  =over 4  =over 4
365    
# Line 423  Line 374 
374  sub LoadFeatureData {  sub LoadFeatureData {
375      # Get this object instance.      # Get this object instance.
376      my ($self) = @_;      my ($self) = @_;
377      # Get the FIG object.      # Get the FIG and Sprout objects.
378      my $fig = $self->{fig};      my $fig = $self->{fig};
379      # Find out if this is a limited run.      my $sprout = $self->{sprout};
     my $limited = $self->{options}->{limitedFeatures};  
380      # Get the table of genome IDs.      # Get the table of genome IDs.
381      my $genomeHash = $self->{genomes};      my $genomeHash = $self->{genomes};
     my $genomeCount = (keys %{$genomeHash});  
     my $featureCount = $genomeCount * 4000;  
382      # Create load objects for each of the tables we're loading.      # Create load objects for each of the tables we're loading.
383      my $loadFeature = $self->_TableLoader('Feature', $featureCount);      my $loadFeature = $self->_TableLoader('Feature');
384      my $loadIsLocatedIn = $self->_TableLoader('IsLocatedIn', $featureCount);      my $loadIsLocatedIn = $self->_TableLoader('IsLocatedIn');
385      my ($loadFeatureAlias, $loadFeatureLink, $loadFeatureTranslation, $loadFeatureUpstream);      my $loadFeatureAlias = $self->_TableLoader('FeatureAlias');
386      if (! $limited) {      my $loadIsAliasOf = $self->_TableLoader('IsAliasOf');
387          $loadFeatureAlias = $self->_TableLoader('FeatureAlias', $featureCount * 6);      my $loadFeatureLink = $self->_TableLoader('FeatureLink');
388          $loadFeatureLink = $self->_TableLoader('FeatureLink', $featureCount * 10);      my $loadFeatureTranslation = $self->_TableLoader('FeatureTranslation');
389          $loadFeatureTranslation = $self->_TableLoader('FeatureTranslation', $featureCount);      my $loadFeatureUpstream = $self->_TableLoader('FeatureUpstream');
390          $loadFeatureUpstream = $self->_TableLoader('FeatureUpstream', $featureCount);      my $loadHasFeature = $self->_TableLoader('HasFeature');
391      }      my $loadHasRoleInSubsystem = $self->_TableLoader('HasRoleInSubsystem');
392        my $loadFeatureEssential = $self->_TableLoader('FeatureEssential');
393        my $loadFeatureVirulent = $self->_TableLoader('FeatureVirulent');
394        my $loadFeatureIEDB = $self->_TableLoader('FeatureIEDB');
395        my $loadCDD = $self->_TableLoader('CDD');
396        my $loadIsPresentOnProteinOf = $self->_TableLoader('IsPresentOnProteinOf');
397        my $loadCellLocation = $self->_TableLoader('CellLocation');
398        my $loadIsPossiblePlaceFor = $self->_TableLoader('IsPossiblePlaceFor');
399        my $loadIsAlsoFoundIn = $self->_TableLoader('IsAlsoFoundIn');
400        my $loadExternalDatabase = $self->_TableLoader('ExternalDatabase');
401        my $loadKeyword = $self->_TableLoader('Keyword');
402        # Get the subsystem hash.
403        my $subHash = $self->{subsystems};
404        # Get the property keys.
405        my $propKeys = $self->{propKeys};
406        # Create a hashes to hold CDD, Cell Location (PSORT), External Database, and alias values.
407        my %CDD = ();
408        my %alias = ();
409        my %cellLocation = ();
410        my %xdb = ();
411        # Create the bio-words object.
412        my $biowords = BioWords->new(cache => 0);
413        # One of the things we have to do here is build the keyword table, and the keyword
414        # table needs to contain the originating text and feature count for each stem. Unfortunately,
415        # the number of distinct keywords is so large it causes PERL to hang if we try to
416        # keep them in memory. As a result, we need to track them using disk files.
417        # Our approach will be to use two sequential files. One will contain stems and phonexes.
418        # Each time a stem occurs in a feature, a record will be written to that file. The stem
419        # file can then be sorted and collated to determine the number of features for each
420        # stem. A separate file will contain keywords and stems. This last file
421        # will be subjected to a sort unique on stem/keyword. The file is then merged
422        # with the stem file to create the keyword table relation (keyword, stem, phonex, count).
423        my $stemFileName = "$FIG_Config::temp/stems$$.tbl";
424        my $keyFileName = "$FIG_Config::temp/keys$$.tbl";
425        my $stemh = Open(undef, "| sort -T\"$FIG_Config::temp\" -t\"\t\" -k1,1 >$stemFileName");
426        my $keyh = Open(undef, "| sort -T\"$FIG_Config::temp\" -t\"\t\" -u -k1,1 -k2,2 >$keyFileName");
427      # Get the maximum sequence size. We need this later for splitting up the      # Get the maximum sequence size. We need this later for splitting up the
428      # locations.      # locations.
429      my $chunkSize = $self->{sprout}->MaxSegment();      my $chunkSize = $self->{sprout}->MaxSegment();
430      Trace("Beginning feature data load.") if T(2);      if ($self->{options}->{loadOnly}) {
431            Trace("Loading from existing files.") if T(2);
432        } else {
433            Trace("Generating feature data.") if T(2);
434      # Now we loop through the genomes, generating the data for each one.      # Now we loop through the genomes, generating the data for each one.
435      for my $genomeID (sort keys %{$genomeHash}) {          my @allGenomes = sort keys %{$genomeHash};
436            Trace(scalar(@allGenomes) . " genomes found in list.") if T(3);
437            for my $genomeID (@allGenomes) {
438          Trace("Loading features for genome $genomeID.") if T(3);          Trace("Loading features for genome $genomeID.") if T(3);
439          $loadFeature->Add("genomeIn");          $loadFeature->Add("genomeIn");
440          # Get the feature list for this genome.          # Get the feature list for this genome.
441          my $features = $fig->all_features_detailed($genomeID);              my $features = $fig->all_features_detailed_fast($genomeID);
442                # Sort and count the list.
443                my @featureTuples = sort { $a->[0] cmp $b->[0] } @{$features};
444                my $count = scalar @featureTuples;
445                my @fids = map { $_->[0] } @featureTuples;
446                Trace("$count features found for genome $genomeID.") if T(3);
447                # Get the attributes for this genome and put them in a hash by feature ID.
448                my $attributes = GetGenomeAttributes($fig, $genomeID, \@fids, $propKeys);
449                Trace("Looping through features for $genomeID.") if T(3);
450                # Set up for our duplicate-feature check.
451                my $oldFeatureID = "";
452          # Loop through the features.          # Loop through the features.
453          for my $featureData (@{$features}) {              for my $featureTuple (@featureTuples) {
             $loadFeature->Add("featureIn");  
454              # Split the tuple.              # Split the tuple.
455              my ($featureID, $locations, $aliases, $type) = @{$featureData};                  my ($featureID, $locations, undef, $type, $minloc, $maxloc, $assignment, $user, $quality) = @{$featureTuple};
456              # Create the feature record.                  # Check for duplicates.
457              $loadFeature->Put($featureID, 1, $type);                  if ($featureID eq $oldFeatureID) {
458              # The next stuff is for a full load only.                      Trace("Duplicate feature $featureID found.") if T(1);
459              if (! $limited) {                  } else {
460                        $oldFeatureID = $featureID;
461                        # Count this feature.
462                        $loadFeature->Add("featureIn");
463                        # Fix the quality. It is almost always a space, but some odd stuff might sneak through, and the
464                        # Sprout database requires a single character.
465                        if (! defined($quality) || $quality eq "") {
466                            $quality = " ";
467                        }
468                        # Begin building the keywords. We start with the genome ID, the
469                        # feature ID, the taxonomy, and the organism name.
470                        my @keywords = ($genomeID, $featureID, $fig->genus_species($genomeID),
471                                        $fig->taxonomy_of($genomeID));
472                  # Create the aliases.                  # Create the aliases.
473                  for my $alias (split /\s*,\s*/, $aliases) {                      for my $alias ($fig->feature_aliases($featureID)) {
474                      $loadFeatureAlias->Put($featureID, $alias);                          #Connect this alias to this feature.
475                            $loadIsAliasOf->Put($alias, $featureID);
476                            push @keywords, $alias;
477                            # If this is a locus tag, also add its natural form as a keyword.
478                            my $naturalName = AliasAnalysis::Type(LocusTag => $alias);
479                            if ($naturalName) {
480                                push @keywords, $naturalName;
481                            }
482                            # If this is the first time for the specified alias, create its
483                            # alias record.
484                            if (! exists $alias{$alias}) {
485                                $loadFeatureAlias->Put($alias);
486                                $alias{$alias} = 1;
487                            }
488                        }
489                        # Add the corresponding IDs. We ask for 2-tuples of the form (id, database).
490                        my @corresponders = $fig->get_corresponding_ids($featureID, 1);
491                        for my $tuple (@corresponders) {
492                            my ($id, $xdb) = @{$tuple};
493                            # Ignore SEED: that's us.
494                            if ($xdb ne 'SEED') {
495                                # Connect this ID to the feature.
496                                $loadIsAlsoFoundIn->Put($featureID, $xdb, $id);
497                                # Add it as a keyword.
498                                push @keywords, $id;
499                                # If this is a new database, create a record for it.
500                                if (! exists $xdb{$xdb}) {
501                                    $xdb{$xdb} = 1;
502                                    $loadExternalDatabase->Put($xdb);
503                                }
504                            }
505                  }                  }
506                        Trace("Assignment for $featureID is: $assignment") if T(4);
507                        # Break the assignment into words and shove it onto the
508                        # keyword list.
509                        push @keywords, split(/\s+/, $assignment);
510                        # Link this feature to the parent genome.
511                        $loadHasFeature->Put($genomeID, $featureID, $type);
512                  # Get the links.                  # Get the links.
513                  my @links = $fig->fid_links($featureID);                  my @links = $fig->fid_links($featureID);
514                  for my $link (@links) {                  for my $link (@links) {
# Line 482  Line 527 
527                          $loadFeatureUpstream->Put($featureID, $upstream);                          $loadFeatureUpstream->Put($featureID, $upstream);
528                      }                      }
529                  }                  }
530                        # Now we need to find the subsystems this feature participates in.
531                        # We also add the subsystems to the keyword list. Before we do that,
532                        # we must convert underscores to spaces.
533                        my @subsystems = $fig->peg_to_subsystems($featureID);
534                        for my $subsystem (@subsystems) {
535                            # Only proceed if we like this subsystem.
536                            if (exists $subHash->{$subsystem}) {
537                                # Store the has-role link.
538                                $loadHasRoleInSubsystem->Put($featureID, $subsystem, $genomeID, $type);
539                                # Save the subsystem's keyword data.
540                                my $subKeywords = $subHash->{$subsystem};
541                                push @keywords, split /\s+/, $subKeywords;
542                                # Now we need to get this feature's role in the subsystem.
543                                my $subObject = $fig->get_subsystem($subsystem);
544                                my @roleColumns = $subObject->get_peg_roles($featureID);
545                                my @allRoles = $subObject->get_roles();
546                                for my $col (@roleColumns) {
547                                    my $role = $allRoles[$col];
548                                    push @keywords, split /\s+/, $role;
549                                    push @keywords, $subObject->get_role_abbr($col);
550              }              }
551                            }
552                        }
553                        # There are three special attributes computed from property
554                        # data that we build next. If the special attribute is non-empty,
555                        # its name will be added to the keyword list. First, we get all
556                        # the attributes for this feature. They will come back as
557                        # 4-tuples: [peg, name, value, URL]. We use a 3-tuple instead:
558                        # [name, value, value with URL]. (We don't need the PEG, since
559                        # we already know it.)
560                        my @attributes = map { [$_->[1], $_->[2], Tracer::CombineURL($_->[2], $_->[3])] }
561                                             @{$attributes->{$featureID}};
562                        # Now we process each of the special attributes.
563                        if (SpecialAttribute($featureID, \@attributes,
564                                             1, [0,2], '^(essential|potential_essential)$',
565                                             $loadFeatureEssential)) {
566                            push @keywords, 'essential';
567                            $loadFeature->Add('essential');
568                        }
569                        if (SpecialAttribute($featureID, \@attributes,
570                                             0, [2], '^virulen',
571                                             $loadFeatureVirulent)) {
572                            push @keywords, 'virulent';
573                            $loadFeature->Add('virulent');
574                        }
575                        if (SpecialAttribute($featureID, \@attributes,
576                                             0, [0,2], '^iedb_',
577                                             $loadFeatureIEDB)) {
578                            push @keywords, 'iedb';
579                            $loadFeature->Add('iedb');
580                        }
581                        # Now we have some other attributes we need to process. To get
582                        # through them, we convert the attribute list for this feature
583                        # into a two-layer hash: key => subkey => value.
584                        my %attributeHash = ();
585                        for my $attrRow (@{$attributes->{$featureID}}) {
586                            my (undef, $key, @values) = @{$attrRow};
587                            my ($realKey, $subKey);
588                            if ($key =~ /^([^:]+)::(.+)/) {
589                                ($realKey, $subKey) = ($1, $2);
590                            } else {
591                                ($realKey, $subKey) = ($key, "");
592                            }
593                            if (exists $attributeHash{$1}) {
594                                $attributeHash{$1}->{$2} = \@values;
595                            } else {
596                                $attributeHash{$1} = {$2 => \@values};
597                            }
598                        }
599                        # First we handle CDD. This is a bit complicated, because
600                        # there are multiple CDDs per protein.
601                        if (exists $attributeHash{CDD}) {
602                            # Get the hash of CDD IDs to scores for this feature. We
603                            # already know it exists because of the above IF.
604                            my $cddHash = $attributeHash{CDD};
605                            my @cddData = sort keys %{$cddHash};
606                            for my $cdd (@cddData) {
607                                # Extract the score for this CDD and decode it.
608                                my ($codeScore) = split(/\s*[,;]\s*/, $cddHash->{$cdd}->[0]);
609                                my $realScore = FIGRules::DecodeScore($codeScore);
610                                # We can't afford to crash because of a bad attribute
611                                # value, hence the IF below.
612                                if (! defined($realScore)) {
613                                    # Bad score, so count it.
614                                    $loadFeature->Add('badCDDscore');
615                                    Trace("CDD score \"$codeScore\" for feature $featureID invalid.") if T(3);
616                                } else {
617                                    # Create the connection.
618                                    $loadIsPresentOnProteinOf->Put($cdd, $featureID, $realScore);
619                                    # If this CDD does not yet exist, create its record.
620                                    if (! exists $CDD{$cdd}) {
621                                        $CDD{$cdd} = 1;
622                                        $loadCDD->Put($cdd);
623                                    }
624                                }
625                            }
626                        }
627                        # Next we do PSORT cell locations. here the confidence value
628                        # could have the value "unknown", which we translate to -1.
629                        if (exists $attributeHash{PSORT}) {
630                            # This will be a hash of cell locations to confidence
631                            # factors.
632                            my $psortHash = $attributeHash{PSORT};
633                            for my $psort (keys %{$psortHash}) {
634                                # Get the confidence, and convert it to a number if necessary.
635                                my $confidence = $psortHash->{$psort};
636                                if ($confidence eq 'unknown') {
637                                    $confidence = -1;
638                                }
639                                $loadIsPossiblePlaceFor->Put($psort, $featureID, $confidence);
640                                # If this cell location does not yet exist, create its record.
641                                if (! exists $cellLocation{$psort}) {
642                                    $cellLocation{$psort} = 1;
643                                    $loadCellLocation->Put($psort);
644                                }
645                                # If this is a significant location, add it as a keyword.
646                                if ($confidence > 2.5) {
647                                    push @keywords, $psort;
648                                }
649                            }
650                        }
651                        # Phobius data is next. This consists of the signal peptide location and
652                        # the transmembrane locations.
653                        my $signalList = "";
654                        my $transList = "";
655                        if (exists $attributeHash{Phobius}) {
656                            # This will be a hash of two keys (transmembrane and signal) to
657                            # location strings. If there's no value, we stuff in an empty string.
658                            $signalList = ($attributeHash{Phobius}->{signal} || "");
659                            $transList = ($attributeHash{Phobius}->{transmembrane} || "");
660                        }
661                        # Here are some more numbers: isoelectric point, molecular weight, and
662                        # the similar-to-human flag.
663                        my $isoelectric = 0;
664                        if (exists $attributeHash{isoelectric_point}) {
665                            $isoelectric = $attributeHash{isoelectric_point}->{""};
666                        }
667                        my $similarToHuman = 0;
668                        if (exists $attributeHash{similar_to_human} && $attributeHash{similar_to_human}->{""} eq 'yes') {
669                            $similarToHuman = 1;
670                        }
671                        my $molecularWeight = 0;
672                        if (exists $attributeHash{molecular_weight}) {
673                            $molecularWeight = $attributeHash{molecular_weight}->{""};
674                        }
675                        # Create the keyword string.
676                        my $keywordString = join(" ", @keywords);
677                        Trace("Real keyword string for $featureID: $keywordString.") if T(4);
678                        # Get rid of annoying punctuation.
679                        $keywordString =~ s/[();@#\/]/ /g;
680                        # Get the list of keywords in the keyword string.
681                        my @realKeywords = grep { $biowords->IsWord($_) } $biowords->Split($keywordString);
682                        # We need to do two things here: create the keyword string for the feature table
683                        # and write records to the keyword and stem files. The stuff we write to
684                        # the files will be taken from the following two hashes. The stuff used
685                        # to create the keyword string will be taken from the list.
686                        my (%keys, %stems, @realStems);
687                        for my $keyword (@realKeywords) {
688                            # Compute the stem and phonex for this keyword.
689                            my ($stem, $phonex) = $biowords->StemLookup($keyword);
690                            # Only proceed if a stem comes back. If no stem came back, it's a
691                            # stop word and we throw it away.
692                            if ($stem) {
693                                $keys{$keyword} = $stem;
694                                $stems{$stem} = $phonex;
695                                push @realStems, $stem;
696                            }
697                        }
698                        # Now create the keyword string.
699                        my $cleanWords = join(" ", @realStems);
700                        Trace("Keyword string for $featureID: $cleanWords") if T(4);
701                        # Write the stem and keyword records.
702                        for my $stem (keys %stems) {
703                            Tracer::PutLine($stemh, [$stem, $stems{$stem}]);
704                        }
705                        for my $key (keys %keys) {
706                            # The stem goes first in this file, because we want to sort
707                            # by stem and then keyword.
708                            Tracer::PutLine($keyh, [$keys{$key}, $key]);
709                        }
710                        # Now we need to process the feature's locations. First, we split them up.
711                        my @locationList = split /\s*,\s*/, $locations;
712                        # Next, we convert them to Sprout location objects.
713                        my @locObjectList = map { BasicLocation->new("$genomeID:$_") } @locationList;
714                        # Assemble them into a sprout location string for later.
715                        my $locationString = join(", ", map { $_->String } @locObjectList);
716                        # We'll store the sequence length in here.
717                        my $sequenceLength = 0;
718              # This part is the roughest. We need to relate the features to contig              # This part is the roughest. We need to relate the features to contig
719              # locations, and the locations must be split so that none of them exceed              # locations, and the locations must be split so that none of them exceed
720              # the maximum segment size. This simplifies the genes_in_region processing              # the maximum segment size. This simplifies the genes_in_region processing
721              # for Sprout.                      # for Sprout. To start, we create the location position indicator.
             my @locationList = map { "$genomeID:$_" } split /\s*,\s*/, $locations;  
             # Create the location position indicator.  
722              my $i = 1;              my $i = 1;
723              # Loop through the locations.              # Loop through the locations.
724              for my $location (@locationList) {                      for my $locObject (@locObjectList) {
725                  # Parse the location.                          # Record the length.
726                  my $locObject = BasicLocation->new($location);                          $sequenceLength += $locObject->Length;
727                  # Split it into a list of chunks.                          # Split this location into a list of chunks.
728                  my @locOList = ();                  my @locOList = ();
729                  while (my $peeling = $locObject->Peel($chunkSize)) {                  while (my $peeling = $locObject->Peel($chunkSize)) {
730                      $loadIsLocatedIn->Add("peeling");                      $loadIsLocatedIn->Add("peeling");
# Line 509  Line 739 
739                      $i++;                      $i++;
740                  }                  }
741              }              }
742                        # Now we get some ancillary flags.
743                        my $locked = $fig->is_locked_fid($featureID);
744                        my $in_genbank = $fig->peg_in_gendb($featureID);
745                        # Create the feature record.
746                        $loadFeature->Put($featureID, 1, $user, $quality, $type, $in_genbank, $isoelectric, $locked, $molecularWeight,
747                                          $sequenceLength, $signalList, $similarToHuman, $assignment, $cleanWords, $locationString,
748                                          $transList);
749                    }
750                }
751                Trace("Genome $genomeID processed.") if T(3);
752            }
753        }
754        Trace("Sorting keywords.") if T(2);
755        # Now we need to load the keyword table from the key and stem files.
756        close $keyh;
757        close $stemh;
758        Trace("Loading keywords.") if T(2);
759        $keyh = Open(undef, "<$keyFileName");
760        $stemh = Open(undef, "<$stemFileName");
761        # We'll count the keywords in here, for tracing purposes.
762        my $count = 0;
763        # These variables track the current stem's data. When an incoming
764        # keyword's stem changes, these will be recomputed.
765        my ($currentStem, $currentPhonex, $currentCount);
766        # Prime the loop by reading the first stem in the stem file.
767        my ($nextStem, $nextPhonex) = Tracer::GetLine($stemh);
768        # Loop through the keyword file.
769        while (! eof $keyh) {
770            # Read this keyword.
771            my ($thisStem, $thisKey) = Tracer::GetLine($keyh);
772            # Check to see if it's the new stem yet.
773            if ($thisStem ne $currentStem) {
774                # Yes. It's a terrible error if it's not also the next stem.
775                if ($thisStem ne $nextStem) {
776                    Confess("Error in stem file. Expected \"$nextStem\", but found \"$thisStem\".");
777                } else {
778                    # Here we're okay.
779                    ($currentStem, $currentPhonex) = ($nextStem, $nextPhonex);
780                    # Count the number of features for this stem.
781                    $currentCount = 0;
782                    while ($nextStem eq $thisStem) {
783                        ($nextStem, $nextPhonex) = Tracer::GetLine($stemh);
784                        $currentCount++;
785          }          }
786      }      }
     # Finish the loads.  
     my $retVal = $self->_FinishAll();  
     return $retVal;  
 }  
   
 =head3 LoadBBHData  
   
 C<< my $stats = $spl->LoadBBHData(); >>  
   
 Load the bidirectional best hit data from FIG into Sprout.  
   
 Sprout does not store information on similarities. Instead, it has only the  
 bi-directional best hits. Even so, the BBH table is one of the largest in  
 the database.  
   
 The following relations are loaded by this method.  
   
     IsBidirectionalBestHitOf  
   
 =over 4  
   
 =item RETURNS  
   
 Returns a statistics object for the loads.  
   
 =back  
   
 =cut  
 #: Return Type $%;  
 sub LoadBBHData {  
     # Get this object instance.  
     my ($self) = @_;  
     # Get the FIG object.  
     my $fig = $self->{fig};  
     # Get the table of genome IDs.  
     my $genomeHash = $self->{genomes};  
     my $genomeCount = (keys %{$genomeHash});  
     my $featureCount = $genomeCount * 4000;  
     # Create load objects for each of the tables we're loading.  
     my $loadIsBidirectionalBestHitOf = $self->_TableLoader('IsBidirectionalBestHitOf',  
                                                            $featureCount * $genomeCount);  
     Trace("Beginning BBH load.") if T(2);  
     # Now we loop through the genomes, generating the data for each one.  
     for my $genomeID (sort keys %{$genomeHash}) {  
         $loadIsBidirectionalBestHitOf->Add("genomeIn");  
         Trace("Processing features for genome $genomeID.") if T(3);  
         # Get the feature list for this genome.  
         my $features = $fig->all_features_detailed($genomeID);  
         # Loop through the features.  
         for my $featureData (@{$features}) {  
             # Split the tuple.  
             my ($featureID, $locations, $aliases, $type) = @{$featureData};  
             # Get the bi-directional best hits.  
             my @bbhList = $fig->bbhs($featureID);  
             for my $bbhEntry (@bbhList) {  
                 # Get the target feature ID and the score.  
                 my ($targetID, $score) = @{$bbhEntry};  
                 # Check the target feature's genome.  
                 my $targetGenomeID = $fig->genome_of($targetID);  
                 # Only proceed if it's one of our genomes.  
                 if ($genomeHash->{$targetGenomeID}) {  
                     $loadIsBidirectionalBestHitOf->Put($featureID, $targetID, $targetGenomeID,  
                                                        $score);  
                 }  
787              }              }
788            # Now $currentStem is the same as $thisStem, and the other $current-vars
789            # contain the stem's data (phonex and count).
790            $loadKeyword->Put($thisKey, $currentCount, $currentPhonex, $currentStem);
791            if (++$count % 1000 == 0 && T(3)) {
792                Trace("$count keywords loaded.");
793          }          }
794      }      }
795        Trace("$count keywords loaded into keyword table.") if T(2);
796      # Finish the loads.      # Finish the loads.
797      my $retVal = $self->_FinishAll();      my $retVal = $self->_FinishAll();
798      return $retVal;      return $retVal;
# Line 585  Line 800 
800    
801  =head3 LoadSubsystemData  =head3 LoadSubsystemData
802    
803  C<< my $stats = $spl->LoadSubsystemData(); >>      my $stats = $spl->LoadSubsystemData();
804    
805  Load the subsystem data from FIG into Sprout.  Load the subsystem data from FIG into Sprout.
806    
# Line 598  Line 813 
813  The following relations are loaded by this method.  The following relations are loaded by this method.
814    
815      Subsystem      Subsystem
816        SubsystemClass
817      Role      Role
818        RoleEC
819        IsIdentifiedByEC
820      SSCell      SSCell
821      ContainsFeature      ContainsFeature
822      IsGenomeOf      IsGenomeOf
# Line 606  Line 824 
824      OccursInSubsystem      OccursInSubsystem
825      ParticipatesIn      ParticipatesIn
826      HasSSCell      HasSSCell
827        ConsistsOfRoles
828        RoleSubset
829        HasRoleSubset
830        ConsistsOfGenomes
831        GenomeSubset
832        HasGenomeSubset
833        Diagram
834        RoleOccursIn
835        SubsystemHopeNotes
836    
837  =over 4  =over 4
838    
# Line 615  Line 842 
842    
843  =back  =back
844    
 B<TO DO>  
   
 Generate RoleName table?  
   
845  =cut  =cut
846  #: Return Type $%;  #: Return Type $%;
847  sub LoadSubsystemData {  sub LoadSubsystemData {
# Line 632  Line 855 
855      # Get the subsystem hash. This lists the subsystems we'll process.      # Get the subsystem hash. This lists the subsystems we'll process.
856      my $subsysHash = $self->{subsystems};      my $subsysHash = $self->{subsystems};
857      my @subsysIDs = sort keys %{$subsysHash};      my @subsysIDs = sort keys %{$subsysHash};
858      my $subsysCount = @subsysIDs;      # Get the map list.
859      my $genomeCount = (keys %{$genomeHash});      my @maps = $fig->all_maps;
     my $featureCount = $genomeCount * 4000;  
860      # Create load objects for each of the tables we're loading.      # Create load objects for each of the tables we're loading.
861      my $loadSubsystem = $self->_TableLoader('Subsystem', $subsysCount);      my $loadDiagram = $self->_TableLoader('Diagram');
862      my $loadRole = $self->_TableLoader('Role', $featureCount * 6);      my $loadRoleOccursIn = $self->_TableLoader('RoleOccursIn');
863      my $loadSSCell = $self->_TableLoader('SSCell', $featureCount * $genomeCount);      my $loadSubsystem = $self->_TableLoader('Subsystem');
864      my $loadContainsFeature = $self->_TableLoader('ContainsFeature', $featureCount * $subsysCount);      my $loadRole = $self->_TableLoader('Role');
865      my $loadIsGenomeOf = $self->_TableLoader('IsGenomeOf', $featureCount * $genomeCount);      my $loadRoleEC = $self->_TableLoader('RoleEC');
866      my $loadIsRoleOf = $self->_TableLoader('IsRoleOf', $featureCount * $genomeCount);      my $loadIsIdentifiedByEC = $self->_TableLoader('IsIdentifiedByEC');
867      my $loadOccursInSubsystem = $self->_TableLoader('OccursInSubsystem', $featureCount * 6);      my $loadCatalyzes = $self->_TableLoader('Catalyzes');
868      my $loadParticipatesIn = $self->_TableLoader('ParticipatesIn', $subsysCount * $genomeCount);      my $loadSSCell = $self->_TableLoader('SSCell');
869      my $loadHasSSCell = $self->_TableLoader('HasSSCell', $featureCount * $genomeCount);      my $loadContainsFeature = $self->_TableLoader('ContainsFeature');
870      Trace("Beginning subsystem data load.") if T(2);      my $loadIsGenomeOf = $self->_TableLoader('IsGenomeOf');
871        my $loadIsRoleOf = $self->_TableLoader('IsRoleOf');
872        my $loadOccursInSubsystem = $self->_TableLoader('OccursInSubsystem');
873        my $loadParticipatesIn = $self->_TableLoader('ParticipatesIn');
874        my $loadHasSSCell = $self->_TableLoader('HasSSCell');
875        my $loadRoleSubset = $self->_TableLoader('RoleSubset');
876        my $loadGenomeSubset = $self->_TableLoader('GenomeSubset');
877        my $loadConsistsOfRoles = $self->_TableLoader('ConsistsOfRoles');
878        my $loadConsistsOfGenomes = $self->_TableLoader('ConsistsOfGenomes');
879        my $loadHasRoleSubset = $self->_TableLoader('HasRoleSubset');
880        my $loadHasGenomeSubset = $self->_TableLoader('HasGenomeSubset');
881        my $loadSubsystemClass = $self->_TableLoader('SubsystemClass');
882        my $loadSubsystemHopeNotes = $self->_TableLoader('SubsystemHopeNotes');
883        if ($self->{options}->{loadOnly}) {
884            Trace("Loading from existing files.") if T(2);
885        } else {
886            Trace("Generating subsystem data.") if T(2);
887            # This hash will contain the roles for each EC. When we're done, this
888            # information will be used to generate the Catalyzes table.
889            my %ecToRoles = ();
890      # Loop through the subsystems. Our first task will be to create the      # Loop through the subsystems. Our first task will be to create the
891      # roles. We do this by looping through the subsystems and creating a      # roles. We do this by looping through the subsystems and creating a
892      # role hash. The hash tracks each role ID so that we don't create      # role hash. The hash tracks each role ID so that we don't create
893      # duplicates. As we move along, we'll connect the roles and subsystems.          # duplicates. As we move along, we'll connect the roles and subsystems
894            # and memorize up the reactions.
895            my ($genomeID, $roleID);
896      my %roleData = ();      my %roleData = ();
897      for my $subsysID (@subsysIDs) {      for my $subsysID (@subsysIDs) {
898                # Get the subsystem object.
899                my $sub = $fig->get_subsystem($subsysID);
900                # Only proceed if the subsystem has a spreadsheet.
901                if (defined($sub) && ! $sub->{empty_ss}) {
902          Trace("Creating subsystem $subsysID.") if T(3);          Trace("Creating subsystem $subsysID.") if T(3);
903          $loadSubsystem->Add("subsystemIn");          $loadSubsystem->Add("subsystemIn");
904          # Create the subsystem record.          # Create the subsystem record.
905          $loadSubsystem->Put($subsysID);                  my $curator = $sub->get_curator();
906          # Get the subsystem's roles.                  my $notes = $sub->get_notes();
907          my @roles = $fig->subsystem_to_roles($subsysID);                  my $version = $sub->get_version();
908          # Connect the roles to the subsystem. If a role is new, we create                  my $description = $sub->get_description();
909          # a role record for it.                  $loadSubsystem->Put($subsysID, $curator, $version, $description, $notes);
910          for my $roleID (@roles) {                  # Add the hope notes.
911                    my $hopeNotes = $sub->get_hope_curation_notes();
912                    if ($hopeNotes) {
913                        $loadSubsystemHopeNotes->Put($sub, $hopeNotes);
914                    }
915                    # Now for the classification string. This comes back as a list
916                    # reference and we convert it to a space-delimited string.
917                    my $classList = $fig->subsystem_classification($subsysID);
918                    my $classString = join($FIG_Config::splitter, grep { $_ } @$classList);
919                    $loadSubsystemClass->Put($subsysID, $classString);
920                    # Connect it to its roles. Each role is a column in the subsystem spreadsheet.
921                    for (my $col = 0; defined($roleID = $sub->get_role($col)); $col++) {
922                        # Get the role's abbreviation.
923                        my $abbr = $sub->get_role_abbr($col);
924                        # Get its essentiality.
925                        my $aux = $fig->is_aux_role_in_subsystem($subsysID, $roleID);
926                        # Get its reaction note.
927                        my $hope_note = $sub->get_hope_reaction_notes($roleID) || "";
928                        # Connect to this role.
929              $loadOccursInSubsystem->Add("roleIn");              $loadOccursInSubsystem->Add("roleIn");
930              $loadOccursInSubsystem->Put($roleID, $subsysID);                      $loadOccursInSubsystem->Put($roleID, $subsysID, $abbr, $aux, $col, $hope_note);
931                        # If it's a new role, add it to the role table.
932              if (! exists $roleData{$roleID}) {              if (! exists $roleData{$roleID}) {
933                            # Get the role's abbreviation.
934                            # Add the role.
935                  $loadRole->Put($roleID);                  $loadRole->Put($roleID);
936                  $roleData{$roleID} = 1;                  $roleData{$roleID} = 1;
937                            # Check for an EC number.
938                            if ($roleID =~ /\(EC (\d+\.\d+\.\d+\.\d+)\s*\)\s*$/) {
939                                my $ec = $1;
940                                $loadIsIdentifiedByEC->Put($roleID, $ec);
941                                # Check to see if this is our first encounter with this EC.
942                                if (exists $ecToRoles{$ec}) {
943                                    # No, so just add this role to the EC list.
944                                    push @{$ecToRoles{$ec}}, $roleID;
945                                } else {
946                                    # Output this EC.
947                                    $loadRoleEC->Put($ec);
948                                    # Create its role list.
949                                    $ecToRoles{$ec} = [$roleID];
950                                }
951                            }
952              }              }
953          }          }
954          # Now all roles for this subsystem have been filled in. We create the                  # Now we create the spreadsheet for the subsystem by matching roles to
955          # spreadsheet by matches roles to genomes. To do this, we need to                  # genomes. Each genome is a row and each role is a column. We may need
956          # get the genomes on the sheet.                  # to actually create the roles as we find them.
957          Trace("Creating subsystem $subsysID spreadsheet.") if T(3);          Trace("Creating subsystem $subsysID spreadsheet.") if T(3);
958          my @genomes = map { $_->[0] } @{$fig->subsystem_genomes($subsysID)};                  for (my $row = 0; defined($genomeID = $sub->get_genome($row)); $row++) {
959          for my $genomeID (@genomes) {                      # Only proceed if this is one of our genomes.
             # Only process this genome if it's one of ours.  
960              if (exists $genomeHash->{$genomeID}) {              if (exists $genomeHash->{$genomeID}) {
961                  # Connect the genome to the subsystem.                          # Count the PEGs and cells found for verification purposes.
962                  $loadParticipatesIn->Put($genomeID, $subsysID);                          my $pegCount = 0;
963                            my $cellCount = 0;
964                            # Create a list for the PEGs we find. This list will be used
965                            # to generate cluster numbers.
966                            my @pegsFound = ();
967                            # Create a hash that maps spreadsheet IDs to PEGs. We will
968                            # use this to generate the ContainsFeature data after we have
969                            # the cluster numbers.
970                            my %cellPegs = ();
971                            # Get the genome's variant code for this subsystem.
972                            my $variantCode = $sub->get_variant_code($row);
973                  # Loop through the subsystem's roles. We use an index because it is                  # Loop through the subsystem's roles. We use an index because it is
974                  # part of the spreadsheet cell ID.                  # part of the spreadsheet cell ID.
975                  for (my $i = 0; $i <= $#roles; $i++) {                          for (my $col = 0; defined($roleID = $sub->get_role($col)); $col++) {
                     my $role = $roles[$i];  
976                      # Get the features in the spreadsheet cell for this genome and role.                      # Get the features in the spreadsheet cell for this genome and role.
977                      my @pegs = $fig->pegs_in_subsystem_cell($subsysID, $genomeID, $i);                              my @pegs = grep { !$fig->is_deleted_fid($_) } $sub->get_pegs_from_cell($row, $col);
978                      # Only proceed if features exist.                      # Only proceed if features exist.
979                      if (@pegs > 0) {                      if (@pegs > 0) {
980                          # Create the spreadsheet cell.                          # Create the spreadsheet cell.
981                          my $cellID = "$subsysID:$genomeID:$i";                                  $cellCount++;
982                                    my $cellID = "$subsysID:$genomeID:$col";
983                          $loadSSCell->Put($cellID);                          $loadSSCell->Put($cellID);
984                          $loadIsGenomeOf->Put($genomeID, $cellID);                          $loadIsGenomeOf->Put($genomeID, $cellID);
985                          $loadIsRoleOf->Put($role, $cellID);                                  $loadIsRoleOf->Put($roleID, $cellID);
986                          $loadHasSSCell->Put($subsysID, $cellID);                          $loadHasSSCell->Put($subsysID, $cellID);
987                          # Attach the features to it.                                  # Remember its features.
988                          for my $pegID (@pegs) {                                  push @pegsFound, @pegs;
989                              $loadContainsFeature->Put($cellID, $pegID);                                  $cellPegs{$cellID} = \@pegs;
990                                    $pegCount += @pegs;
991                                }
992                            }
993                            # If we found some cells for this genome, we need to compute clusters and
994                            # denote it participates in the subsystem.
995                            if ($pegCount > 0) {
996                                Trace("$pegCount PEGs in $cellCount cells for $genomeID.") if T(3);
997                                $loadParticipatesIn->Put($genomeID, $subsysID, $variantCode);
998                                # Create a hash mapping PEG IDs to cluster numbers.
999                                # We default to -1 for all of them.
1000                                my %clusterOf = map { $_ => -1 } @pegsFound;
1001                                # Partition the PEGs found into clusters.
1002                                my @clusters = $fig->compute_clusters([keys %clusterOf], $sub);
1003                                for (my $i = 0; $i <= $#clusters; $i++) {
1004                                    my $subList = $clusters[$i];
1005                                    for my $peg (@{$subList}) {
1006                                        $clusterOf{$peg} = $i;
1007                                    }
1008                                }
1009                                # Create the ContainsFeature data.
1010                                for my $cellID (keys %cellPegs) {
1011                                    my $cellList = $cellPegs{$cellID};
1012                                    for my $cellPeg (@$cellList) {
1013                                        $loadContainsFeature->Put($cellID, $cellPeg, $clusterOf{$cellPeg});
1014                          }                          }
1015                      }                      }
1016                  }                  }
1017              }              }
1018          }          }
1019                    # Now we need to generate the subsets. The subset names must be concatenated to
1020                    # the subsystem name to make them unique keys. There are two types of subsets:
1021                    # genome subsets and role subsets. We do the role subsets first.
1022                    my @subsetNames = $sub->get_subset_names();
1023                    for my $subsetID (@subsetNames) {
1024                        # Create the subset record.
1025                        my $actualID = "$subsysID:$subsetID";
1026                        $loadRoleSubset->Put($actualID);
1027                        # Connect the subset to the subsystem.
1028                        $loadHasRoleSubset->Put($subsysID, $actualID);
1029                        # Connect the subset to its roles.
1030                        my @roles = $sub->get_subsetC_roles($subsetID);
1031                        for my $roleID (@roles) {
1032                            $loadConsistsOfRoles->Put($actualID, $roleID);
1033      }      }
     # Finish the load.  
     my $retVal = $self->_FinishAll();  
     return $retVal;  
1034  }  }
1035                    # Next the genome subsets.
1036  =head3 LoadDiagramData                  @subsetNames = $sub->get_subset_namesR();
1037                    for my $subsetID (@subsetNames) {
1038  C<< my $stats = $spl->LoadDiagramData(); >>                      # Create the subset record.
1039                        my $actualID = "$subsysID:$subsetID";
1040  Load the diagram data from FIG into Sprout.                      $loadGenomeSubset->Put($actualID);
1041                        # Connect the subset to the subsystem.
1042  Diagrams are used to organize functional roles. The diagram shows the                      $loadHasGenomeSubset->Put($subsysID, $actualID);
1043  connections between chemicals that interact with a subsystem.                      # Connect the subset to its genomes.
1044                        my @genomes = $sub->get_subsetR($subsetID);
1045  The following relations are loaded by this method.                      for my $genomeID (@genomes) {
1046                            $loadConsistsOfGenomes->Put($actualID, $genomeID);
1047      Diagram                      }
1048      RoleOccursIn                  }
1049                }
1050  =over 4          }
1051            # Now we loop through the diagrams. We need to create the diagram records
1052  =item RETURNS          # and link each diagram to its roles. Note that only roles which occur
1053            # in subsystems (and therefore appear in the %ecToRoles hash) are
1054  Returns a statistics object for the loads.          # included.
1055            for my $map (@maps) {
 =back  
   
 =cut  
 #: Return Type $%;  
 sub LoadDiagramData {  
     # Get this object instance.  
     my ($self) = @_;  
     # Get the FIG object.  
     my $fig = $self->{fig};  
     # Get the map list.  
     my @maps = $fig->all_maps;  
     my $mapCount = @maps;  
     my $genomeCount = (keys %{$self->{genomes}});  
     my $featureCount = $genomeCount * 4000;  
     # Create load objects for each of the tables we're loading.  
     my $loadDiagram = $self->_TableLoader('Diagram', $mapCount);  
     my $loadRoleOccursIn = $self->_TableLoader('RoleOccursIn', $featureCount * 6);  
     Trace("Beginning diagram data load.") if T(2);  
     # Loop through the diagrams.  
     for my $map ($fig->all_maps) {  
1056          Trace("Loading diagram $map.") if T(3);          Trace("Loading diagram $map.") if T(3);
1057          # Get the diagram's descriptive name.          # Get the diagram's descriptive name.
1058          my $name = $fig->map_name($map);          my $name = $fig->map_name($map);
# Line 753  Line 1060 
1060          # Now we need to link all the map's roles to it.          # Now we need to link all the map's roles to it.
1061          # A hash is used to prevent duplicates.          # A hash is used to prevent duplicates.
1062          my %roleHash = ();          my %roleHash = ();
1063          for my $role ($fig->map_to_ecs($map)) {              for my $ec ($fig->map_to_ecs($map)) {
1064                    if (exists $ecToRoles{$ec}) {
1065                        for my $role (@{$ecToRoles{$ec}}) {
1066              if (! $roleHash{$role}) {              if (! $roleHash{$role}) {
1067                  $loadRoleOccursIn->Put($role, $map);                  $loadRoleOccursIn->Put($role, $map);
1068                  $roleHash{$role} = 1;                  $roleHash{$role} = 1;
1069              }              }
1070          }          }
1071      }      }
1072                }
1073            }
1074        }
1075      # Finish the load.      # Finish the load.
1076      my $retVal = $self->_FinishAll();      my $retVal = $self->_FinishAll();
1077      return $retVal;      return $retVal;
# Line 767  Line 1079 
1079    
1080  =head3 LoadPropertyData  =head3 LoadPropertyData
1081    
1082  C<< my $stats = $spl->LoadPropertyData(); >>      my $stats = $spl->LoadPropertyData();
1083    
1084  Load the attribute data from FIG into Sprout.  Load the attribute data from FIG into Sprout.
1085    
# Line 801  Line 1113 
1113      my $fig = $self->{fig};      my $fig = $self->{fig};
1114      # Get the genome hash.      # Get the genome hash.
1115      my $genomeHash = $self->{genomes};      my $genomeHash = $self->{genomes};
     my $genomeCount = (keys %{$genomeHash});  
1116      # Create load objects for each of the tables we're loading.      # Create load objects for each of the tables we're loading.
1117      my $loadProperty = $self->_TableLoader('Property', $genomeCount * 1500);      my $loadProperty = $self->_TableLoader('Property');
1118      my $loadHasProperty = $self->_TableLoader('HasProperty', $genomeCount * 1500);      my $loadHasProperty = $self->_TableLoader('HasProperty');
1119      Trace("Beginning property data load.") if T(2);      if ($self->{options}->{loadOnly}) {
1120            Trace("Loading from existing files.") if T(2);
1121        } else {
1122            Trace("Generating property data.") if T(2);
1123      # Create a hash for storing property IDs.      # Create a hash for storing property IDs.
1124      my %propertyKeys = ();      my %propertyKeys = ();
1125      my $nextID = 1;      my $nextID = 1;
1126            # Get the attributes we intend to store in the property table.
1127            my $propKeys = $self->{propKeys};
1128      # Loop through the genomes.      # Loop through the genomes.
1129      for my $genomeID (keys %{$genomeHash}) {          for my $genomeID (sort keys %{$genomeHash}) {
1130          $loadProperty->Add("genomeIn");          $loadProperty->Add("genomeIn");
1131          # Get the genome's features. The feature ID is the first field in the              Trace("Generating properties for $genomeID.") if T(3);
1132          # tuples returned by "all_features_detailed". We use "all_features_detailed"              # Initialize a counter.
1133          # rather than "all_features" because we want all features regardless of type.              my $propertyCount = 0;
1134          my @features = map { $_->[0] } @{$fig->all_features_detailed($genomeID)};              # Get the properties for this genome's features.
1135          # Loop through the features, creating HasProperty records.              my @attributes = $fig->get_attributes("fig|$genomeID%", $propKeys);
1136          for my $fid (@features) {              Trace("Property list built for $genomeID.") if T(3);
1137              $loadProperty->Add("featureIn");              # Loop through the results, creating HasProperty records.
1138              # Get all attributes for this feature. We do this one feature at a time              for my $attributeData (@attributes) {
1139              # to insure we do not get any genome attributes.                  # Pull apart the attribute tuple.
1140              my @attributeList = $fig->get_attributes($fid, '', '', '');                  my ($fid, $key, $value, $url) = @{$attributeData};
             # Loop through the attributes.  
             for my $tuple (@attributeList) {  
                 # Get this attribute value's data. Note that we throw away the FID,  
                 # since it will always be the same as the value if "$fid".  
                 my (undef, $key, $value, $url) = @{$tuple};  
1141                  # Concatenate the key and value and check the "propertyKeys" hash to                  # Concatenate the key and value and check the "propertyKeys" hash to
1142                  # see if we already have an ID for it. We use a tab for the separator                  # see if we already have an ID for it. We use a tab for the separator
1143                  # character.                  # character.
# Line 844  Line 1155 
1155                  # Create the HasProperty entry for this feature/property association.                  # Create the HasProperty entry for this feature/property association.
1156                  $loadHasProperty->Put($fid, $propertyID, $url);                  $loadHasProperty->Put($fid, $propertyID, $url);
1157              }              }
1158                # Update the statistics.
1159                Trace("$propertyCount attributes processed.") if T(3);
1160                $loadHasProperty->Add("propertiesIn", $propertyCount);
1161          }          }
1162      }      }
1163      # Finish the load.      # Finish the load.
# Line 853  Line 1167 
1167    
1168  =head3 LoadAnnotationData  =head3 LoadAnnotationData
1169    
1170  C<< my $stats = $spl->LoadAnnotationData(); >>      my $stats = $spl->LoadAnnotationData();
1171    
1172  Load the annotation data from FIG into Sprout.  Load the annotation data from FIG into Sprout.
1173    
# Line 885  Line 1199 
1199      my $fig = $self->{fig};      my $fig = $self->{fig};
1200      # Get the genome hash.      # Get the genome hash.
1201      my $genomeHash = $self->{genomes};      my $genomeHash = $self->{genomes};
     my $genomeCount = (keys %{$genomeHash});  
1202      # Create load objects for each of the tables we're loading.      # Create load objects for each of the tables we're loading.
1203      my $loadAnnotation = $self->_TableLoader('Annotation', $genomeCount * 4000);      my $loadAnnotation = $self->_TableLoader('Annotation');
1204      my $loadIsTargetOfAnnotation = $self->_TableLoader('IsTargetOfAnnotation', $genomeCount * 4000);      my $loadIsTargetOfAnnotation = $self->_TableLoader('IsTargetOfAnnotation');
1205      my $loadSproutUser = $self->_TableLoader('SproutUser', 100);      my $loadSproutUser = $self->_TableLoader('SproutUser');
1206      my $loadUserAccess = $self->_TableLoader('UserAccess', 1000);      my $loadUserAccess = $self->_TableLoader('UserAccess');
1207      my $loadMadeAnnotation = $self->_TableLoader('MadeAnnotation', $genomeCount * 4000);      my $loadMadeAnnotation = $self->_TableLoader('MadeAnnotation');
1208      Trace("Beginning annotation data load.") if T(2);      if ($self->{options}->{loadOnly}) {
1209            Trace("Loading from existing files.") if T(2);
1210        } else {
1211            Trace("Generating annotation data.") if T(2);
1212      # Create a hash of user names. We'll use this to prevent us from generating duplicate      # Create a hash of user names. We'll use this to prevent us from generating duplicate
1213      # user records.      # user records.
1214      my %users = ( FIG => 1, master => 1 );      my %users = ( FIG => 1, master => 1 );
# Line 906  Line 1222 
1222      # Loop through the genomes.      # Loop through the genomes.
1223      for my $genomeID (sort keys %{$genomeHash}) {      for my $genomeID (sort keys %{$genomeHash}) {
1224          Trace("Processing $genomeID.") if T(3);          Trace("Processing $genomeID.") if T(3);
         # Get the genome's PEGs.  
         my @pegs = $fig->pegs_of($genomeID);  
         for my $peg (@pegs) {  
             Trace("Processing $peg.") if T(4);  
1225              # Create a hash of timestamps. We use this to prevent duplicate time stamps              # Create a hash of timestamps. We use this to prevent duplicate time stamps
1226              # from showing up for a single PEG's annotations.              # from showing up for a single PEG's annotations.
1227              my %seenTimestamps = ();              my %seenTimestamps = ();
1228              # Check for a functional assignment.              # Get the genome's annotations.
1229              my $func = $fig->function_of($peg);              my @annotations = $fig->read_all_annotations($genomeID);
1230              if ($func) {              Trace("Processing annotations.") if T(2);
1231                  # If this is NOT a hypothetical assignment, we create an              for my $tuple (@annotations) {
1232                  # assignment annotation for it.                  # Get the annotation tuple.
1233                  if (! FIG::hypo($peg)) {                  my ($peg, $timestamp, $user, $text) = @{$tuple};
                     # Note that we double the slashes so that what goes into the database is  
                     # a new-line escape sequence rather than an actual new-line.  
                     $loadAnnotation->Put("$peg:$time", $time, "FIG\\nSet function to\\n$func");  
                     $loadIsTargetOfAnnotation->Put($peg, "$peg:$time");  
                     $loadMadeAnnotation->Put("FIG", "$peg:$time");  
                     # Denote we've seen this timestamp.  
                     $seenTimestamps{$time} = 1;  
                 }  
                 # Now loop through the real annotations.  
                 for my $tuple ($fig->feature_annotations($peg, "raw")) {  
                     my ($fid, $timestamp, $user, $text) = @{$tuple};  
1234                      # Here we fix up the annotation text. "\r" is removed,                      # Here we fix up the annotation text. "\r" is removed,
1235                      # and "\t" and "\n" are escaped. Note we use the "s"                  # and "\t" and "\n" are escaped. Note we use the "gs"
1236                      # modifier so that new-lines inside the text do not                      # modifier so that new-lines inside the text do not
1237                      # stop the substitution search.                      # stop the substitution search.
1238                      $text =~ s/\r//gs;                      $text =~ s/\r//gs;
# Line 941  Line 1242 
1242                      $text =~ s/Set master function/Set FIG function/s;                      $text =~ s/Set master function/Set FIG function/s;
1243                      # Insure the time stamp is valid.                      # Insure the time stamp is valid.
1244                      if ($timestamp =~ /^\d+$/) {                      if ($timestamp =~ /^\d+$/) {
1245                          # Here it's a number. We need to insure it's unique.                      # Here it's a number. We need to insure the one we use to form
1246                          while ($seenTimestamps{$timestamp}) {                      # the key is unique.
1247                              $timestamp++;                      my $keyStamp = $timestamp;
1248                        while ($seenTimestamps{"$peg:$keyStamp"}) {
1249                            $keyStamp++;
1250                          }                          }
1251                          $seenTimestamps{$timestamp} = 1;                      my $annotationID = "$peg:$keyStamp";
1252                          my $annotationID = "$peg:$timestamp";                      $seenTimestamps{$annotationID} = 1;
1253                          # Insure the user exists.                          # Insure the user exists.
1254                          if (! $users{$user}) {                          if (! $users{$user}) {
1255                              $loadSproutUser->Put($user, "SEED user");                              $loadSproutUser->Put($user, "SEED user");
# Line 954  Line 1257 
1257                              $users{$user} = 1;                              $users{$user} = 1;
1258                          }                          }
1259                          # Generate the annotation.                          # Generate the annotation.
1260                          $loadAnnotation->Put($annotationID, $timestamp, "$user\\n$text");                      $loadAnnotation->Put($annotationID, $timestamp, $text);
1261                          $loadIsTargetOfAnnotation->Put($peg, $annotationID);                          $loadIsTargetOfAnnotation->Put($peg, $annotationID);
1262                          $loadMadeAnnotation->Put($user, $annotationID);                          $loadMadeAnnotation->Put($user, $annotationID);
1263                      } else {                      } else {
# Line 964  Line 1267 
1267                  }                  }
1268              }              }
1269          }          }
     }  
1270      # Finish the load.      # Finish the load.
1271      my $retVal = $self->_FinishAll();      my $retVal = $self->_FinishAll();
1272      return $retVal;      return $retVal;
# Line 972  Line 1274 
1274    
1275  =head3 LoadSourceData  =head3 LoadSourceData
1276    
1277  C<< my $stats = $spl->LoadSourceData(); >>      my $stats = $spl->LoadSourceData();
1278    
1279  Load the source data from FIG into Sprout.  Load the source data from FIG into Sprout.
1280    
# Line 1005  Line 1307 
1307      my $fig = $self->{fig};      my $fig = $self->{fig};
1308      # Get the genome hash.      # Get the genome hash.
1309      my $genomeHash = $self->{genomes};      my $genomeHash = $self->{genomes};
     my $genomeCount = (keys %{$genomeHash});  
1310      # Create load objects for each of the tables we're loading.      # Create load objects for each of the tables we're loading.
1311      my $loadComesFrom = $self->_TableLoader('ComesFrom', $genomeCount * 4);      my $loadComesFrom = $self->_TableLoader('ComesFrom');
1312      my $loadSource = $self->_TableLoader('Source', $genomeCount * 4);      my $loadSource = $self->_TableLoader('Source');
1313      my $loadSourceURL = $self->_TableLoader('SourceURL', $genomeCount * 8);      my $loadSourceURL = $self->_TableLoader('SourceURL');
1314      Trace("Beginning source data load.") if T(2);      if ($self->{options}->{loadOnly}) {
1315            Trace("Loading from existing files.") if T(2);
1316        } else {
1317            Trace("Generating annotation data.") if T(2);
1318      # Create hashes to collect the Source information.      # Create hashes to collect the Source information.
1319      my %sourceURL = ();      my %sourceURL = ();
1320      my %sourceDesc = ();      my %sourceDesc = ();
# Line 1024  Line 1328 
1328              chomp $line;              chomp $line;
1329              my($sourceID, $desc, $url) = split(/\t/,$line);              my($sourceID, $desc, $url) = split(/\t/,$line);
1330              $loadComesFrom->Put($genomeID, $sourceID);              $loadComesFrom->Put($genomeID, $sourceID);
1331              if ($url && ! exists $sourceURL{$genomeID}) {                  if ($url && ! exists $sourceURL{$sourceID}) {
1332                  $loadSourceURL->Put($sourceID, $url);                  $loadSourceURL->Put($sourceID, $url);
1333                  $sourceURL{$sourceID} = 1;                  $sourceURL{$sourceID} = 1;
1334              }              }
1335              if ($desc && ! exists $sourceDesc{$sourceID}) {                  if ($desc) {
1336                  $loadSource->Put($sourceID, $desc);                      $sourceDesc{$sourceID} = $desc;
1337                  $sourceDesc{$sourceID} = 1;                  } elsif (! exists $sourceDesc{$sourceID}) {
1338                        $sourceDesc{$sourceID} = $sourceID;
1339              }              }
1340          }          }
1341          close TMP;          close TMP;
1342      }      }
1343            # Write the source descriptions.
1344            for my $sourceID (keys %sourceDesc) {
1345                $loadSource->Put($sourceID, $sourceDesc{$sourceID});
1346            }
1347        }
1348      # Finish the load.      # Finish the load.
1349      my $retVal = $self->_FinishAll();      my $retVal = $self->_FinishAll();
1350      return $retVal;      return $retVal;
# Line 1042  Line 1352 
1352    
1353  =head3 LoadExternalData  =head3 LoadExternalData
1354    
1355  C<< my $stats = $spl->LoadExternalData(); >>      my $stats = $spl->LoadExternalData();
1356    
1357  Load the external data from FIG into Sprout.  Load the external data from FIG into Sprout.
1358    
# Line 1074  Line 1384 
1384      my $fig = $self->{fig};      my $fig = $self->{fig};
1385      # Get the genome hash.      # Get the genome hash.
1386      my $genomeHash = $self->{genomes};      my $genomeHash = $self->{genomes};
     my $genomeCount = (keys %{$genomeHash});  
1387      # Convert the genome hash. We'll get the genus and species for each genome and make      # Convert the genome hash. We'll get the genus and species for each genome and make
1388      # it the key.      # it the key.
1389      my %speciesHash = map { $fig->genus_species($_) => $_ } (keys %{$genomeHash});      my %speciesHash = map { $fig->genus_species($_) => $_ } (keys %{$genomeHash});
1390      # Create load objects for each of the tables we're loading.      # Create load objects for each of the tables we're loading.
1391      my $loadExternalAliasFunc = $self->_TableLoader('ExternalAliasFunc', $genomeCount * 4000);      my $loadExternalAliasFunc = $self->_TableLoader('ExternalAliasFunc');
1392      my $loadExternalAliasOrg = $self->_TableLoader('ExternalAliasOrg', $genomeCount * 4000);      my $loadExternalAliasOrg = $self->_TableLoader('ExternalAliasOrg');
1393      Trace("Beginning external data load.") if T(2);      if ($self->{options}->{loadOnly}) {
1394            Trace("Loading from existing files.") if T(2);
1395        } else {
1396            Trace("Generating external data.") if T(2);
1397      # We loop through the files one at a time. First, the organism file.      # We loop through the files one at a time. First, the organism file.
1398      Open(\*ORGS, "<$FIG_Config::global/ext_org.table");          Open(\*ORGS, "sort +0 -1 -u -t\"\t\" $FIG_Config::global/ext_org.table |");
1399      my $orgLine;      my $orgLine;
1400      while (defined($orgLine = <ORGS>)) {      while (defined($orgLine = <ORGS>)) {
1401          # Clean the input line.          # Clean the input line.
# Line 1095  Line 1407 
1407      close ORGS;      close ORGS;
1408      # Now the function file.      # Now the function file.
1409      my $funcLine;      my $funcLine;
1410      Open(\*FUNCS, "<$FIG_Config::global/ext_func.table");          Open(\*FUNCS, "sort +0 -1 -u -t\"\t\" $FIG_Config::global/ext_func.table |");
1411      while (defined($funcLine = <FUNCS>)) {      while (defined($funcLine = <FUNCS>)) {
1412          # Clean the line ending.          # Clean the line ending.
1413          chomp $funcLine;          chomp $funcLine;
# Line 1111  Line 1423 
1423              $loadExternalAliasFunc->Put(@funcFields[0,1]);              $loadExternalAliasFunc->Put(@funcFields[0,1]);
1424          }          }
1425      }      }
1426        }
1427      # Finish the load.      # Finish the load.
1428      my $retVal = $self->_FinishAll();      my $retVal = $self->_FinishAll();
1429      return $retVal;      return $retVal;
1430  }  }
1431    
 =head3 LoadGroupData  
1432    
1433  C<< my $stats = $spl->LoadGroupData(); >>  =head3 LoadReactionData
1434    
1435        my $stats = $spl->LoadReactionData();
1436    
1437    Load the reaction data from FIG into Sprout.
1438    
1439  Load the genome Groups into Sprout.  Reaction data connects reactions to the compounds that participate in them.
1440    
1441  The following relations are loaded by this method.  The following relations are loaded by this method.
1442    
1443      GenomeGroups      Reaction
1444        ReactionURL
1445        Compound
1446        CompoundName
1447        CompoundCAS
1448        IsIdentifiedByCAS
1449        HasCompoundName
1450        IsAComponentOf
1451        Scenario
1452        Catalyzes
1453        HasScenario
1454        IsInputFor
1455        IsOutputOf
1456        ExcludesReaction
1457        IncludesReaction
1458        IsOnDiagram
1459        IncludesReaction
1460    
1461  There is no direct support for genome groups in FIG, so we access the SEED  This method proceeds reaction by reaction rather than genome by genome.
 files directly.  
1462    
1463  =over 4  =over 4
1464    
# Line 1139  Line 1470 
1470    
1471  =cut  =cut
1472  #: Return Type $%;  #: Return Type $%;
1473  sub LoadGroupData {  sub LoadReactionData {
1474        # Get this object instance.
1475        my ($self) = @_;
1476        # Get the FIG object.
1477        my $fig = $self->{fig};
1478        # Create load objects for each of the tables we're loading.
1479        my $loadReaction = $self->_TableLoader('Reaction');
1480        my $loadReactionURL = $self->_TableLoader('ReactionURL');
1481        my $loadCompound = $self->_TableLoader('Compound');
1482        my $loadCompoundName = $self->_TableLoader('CompoundName');
1483        my $loadCompoundCAS = $self->_TableLoader('CompoundCAS');
1484        my $loadIsAComponentOf = $self->_TableLoader('IsAComponentOf');
1485        my $loadIsIdentifiedByCAS = $self->_TableLoader('IsIdentifiedByCAS');
1486        my $loadHasCompoundName = $self->_TableLoader('HasCompoundName');
1487        my $loadScenario = $self->_TableLoader('Scenario');
1488        my $loadHasScenario = $self->_TableLoader('HasScenario');
1489        my $loadIsInputFor = $self->_TableLoader('IsInputFor');
1490        my $loadIsOutputOf = $self->_TableLoader('IsOutputOf');
1491        my $loadIsOnDiagram = $self->_TableLoader('IsOnDiagram');
1492        my $loadIncludesReaction = $self->_TableLoader('IncludesReaction');
1493        my $loadExcludesReaction = $self->_TableLoader('ExcludesReaction');
1494        my $loadCatalyzes = $self->_TableLoader('Catalyzes');
1495        if ($self->{options}->{loadOnly}) {
1496            Trace("Loading from existing files.") if T(2);
1497        } else {
1498            Trace("Generating reaction data.") if T(2);
1499            # We need some hashes to prevent duplicates.
1500            my %compoundNames = ();
1501            my %compoundCASes = ();
1502            # First we create the compounds.
1503            my %compounds = map { $_ => 1 } $fig->all_compounds();
1504            for my $cid (keys %compounds) {
1505                # Check for names.
1506                my @names = $fig->names_of_compound($cid);
1507                # Each name will be given a priority number, starting with 1.
1508                my $prio = 1;
1509                for my $name (@names) {
1510                    if (! exists $compoundNames{$name}) {
1511                        $loadCompoundName->Put($name);
1512                        $compoundNames{$name} = 1;
1513                    }
1514                    $loadHasCompoundName->Put($cid, $name, $prio++);
1515                }
1516                # Create the main compound record. Note that the first name
1517                # becomes the label.
1518                my $label = (@names > 0 ? $names[0] : $cid);
1519                $loadCompound->Put($cid, $label);
1520                # Check for a CAS ID.
1521                my $cas = $fig->cas($cid);
1522                if ($cas) {
1523                    $loadIsIdentifiedByCAS->Put($cid, $cas);
1524                    if (! exists $compoundCASes{$cas}) {
1525                        $loadCompoundCAS->Put($cas);
1526                        $compoundCASes{$cas} = 1;
1527                    }
1528                }
1529            }
1530            # All the compounds are set up, so we need to loop through the reactions next. First,
1531            # we initialize the discriminator index. This is a single integer used to insure
1532            # duplicate elements in a reaction are not accidentally collapsed.
1533            my $discrim = 0;
1534            my %reactions = map { $_ => 1 } $fig->all_reactions();
1535            for my $reactionID (keys %reactions) {
1536                # Create the reaction record.
1537                $loadReaction->Put($reactionID, $fig->reversible($reactionID));
1538                # Compute the reaction's URL.
1539                my $url = HTML::reaction_link($reactionID);
1540                # Put it in the ReactionURL table.
1541                $loadReactionURL->Put($reactionID, $url);
1542                # Now we need all of the reaction's compounds. We get these in two phases,
1543                # substrates first and then products.
1544                for my $product (0, 1) {
1545                    # Get the compounds of the current type for the current reaction. FIG will
1546                    # give us 3-tuples: [ID, stoichiometry, main-flag]. At this time we do not
1547                    # have location data in SEED, so it defaults to the empty string.
1548                    my @compounds = $fig->reaction2comp($reactionID, $product);
1549                    for my $compData (@compounds) {
1550                        # Extract the compound data from the current tuple.
1551                        my ($cid, $stoich, $main) = @{$compData};
1552                        # Link the compound to the reaction.
1553                        $loadIsAComponentOf->Put($cid, $reactionID, $discrim++, "", $main,
1554                                                 $product, $stoich);
1555                    }
1556                }
1557            }
1558            # Now we run through the subsystems and roles, generating the scenarios
1559            # and connecting the reactions. We'll need some hashes to prevent
1560            # duplicates and a counter for compound group keys.
1561            my %roles = ();
1562            my %scenarios = ();
1563            my @subsystems = $fig->all_subsystems();
1564            for my $subName (@subsystems) {
1565                my $sub = $fig->get_subsystem($subName);
1566                Trace("Processing $subName reactions.") if T(3);
1567                # Get the subsystem's reactions.
1568                my %reactions = $sub->get_hope_reactions();
1569                # Loop through the roles, connecting them to the reactions.
1570                for my $role (keys %reactions) {
1571                    # Only process this role if it is new.
1572                    if (! $roles{$role}) {
1573                        $roles{$role} = 1;
1574                        my @reactions = @{$reactions{$role}};
1575                        for my $reaction (@reactions) {
1576                            $loadCatalyzes->Put($role, $reaction);
1577                        }
1578                    }
1579                }
1580                Trace("Processing $subName scenarios.") if T(3);
1581                # Get the subsystem's scenarios.
1582                my @scenarioNames = $sub->get_hope_scenario_names();
1583                # Loop through the scenarios, creating scenario data.
1584                for my $scenarioName (@scenarioNames) {
1585                    # Link this scenario to this subsystem.
1586                    $loadHasScenario->Put($subName, $scenarioName);
1587                    # If this scenario is new, we need to create it.
1588                    if (! $scenarios{$scenarioName}) {
1589                        Trace("Creating scenario $scenarioName.") if T(3);
1590                        $scenarios{$scenarioName} = 1;
1591                        # Create the scenario itself.
1592                        $loadScenario->Put($scenarioName);
1593                        # Attach the input compounds.
1594                        for my $input ($sub->get_hope_input_compounds($scenarioName)) {
1595                            $loadIsInputFor->Put($input, $scenarioName);
1596                        }
1597                        # Now we need to set up the output compounds. They come in two
1598                        # groups, which we mark 0 and 1.
1599                        my $outputGroup = 0;
1600                        # Set up the output compounds.
1601                        for my $outputGroup ($sub->get_hope_output_compounds($scenarioName)) {
1602                            # Attach the compounds.
1603                            for my $compound (@$outputGroup) {
1604                                $loadIsOutputOf->Put($scenarioName, $compound, $outputGroup);
1605                            }
1606                        }
1607                        # Create the reaction lists.
1608                        my @addReactions = $sub->get_hope_additional_reactions($scenarioName);
1609                        for my $reaction (@addReactions) {
1610                            $loadIncludesReaction->Put($scenarioName, $reaction);
1611                        }
1612                        my @notReactions = $sub->get_hope_ignore_reactions($scenarioName);
1613                        for my $reaction (@notReactions) {
1614                            $loadExcludesReaction->Put($scenarioName, $reaction);
1615                        }
1616                        # Link the maps.
1617                        my @maps = $sub->get_hope_map_ids($scenarioName);
1618                        for my $map (@maps) {
1619                            $loadIsOnDiagram->Put($scenarioName, "map$map");
1620                        }
1621                    }
1622                }
1623            }
1624        }
1625        # Finish the load.
1626        my $retVal = $self->_FinishAll();
1627        return $retVal;
1628    }
1629    
1630    =head3 LoadSynonymData
1631    
1632        my $stats = $spl->LoadSynonymData();
1633    
1634    Load the synonym groups into Sprout.
1635    
1636    The following relations are loaded by this method.
1637    
1638        SynonymGroup
1639        IsSynonymGroupFor
1640    
1641    The source information for these relations is taken from the C<maps_to_id> method
1642    of the B<FIG> object. Unfortunately, to make this work, we need to use direct
1643    SQL against the FIG database.
1644    
1645    =over 4
1646    
1647    =item RETURNS
1648    
1649    Returns a statistics object for the loads.
1650    
1651    =back
1652    
1653    =cut
1654    #: Return Type $%;
1655    sub LoadSynonymData {
1656      # Get this object instance.      # Get this object instance.
1657      my ($self) = @_;      my ($self) = @_;
1658      # Get the FIG object.      # Get the FIG object.
1659      my $fig = $self->{fig};      my $fig = $self->{fig};
1660      # Get the genome hash.      # Get the genome hash.
1661      my $genomeHash = $self->{genomes};      my $genomeHash = $self->{genomes};
     my $genomeCount = (keys %{$genomeHash});  
1662      # Create a load object for the table we're loading.      # Create a load object for the table we're loading.
1663      my $loadGenomeGroups = $self->_TableLoader('GenomeGroups', $genomeCount * 4);      my $loadSynonymGroup = $self->_TableLoader('SynonymGroup');
1664      Trace("Beginning group data load.") if T(2);      my $loadIsSynonymGroupFor = $self->_TableLoader('IsSynonymGroupFor');
1665        if ($self->{options}->{loadOnly}) {
1666            Trace("Loading from existing files.") if T(2);
1667        } else {
1668            Trace("Generating synonym group data.") if T(2);
1669            # Get the database handle.
1670            my $dbh = $fig->db_handle();
1671            # Ask for the synonyms. Note that "maps_to" is a group name, and "syn_id" is a PEG ID or alias.
1672            my $sth = $dbh->prepare_command("SELECT maps_to, syn_id FROM peg_synonyms ORDER BY maps_to");
1673            my $result = $sth->execute();
1674            if (! defined($result)) {
1675                Confess("Database error in Synonym load: " . $sth->errstr());
1676            } else {
1677                Trace("Processing synonym results.") if T(2);
1678                # Remember the current synonym.
1679                my $current_syn = "";
1680                # Count the features.
1681                my $featureCount = 0;
1682                my $entryCount = 0;
1683                # Loop through the synonym/peg pairs.
1684                while (my @row = $sth->fetchrow()) {
1685                    # Get the synonym group ID and feature ID.
1686                    my ($syn_id, $peg) = @row;
1687                    # Count this row.
1688                    $entryCount++;
1689                    if ($entryCount % 1000 == 0) {
1690                        Trace("$entryCount rows processed.") if T(3);
1691                    }
1692                    # Insure it's for one of our genomes.
1693                    my $genomeID = FIG::genome_of($peg);
1694                    if (exists $genomeHash->{$genomeID}) {
1695                        # Verify the synonym.
1696                        if ($syn_id ne $current_syn) {
1697                            # It's new, so put it in the group table.
1698                            $loadSynonymGroup->Put($syn_id);
1699                            $current_syn = $syn_id;
1700                        }
1701                        # Connect the synonym to the peg.
1702                        $loadIsSynonymGroupFor->Put($syn_id, $peg);
1703                        # Count this feature.
1704                        $featureCount++;
1705                        if ($featureCount % 1000 == 0) {
1706                            Trace("$featureCount features processed.") if T(3);
1707                        }
1708                    }
1709                }
1710                Trace("$entryCount rows produced $featureCount features.") if T(2);
1711            }
1712        }
1713        # Finish the load.
1714        my $retVal = $self->_FinishAll();
1715        return $retVal;
1716    }
1717    
1718    =head3 LoadFamilyData
1719    
1720        my $stats = $spl->LoadFamilyData();
1721    
1722    Load the protein families into Sprout.
1723    
1724    The following relations are loaded by this method.
1725    
1726        Family
1727        IsFamilyForFeature
1728    
1729    The source information for these relations is taken from the C<families_for_protein>,
1730    C<family_function>, and C<sz_family> methods of the B<FIG> object.
1731    
1732    =over 4
1733    
1734    =item RETURNS
1735    
1736    Returns a statistics object for the loads.
1737    
1738    =back
1739    
1740    =cut
1741    #: Return Type $%;
1742    sub LoadFamilyData {
1743        # Get this object instance.
1744        my ($self) = @_;
1745        # Get the FIG object.
1746        my $fig = $self->{fig};
1747        # Get the genome hash.
1748        my $genomeHash = $self->{genomes};
1749        # Create load objects for the tables we're loading.
1750        my $loadFamily = $self->_TableLoader('Family');
1751        my $loadIsFamilyForFeature = $self->_TableLoader('IsFamilyForFeature');
1752        if ($self->{options}->{loadOnly}) {
1753            Trace("Loading from existing files.") if T(2);
1754        } else {
1755            Trace("Generating family data.") if T(2);
1756            # Create a hash for the family IDs.
1757            my %familyHash = ();
1758      # Loop through the genomes.      # Loop through the genomes.
1759      my $line;          for my $genomeID (sort keys %{$genomeHash}) {
1760      for my $genomeID (keys %{$genomeHash}) {              Trace("Processing features for $genomeID.") if T(2);
1761          Trace("Processing $genomeID.") if T(3);              # Loop through this genome's PEGs.
1762          # Open the NMPDR group file for this genome.              for my $fid ($fig->all_features($genomeID, "peg")) {
1763          if (open(TMP, "<$FIG_Config::organisms/$genomeID/NMPDR") &&                  $loadIsFamilyForFeature->Add("features", 1);
1764              defined($line = <TMP>)) {                  # Get this feature's families.
1765              # Clean the line ending.                  my @families = $fig->families_for_protein($fid);
1766              chomp $line;                  # Loop through the families, connecting them to the feature.
1767              # Add the group to the table. Note that there can only be one group                  for my $family (@families) {
1768              # per genome.                      $loadIsFamilyForFeature->Put($family, $fid);
1769              $loadGenomeGroups->Put($genomeID, $line);                      # If this is a new family, create a record for it.
1770                        if (! exists $familyHash{$family}) {
1771                            $familyHash{$family} = 1;
1772                            $loadFamily->Add("families", 1);
1773                            my $size = $fig->sz_family($family);
1774                            my $func = $fig->family_function($family);
1775                            $loadFamily->Put($family, $size, $func);
1776          }          }
1777          close TMP;                  }
1778                }
1779            }
1780        }
1781        # Finish the load.
1782        my $retVal = $self->_FinishAll();
1783        return $retVal;
1784    }
1785    
1786    =head3 LoadDrugData
1787    
1788        my $stats = $spl->LoadDrugData();
1789    
1790    Load the drug target data into Sprout.
1791    
1792    The following relations are loaded by this method.
1793    
1794        PDB
1795        DocksWith
1796        IsProteinForFeature
1797        Ligand
1798    
1799    The source information for these relations is taken from attributes. The
1800    C<PDB> attribute links a PDB to a feature, and is used to build B<IsProteinForFeature>.
1801    The C<zinc_name> attribute describes the ligands. The C<docking_results>
1802    attribute contains the information for the B<DocksWith> relationship. It is
1803    expected that additional attributes and tables will be added in the future.
1804    
1805    =over 4
1806    
1807    =item RETURNS
1808    
1809    Returns a statistics object for the loads.
1810    
1811    =back
1812    
1813    =cut
1814    #: Return Type $%;
1815    sub LoadDrugData {
1816        # Get this object instance.
1817        my ($self) = @_;
1818        # Get the FIG object.
1819        my $fig = $self->{fig};
1820        # Get the genome hash.
1821        my $genomeHash = $self->{genomes};
1822        # Create load objects for the tables we're loading.
1823        my $loadPDB = $self->_TableLoader('PDB');
1824        my $loadLigand = $self->_TableLoader('Ligand');
1825        my $loadIsProteinForFeature = $self->_TableLoader('IsProteinForFeature');
1826        my $loadDocksWith = $self->_TableLoader('DocksWith');
1827        if ($self->{options}->{loadOnly}) {
1828            Trace("Loading from existing files.") if T(2);
1829        } else {
1830            Trace("Generating drug target data.") if T(2);
1831            # First comes the "DocksWith" relationship. This will give us a list of PDBs.
1832            # We can also encounter PDBs when we process "IsProteinForFeature". To manage
1833            # this process, PDB information is collected in a hash table and then
1834            # unspooled after both relationships are created.
1835            my %pdbHash = ();
1836            Trace("Generating docking data.") if T(2);
1837            # Get all the docking data. This may cause problems if there are too many PDBs,
1838            # at which point we'll need another algorithm. The indicator that this is
1839            # happening will be a timeout error in the next statement.
1840            my @dockData = $fig->query_attributes('$key = ? AND $value < ?',
1841                                                  ['docking_results', $FIG_Config::dockLimit]);
1842            Trace(scalar(@dockData) . " rows of docking data found.") if T(3);
1843            for my $dockData (@dockData) {
1844                # Get the docking data components.
1845                my ($pdbID, $docking_key, @valueData) = @{$dockData};
1846                # Fix the PDB ID. It's supposed to be lower-case, but this does not always happen.
1847                $pdbID = lc $pdbID;
1848                # Strip off the object type.
1849                $pdbID =~ s/pdb://;
1850                # Extract the ZINC ID from the docking key. Note that there are two possible
1851                # formats.
1852                my (undef, $zinc_id) = $docking_key =~ /^docking_results::(ZINC)?(\d+)$/;
1853                if (! $zinc_id) {
1854                    Trace("Invalid docking result key $docking_key for $pdbID.") if T(0);
1855                    $loadDocksWith->Add("errors");
1856                } else {
1857                    # Get the pieces of the value and parse the energy.
1858                    # Note that we don't care about the rank, since
1859                    # we can sort on the energy level itself in our database.
1860                    my ($energy, $tool, $type) = @valueData;
1861                    my ($rank, $total, $vanderwaals, $electrostatic) = split /\s*;\s*/, $energy;
1862                    # Ignore predicted results.
1863                    if ($type ne "Predicted") {
1864                        # Count this docking result.
1865                        if (! exists $pdbHash{$pdbID}) {
1866                            $pdbHash{$pdbID} = 1;
1867                        } else {
1868                            $pdbHash{$pdbID}++;
1869                        }
1870                        # Write the result to the output.
1871                        $loadDocksWith->Put($pdbID, $zinc_id, $electrostatic, $type, $tool,
1872                                            $total, $vanderwaals);
1873                    }
1874                }
1875            }
1876            Trace("Connecting features.") if T(2);
1877            # Loop through the genomes.
1878            for my $genome (sort keys %{$genomeHash}) {
1879                Trace("Generating PDBs for $genome.") if T(3);
1880                # Get all of the PDBs that BLAST against this genome's features.
1881                my @attributeData = $fig->get_attributes("fig|$genome%", 'PDB::%');
1882                for my $pdbData (@attributeData) {
1883                    # The PDB ID is coded as a subkey.
1884                    if ($pdbData->[1] !~ /PDB::(.+)/i) {
1885                        Trace("Invalid PDB ID \"$pdbData->[1]\" in attribute table.") if T(0);
1886                        $loadPDB->Add("errors");
1887                    } else {
1888                        my $pdbID = $1;
1889                        # Insure the PDB is in the hash.
1890                        if (! exists $pdbHash{$pdbID}) {
1891                            $pdbHash{$pdbID} = 0;
1892                        }
1893                        # The score and locations are coded in the attribute value.
1894                        if ($pdbData->[2] !~ /^([^;]+)(.*)$/) {
1895                            Trace("Invalid PDB data for $pdbID and feature $pdbData->[0].") if T(0);
1896                            $loadIsProteinForFeature->Add("errors");
1897                        } else {
1898                            my ($score, $locData) = ($1,$2);
1899                            # The location data may not be present, so we have to start with some
1900                            # defaults and then check.
1901                            my ($start, $end) = (1, 0);
1902                            if ($locData) {
1903                                $locData =~ /(\d+)-(\d+)/;
1904                                $start = $1;
1905                                $end = $2;
1906                            }
1907                            # If we still don't have the end location, compute it from
1908                            # the feature length.
1909                            if (! $end) {
1910                                # Most features have one location, but we do a list iteration
1911                                # just in case.
1912                                my @locations = $fig->feature_location($pdbData->[0]);
1913                                $end = 0;
1914                                for my $loc (@locations) {
1915                                    my $locObject = BasicLocation->new($loc);
1916                                    $end += $locObject->Length;
1917                                }
1918                            }
1919                            # Decode the score.
1920                            my $realScore = FIGRules::DecodeScore($score);
1921                            # Connect the PDB to the feature.
1922                            $loadIsProteinForFeature->Put($pdbID, $pdbData->[0], $start, $realScore, $end);
1923                        }
1924                    }
1925                }
1926            }
1927            # We've got all our PDBs now, so we unspool them from the hash.
1928            Trace("Generating PDBs. " . scalar(keys %pdbHash) . " found.") if T(2);
1929            my $count = 0;
1930            for my $pdbID (sort keys %pdbHash) {
1931                $loadPDB->Put($pdbID, $pdbHash{$pdbID});
1932                $count++;
1933                Trace("$count PDBs processed.") if T(3) && ($count % 500 == 0);
1934            }
1935            # Finally we create the ligand table. This information can be found in the
1936            # zinc_name attribute.
1937            Trace("Loading ligands.") if T(2);
1938            # The ligand list is huge, so we have to get it in pieces. We also have to check for duplicates.
1939            my $last_zinc_id = "";
1940            my $zinc_id = "";
1941            my $done = 0;
1942            while (! $done) {
1943                # Get the next 10000 ligands. We insist that the object ID is greater than
1944                # the last ID we processed.
1945                Trace("Loading batch starting with ZINC:$zinc_id.") if T(3);
1946                my @attributeData = $fig->query_attributes('$object > ? AND $key = ? ORDER BY $object LIMIT 10000',
1947                                                           ["ZINC:$zinc_id", "zinc_name"]);
1948                Trace(scalar(@attributeData) . " attribute rows returned.") if T(3);
1949                if (! @attributeData) {
1950                    # Here there are no attributes left, so we quit the loop.
1951                    $done = 1;
1952                } else {
1953                    # Process the attribute data we've received.
1954                    for my $zinc_data (@attributeData) {
1955                        # The ZINC ID is found in the first return column, prefixed with the word ZINC.
1956                        if ($zinc_data->[0] =~ /^ZINC:(\d+)$/) {
1957                            $zinc_id = $1;
1958                            # Check for a duplicate.
1959                            if ($zinc_id eq $last_zinc_id) {
1960                                $loadLigand->Add("duplicate");
1961                            } else {
1962                                # Here it's safe to output the ligand. The ligand name is the attribute value
1963                                # (third column in the row).
1964                                $loadLigand->Put($zinc_id, $zinc_data->[2]);
1965                                # Insure we don't try to add this ID again.
1966                                $last_zinc_id = $zinc_id;
1967                            }
1968                        } else {
1969                            Trace("Invalid zinc ID \"$zinc_data->[0]\" in attribute table.") if T(0);
1970                            $loadLigand->Add("errors");
1971                        }
1972                    }
1973                }
1974            }
1975            Trace("Ligands loaded.") if T(2);
1976      }      }
1977      # Finish the load.      # Finish the load.
1978      my $retVal = $self->_FinishAll();      my $retVal = $self->_FinishAll();
1979      return $retVal;      return $retVal;
1980  }  }
1981    
1982    
1983  =head2 Internal Utility Methods  =head2 Internal Utility Methods
1984    
1985    =head3 SpecialAttribute
1986    
1987        my $count = SproutLoad::SpecialAttribute($id, \@attributes, $idxMatch, \@idxValues, $pattern, $loader);
1988    
1989    Look for special attributes of a given type. A special attribute is found by comparing one of
1990    the columns of the incoming attribute list to a search pattern. If a match is found, then
1991    a set of columns is put into an output table connected to the specified ID.
1992    
1993    For example, when processing features, the attribute list we look at has three columns: attribute
1994    name, attribute value, and attribute value HTML. The IEDB attribute exists if the attribute name
1995    begins with C<iedb_>. The call signature is therefore
1996    
1997        my $found = SpecialAttribute($fid, \@attributeList, 0, [0,2], '^iedb_', $loadFeatureIEDB);
1998    
1999    The pattern is matched against column 0, and if we have a match, then column 2's value is put
2000    to the output along with the specified feature ID.
2001    
2002    =over 4
2003    
2004    =item id
2005    
2006    ID of the object whose special attributes are being loaded. This forms the first column of the
2007    output.
2008    
2009    =item attributes
2010    
2011    Reference to a list of tuples.
2012    
2013    =item idxMatch
2014    
2015    Index in each tuple of the column to be matched against the pattern. If the match is
2016    successful, an output record will be generated.
2017    
2018    =item idxValues
2019    
2020    Reference to a list containing the indexes in each tuple of the columns to be put as
2021    the second column of the output.
2022    
2023    =item pattern
2024    
2025    Pattern to be matched against the specified column. The match will be case-insensitive.
2026    
2027    =item loader
2028    
2029    An object to which each output record will be put. Usually this is an B<ERDBLoad> object,
2030    but technically it could be anything with a C<Put> method.
2031    
2032    =item RETURN
2033    
2034    Returns a count of the matches found.
2035    
2036    =item
2037    
2038    =back
2039    
2040    =cut
2041    
2042    sub SpecialAttribute {
2043        # Get the parameters.
2044        my ($id, $attributes, $idxMatch, $idxValues, $pattern, $loader) = @_;
2045        # Declare the return variable.
2046        my $retVal = 0;
2047        # Loop through the attribute rows.
2048        for my $row (@{$attributes}) {
2049            # Check for a match.
2050            if ($row->[$idxMatch] =~ m/$pattern/i) {
2051                # We have a match, so output a row. This is a bit tricky, since we may
2052                # be putting out multiple columns of data from the input.
2053                my $value = join(" ", map { $row->[$_] } @{$idxValues});
2054                $loader->Put($id, $value);
2055                $retVal++;
2056            }
2057        }
2058        Trace("$retVal special attributes found for $id and loader " . $loader->RelName() . ".") if T(4) && $retVal;
2059        # Return the number of matches.
2060        return $retVal;
2061    }
2062    
2063  =head3 TableLoader  =head3 TableLoader
2064    
2065  Create an ERDBLoad object for the specified table. The object is also added to  Create an ERDBLoad object for the specified table. The object is also added to
# Line 1186  Line 2074 
2074    
2075  Name of the table (relation) being loaded.  Name of the table (relation) being loaded.
2076    
 =item rowCount (optional)  
   
 Estimated maximum number of rows in the table.  
   
2077  =item RETURN  =item RETURN
2078    
2079  Returns an ERDBLoad object for loading the specified table.  Returns an ERDBLoad object for loading the specified table.
# Line 1200  Line 2084 
2084    
2085  sub _TableLoader {  sub _TableLoader {
2086      # Get the parameters.      # Get the parameters.
2087      my ($self, $tableName, $rowCount) = @_;      my ($self, $tableName) = @_;
2088      # Create the load object.      # Create the load object.
2089      my $retVal = ERDBLoad->new($self->{erdb}, $tableName, $self->{loadDirectory}, $rowCount);      my $retVal = ERDBLoad->new($self->{erdb}, $tableName, $self->{loadDirectory}, $self->LoadOnly);
2090      # Cache it in the loader list.      # Cache it in the loader list.
2091      push @{$self->{loaders}}, $retVal;      push @{$self->{loaders}}, $retVal;
2092      # Return it to the caller.      # Return it to the caller.
# Line 1236  Line 2120 
2120      my $retVal = Stats->new();      my $retVal = Stats->new();
2121      # Get the loader list.      # Get the loader list.
2122      my $loadList = $self->{loaders};      my $loadList = $self->{loaders};
2123        # Create a hash to hold the statistics objects, keyed on relation name.
2124        my %loaderHash = ();
2125      # Loop through the list, finishing the loads. Note that if the finish fails, we die      # Loop through the list, finishing the loads. Note that if the finish fails, we die
2126      # ignominiously. At some future point, we want to make the loads restartable.      # ignominiously. At some future point, we want to make the loads more restartable.
2127      while (my $loader = pop @{$loadList}) {      while (my $loader = pop @{$loadList}) {
2128            # Get the relation name.
2129            my $relName = $loader->RelName;
2130            # Check the ignore flag.
2131            if ($loader->Ignore) {
2132                Trace("Relation $relName not loaded.") if T(2);
2133            } else {
2134                # Here we really need to finish.
2135                Trace("Finishing $relName.") if T(2);
2136          my $stats = $loader->Finish();          my $stats = $loader->Finish();
2137                $loaderHash{$relName} = $stats;
2138            }
2139        }
2140        # Now we loop through again, actually loading the tables. We want to finish before
2141        # loading so that if something goes wrong at this point, all the load files are usable
2142        # and we don't have to redo all that work.
2143        for my $relName (sort keys %loaderHash) {
2144            # Get the statistics for this relation.
2145            my $stats = $loaderHash{$relName};
2146            # Check for a database load.
2147            if ($self->{options}->{dbLoad}) {
2148                # Here we want to use the load file just created to load the database.
2149                Trace("Loading relation $relName.") if T(2);
2150                my $newStats = $self->{sprout}->LoadUpdate(1, [$relName]);
2151                # Accumulate the statistics from the DB load.
2152                $stats->Accumulate($newStats);
2153            }
2154          $retVal->Accumulate($stats);          $retVal->Accumulate($stats);
         my $relName = $loader->RelName;  
2155          Trace("Statistics for $relName:\n" . $stats->Show()) if T(2);          Trace("Statistics for $relName:\n" . $stats->Show()) if T(2);
2156      }      }
2157      # Return the load statistics.      # Return the load statistics.
2158      return $retVal;      return $retVal;
2159  }  }
2160    
2161    =head3 GetGenomeAttributes
2162    
2163        my $aHashRef = GetGenomeAttributes($fig, $genomeID, \@fids, \@propKeys);
2164    
2165    Return a hash of attributes keyed on feature ID. This method gets all the NMPDR-related
2166    attributes for all the features of a genome in a single call, then organizes them into
2167    a hash.
2168    
2169    =over 4
2170    
2171    =item fig
2172    
2173    FIG-like object for accessing attributes.
2174    
2175    =item genomeID
2176    
2177    ID of the genome who's attributes are desired.
2178    
2179    =item fids
2180    
2181    Reference to a list of the feature IDs whose attributes are to be kept.
2182    
2183    =item propKeys
2184    
2185    A list of the keys to retrieve.
2186    
2187    =item RETURN
2188    
2189    Returns a reference to a hash. The key of the hash is the feature ID. The value is the
2190    reference to a list of the feature's attribute tuples. Each tuple contains the feature ID,
2191    the attribute key, and one or more attribute values.
2192    
2193    =back
2194    
2195    =cut
2196    
2197    sub GetGenomeAttributes {
2198        # Get the parameters.
2199        my ($fig, $genomeID, $fids, $propKeys) = @_;
2200        # Declare the return variable.
2201        my $retVal = {};
2202        # Initialize the hash. This not only enables us to easily determine which FIDs to
2203        # keep, it insures that the caller sees a list reference for every known fid,
2204        # simplifying the logic.
2205        for my $fid (@{$fids}) {
2206            $retVal->{$fid} = [];
2207        }
2208        # Get the attributes. If ev_code_cron is running, we may get a timeout error, so
2209        # an eval is used.
2210        my @aList = ();
2211        eval {
2212            @aList = $fig->get_attributes("fig|$genomeID%", $propKeys);
2213            Trace(scalar(@aList) . " attributes returned for genome $genomeID.") if T(3);
2214        };
2215        # Check for a problem.
2216        if ($@) {
2217            Trace("Retrying attributes for $genomeID due to error: $@") if T(1);
2218            # Our fallback plan is to process the attributes in blocks of 100. This is much slower,
2219            # but allows us to continue processing.
2220            my $nFids = scalar @{$fids};
2221            for (my $i = 0; $i < $nFids; $i += 100) {
2222                # Determine the index of the last feature ID we'll be specifying on this pass.
2223                # Normally it's $i + 99, but if we're close to the end it may be less.
2224                my $end = ($i + 100 > $nFids ? $nFids - 1 : $i + 99);
2225                # Get a slice of the fid list.
2226                my @slice = @{$fids}[$i .. $end];
2227                # Get the relevant attributes.
2228                Trace("Retrieving attributes for fids $i to $end.") if T(3);
2229                my @aShort = $fig->get_attributes(\@slice, $propKeys);
2230                Trace(scalar(@aShort) . " attributes returned for fids $i to $end.") if T(3);
2231                push @aList, @aShort;
2232            }
2233        }
2234        # Now we should have all the interesting attributes in @aList. Populate the hash with
2235        # them.
2236        for my $aListEntry (@aList) {
2237            my $fid = $aListEntry->[0];
2238            if (exists $retVal->{$fid}) {
2239                push @{$retVal->{$fid}}, $aListEntry;
2240            }
2241        }
2242        # Return the result.
2243        return $retVal;
2244    }
2245    
2246    
2247  1;  1;

Legend:
Removed from v.1.9  
changed lines
  Added in v.1.93

MCS Webmaster
ViewVC Help
Powered by ViewVC 1.0.3