2 * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
3 * Copyright (C) $$Year-Rel$$ The Jalview Authors
5 * This file is part of Jalview.
7 * Jalview is free software: you can redistribute it and/or
8 * modify it under the terms of the GNU General Public License
9 * as published by the Free Software Foundation, either version 3
10 * of the License, or (at your option) any later version.
12 * Jalview is distributed in the hope that it will be useful, but
13 * WITHOUT ANY WARRANTY; without even the implied warranty
14 * of MERCHANTABILITY or FITNESS FOR A PARTICULAR
15 * PURPOSE. See the GNU General Public License for more details.
17 * You should have received a copy of the GNU General Public License
18 * along with Jalview. If not, see <http://www.gnu.org/licenses/>.
19 * The Jalview Authors are detailed in the 'AUTHORS' file.
23 import jalview.analysis.AlignSeq;
24 import jalview.bin.Cache;
25 import jalview.datamodel.AlignmentI;
26 import jalview.datamodel.DBRefEntry;
27 import jalview.datamodel.DBRefSource;
28 import jalview.datamodel.Mapping;
29 import jalview.datamodel.SequenceI;
30 import jalview.gui.CutAndPasteTransfer;
31 import jalview.gui.DasSourceBrowser;
32 import jalview.gui.Desktop;
33 import jalview.gui.FeatureSettings;
34 import jalview.gui.IProgressIndicator;
35 import jalview.gui.OOMWarning;
36 import jalview.util.DBRefUtils;
37 import jalview.util.MessageManager;
38 import jalview.ws.dbsources.das.api.jalviewSourceI;
39 import jalview.ws.dbsources.das.datamodel.DasSequenceSource;
40 import jalview.ws.seqfetcher.DbSourceProxy;
42 import java.util.ArrayList;
43 import java.util.Arrays;
44 import java.util.Enumeration;
45 import java.util.Hashtable;
46 import java.util.List;
47 import java.util.StringTokenizer;
48 import java.util.Vector;
50 import uk.ac.ebi.picr.model.UPEntry;
51 import uk.ac.ebi.www.picr.AccessionMappingService.AccessionMapperServiceLocator;
54 * Implements a runnable for validating a sequence against external databases
55 * and then propagating references and features onto the sequence(s)
60 public class DBRefFetcher implements Runnable
62 private static final String NEWLINE = System.lineSeparator();
64 public interface FetchFinishedListenerI
71 IProgressIndicator progressWindow;
73 CutAndPasteTransfer output = new CutAndPasteTransfer();
75 boolean running = false;
78 * picr client instance
80 uk.ac.ebi.www.picr.AccessionMappingService.AccessionMapperInterface picrClient = null;
82 // This will be a collection of Vectors of sequenceI refs.
83 // The key will be the seq name or accession id of the seq
84 Hashtable<String, Vector<SequenceI>> seqRefs;
86 DbSourceProxy[] dbSources;
88 SequenceFetcher sfetcher;
90 private List<FetchFinishedListenerI> listeners;
92 private SequenceI[] alseqs;
95 * when true - retrieved sequences will be trimmed to cover longest derived
98 private boolean trimDsSeqs = true;
101 * Creates a new DBRefFetcher object and fetches from the currently selected
102 * set of databases, if this is null then it fetches based on feature settings
105 * fetch references for these SequenceI array
106 * @param progressIndicatorFrame
107 * the frame for progress bar monitoring
109 * array of DbSourceProxy to query references form
110 * @param featureSettings
111 * FeatureSettings to get alternative DbSourceProxy from
112 * @param isNucleotide
113 * indicates if the array of SequenceI are Nucleotides or not
115 public DBRefFetcher(SequenceI[] seqs,
116 IProgressIndicator progressIndicatorFrame,
117 DbSourceProxy[] sources, FeatureSettings featureSettings,
118 boolean isNucleotide)
120 listeners = new ArrayList<FetchFinishedListenerI>();
121 this.progressWindow = progressIndicatorFrame;
122 alseqs = new SequenceI[seqs.length];
123 SequenceI[] ds = new SequenceI[seqs.length];
124 for (int i = 0; i < seqs.length; i++)
127 if (seqs[i].getDatasetSequence() != null)
129 ds[i] = seqs[i].getDatasetSequence();
137 // TODO Jalview 2.5 lots of this code should be in the gui package!
138 sfetcher = jalview.gui.SequenceFetcher
139 .getSequenceFetcherSingleton(progressIndicatorFrame);
140 // set default behaviour for transferring excess sequence data to the
142 trimDsSeqs = Cache.getDefault("TRIM_FETCHED_DATASET_SEQS", true);
145 setDatabaseSources(featureSettings, isNucleotide);
149 // we assume the caller knows what they're doing and ensured that all the
150 // db source names are valid
156 * Helper method to configure the list of database sources to query
158 * @param featureSettings
159 * @param forNucleotide
161 void setDatabaseSources(FeatureSettings featureSettings,
162 boolean forNucleotide)
164 // af.featureSettings_actionPerformed(null);
165 String[] defdb = null;
166 List<DbSourceProxy> selsources = new ArrayList<DbSourceProxy>();
167 Vector<jalviewSourceI> dasselsrc = (featureSettings != null)
168 ? featureSettings.getSelectedSources()
169 : new DasSourceBrowser().getSelectedSources();
171 for (jalviewSourceI src : dasselsrc)
173 List<DbSourceProxy> sp = src.getSequenceSourceProxies();
176 selsources.addAll(sp);
179 Cache.log.debug("Added many Db Sources for :" + src.getTitle());
183 // select appropriate databases based on alignFrame context.
186 defdb = DBRefSource.DNACODINGDBS;
190 defdb = DBRefSource.PROTEINDBS;
192 List<DbSourceProxy> srces = new ArrayList<DbSourceProxy>();
193 for (String ddb : defdb)
195 List<DbSourceProxy> srcesfordb = sfetcher.getSourceProxy(ddb);
196 if (srcesfordb != null)
198 for (DbSourceProxy src : srcesfordb)
200 if (!srces.contains(src))
202 srces.addAll(srcesfordb);
207 // append the PDB data source, since it is 'special', catering for both
208 // nucleotide and protein
209 // srces.addAll(sfetcher.getSourceProxy(DBRefSource.PDB));
211 srces.addAll(selsources);
212 dbSources = srces.toArray(new DbSourceProxy[srces.size()]);
216 * Constructor with only sequences provided
220 public DBRefFetcher(SequenceI[] sequences)
222 this(sequences, null, null, null, false);
226 * Add a listener to be notified when sequence fetching is complete
230 public void addListener(FetchFinishedListenerI l)
236 * retrieve all the das sequence sources and add them to the list of db
237 * sources to retrieve from
239 public void appendAllDasSources()
241 if (dbSources == null)
243 dbSources = new DbSourceProxy[0];
245 // append additional sources
246 DbSourceProxy[] otherdb = sfetcher
247 .getDbSourceProxyInstances(DasSequenceSource.class);
248 if (otherdb != null && otherdb.length > 0)
250 DbSourceProxy[] newsrc = new DbSourceProxy[dbSources.length
252 System.arraycopy(dbSources, 0, newsrc, 0, dbSources.length);
253 System.arraycopy(otherdb, 0, newsrc, dbSources.length,
260 * start the fetcher thread
262 * @param waitTillFinished
263 * true to block until the fetcher has finished
265 public void fetchDBRefs(boolean waitTillFinished)
267 // TODO can we not simply write
268 // if (waitTillFinished) { run(); } else { new Thread(this).start(); }
270 Thread thread = new Thread(this);
274 if (waitTillFinished)
281 } catch (Exception ex)
289 * The sequence will be added to a vector of sequences belonging to key which
290 * could be either seq name or dbref id
297 void addSeqId(SequenceI seq, String key)
299 key = key.toUpperCase();
301 Vector<SequenceI> seqs;
302 if (seqRefs.containsKey(key))
304 seqs = seqRefs.get(key);
306 if (seqs != null && !seqs.contains(seq))
308 seqs.addElement(seq);
310 else if (seqs == null)
312 seqs = new Vector<SequenceI>();
313 seqs.addElement(seq);
319 seqs = new Vector<SequenceI>();
320 seqs.addElement(seq);
323 seqRefs.put(key, seqs);
332 if (dbSources == null)
334 throw new Error(MessageManager
335 .getString("error.implementation_error_must_init_dbsources"));
338 long startTime = System.currentTimeMillis();
339 if (progressWindow != null)
341 progressWindow.setProgressBar(
342 MessageManager.getString("status.fetching_db_refs"),
347 if (Cache.getDefault("DBREFFETCH_USEPICR", false))
349 picrClient = new AccessionMapperServiceLocator()
350 .getAccessionMapperPort();
352 } catch (Exception e)
354 System.err.println("Couldn't locate PICR service instance.\n");
358 Vector<SequenceI> sdataset = new Vector<SequenceI>(
359 Arrays.asList(dataset));
360 List<String> warningMessages = new ArrayList<String>();
363 while (sdataset.size() > 0 && db < dbSources.length)
365 int maxqlen = 1; // default number of queries made at one time
366 System.out.println("Verifying against " + dbSources[db].getDbName());
368 // iterate through db for each remaining un-verified sequence
369 SequenceI[] currSeqs = new SequenceI[sdataset.size()];
370 sdataset.copyInto(currSeqs);// seqs that are to be validated against
372 Vector<String> queries = new Vector<String>(); // generated queries curSeq
373 seqRefs = new Hashtable<String, Vector<SequenceI>>();
377 DbSourceProxy dbsource = dbSources[db];
378 // for moment, we dumbly iterate over all retrieval sources for a
379 // particular database
380 // TODO: introduce multithread multisource queries and logic to remove a
381 // query from other sources if any source for a database returns a
383 maxqlen = dbsource.getMaximumQueryCount();
385 while (queries.size() > 0 || seqIndex < currSeqs.length)
387 if (queries.size() > 0)
389 // Still queries to make for current seqIndex
390 StringBuffer queryString = new StringBuffer("");
392 int nqSize = (maxqlen > queries.size()) ? queries.size()
395 while (queries.size() > 0 && numq < nqSize)
397 String query = queries.elementAt(0);
398 if (dbsource.isValidReference(query))
401 (numq == 0) ? "" : dbsource.getAccessionSeparator());
402 queryString.append(query);
405 // remove the extracted query string
406 queries.removeElementAt(0);
408 // make the queries and process the response
409 AlignmentI retrieved = null;
412 if (Cache.log.isDebugEnabled())
414 Cache.log.debug("Querying " + dbsource.getDbName()
415 + " with : '" + queryString.toString() + "'");
417 retrieved = dbsource.getSequenceRecords(queryString.toString());
418 } catch (Exception ex)
420 ex.printStackTrace();
421 } catch (OutOfMemoryError err)
423 new OOMWarning("retrieving database references ("
424 + queryString.toString() + ")", err);
426 if (retrieved != null)
428 transferReferences(sdataset, dbsource.getDbSource(), retrieved,
429 trimDsSeqs, warningMessages);
434 // make some more strings for use as queries
435 for (int i = 0; (seqIndex < dataset.length)
436 && (i < 50); seqIndex++, i++)
438 SequenceI sequence = dataset[seqIndex];
439 DBRefEntry[] uprefs = DBRefUtils
440 .selectRefs(sequence.getDBRefs(), new String[]
441 { dbsource.getDbSource() }); // jalview.datamodel.DBRefSource.UNIPROT
443 // check for existing dbrefs to use
444 if (uprefs != null && uprefs.length > 0)
446 for (int j = 0; j < uprefs.length; j++)
448 addSeqId(sequence, uprefs[j].getAccessionId());
450 uprefs[j].getAccessionId().toUpperCase());
455 // generate queries from sequence ID string
456 StringTokenizer st = new StringTokenizer(sequence.getName(),
458 while (st.hasMoreTokens())
460 String token = st.nextToken();
461 UPEntry[] presp = null;
462 if (picrClient != null)
464 // resolve the string against PICR to recover valid IDs
467 presp = picrClient.getUPIForAccession(token, null,
468 picrClient.getMappedDatabaseNames(), null,
470 } catch (Exception e)
473 "Exception with Picr for '" + token + "'\n");
477 if (presp != null && presp.length > 0)
479 for (int id = 0; id < presp.length; id++)
481 // construct sequences from response if sequences are
482 // present, and do a transferReferences
483 // otherwise transfer non sequence x-references directly.
486 "Validated ID against PICR... (for what its worth):"
488 addSeqId(sequence, token);
489 queries.addElement(token.toUpperCase());
494 // System.out.println("Not querying source with
495 // token="+token+"\n");
496 addSeqId(sequence, token);
497 queries.addElement(token.toUpperCase());
504 // advance to next database
506 } // all databases have been queried
507 if (!warningMessages.isEmpty())
509 StringBuilder sb = new StringBuilder(warningMessages.size() * 30);
510 sb.append(MessageManager
511 .getString("label.your_sequences_have_been_verified"));
512 for (String msg : warningMessages)
514 sb.append(msg).append(NEWLINE);
516 output.setText(sb.toString());
518 Desktop.addInternalFrame(output,
519 MessageManager.getString("label.sequences_updated"), 600,
521 // The above is the dataset, we must now find out the index
522 // of the viewed sequence
525 if (progressWindow != null)
527 progressWindow.setProgressBar(
528 MessageManager.getString("label.dbref_search_completed"),
532 for (FetchFinishedListenerI listener : listeners)
540 * Verify local sequences in seqRefs against the retrieved sequence database
541 * records. Returns true if any sequence was modified as a result (start/end
542 * changed and/or sequence enlarged), else false.
545 * dataset sequences we are retrieving for
547 * database source we are retrieving from
549 * retrieved sequences as alignment
550 * @param trimDatasetSeqs
551 * if true, sequences will not be enlarged to match longer retrieved
552 * sequences, only their start/end adjusted
553 * @param warningMessages
554 * a list of messages to add to
556 boolean transferReferences(Vector<SequenceI> sdataset, String dbSource,
557 AlignmentI retrievedAl, boolean trimDatasetSeqs,
558 List<String> warningMessages)
560 // System.out.println("trimming ? " + trimDatasetSeqs);
561 if (retrievedAl == null || retrievedAl.getHeight() == 0)
566 boolean modified = false;
567 SequenceI[] retrieved = recoverDbSequences(
568 retrievedAl.getSequencesArray());
569 SequenceI sequence = null;
571 for (SequenceI retrievedSeq : retrieved)
573 // Work out which sequences this sequence matches,
574 // taking into account all accessionIds and names in the file
575 Vector<SequenceI> sequenceMatches = new Vector<SequenceI>();
576 // look for corresponding accession ids
577 DBRefEntry[] entryRefs = DBRefUtils
578 .selectRefs(retrievedSeq.getDBRefs(), new String[]
580 if (entryRefs == null)
583 .println("Dud dbSource string ? no entryrefs selected for "
584 + dbSource + " on " + retrievedSeq.getName());
587 for (int j = 0; j < entryRefs.length; j++)
589 String accessionId = entryRefs[j].getAccessionId();
590 // match up on accessionId
591 if (seqRefs.containsKey(accessionId.toUpperCase()))
593 Vector<SequenceI> seqs = seqRefs.get(accessionId);
594 for (int jj = 0; jj < seqs.size(); jj++)
596 sequence = seqs.elementAt(jj);
597 if (!sequenceMatches.contains(sequence))
599 sequenceMatches.addElement(sequence);
604 if (sequenceMatches.isEmpty())
606 // failed to match directly on accessionId==query so just compare all
607 // sequences to entry
608 Enumeration<String> e = seqRefs.keys();
609 while (e.hasMoreElements())
611 Vector<SequenceI> sqs = seqRefs.get(e.nextElement());
612 if (sqs != null && sqs.size() > 0)
614 Enumeration<SequenceI> sqe = sqs.elements();
615 while (sqe.hasMoreElements())
617 sequenceMatches.addElement(sqe.nextElement());
622 // look for corresponding names
623 // this is uniprot specific ?
624 // could be useful to extend this so we try to find any 'significant'
625 // information in common between two sequence objects.
627 * DBRefEntry[] entryRefs =
628 * jalview.util.DBRefUtils.selectRefs(entry.getDBRef(), new String[] {
629 * dbSource }); for (int j = 0; j < entry.getName().size(); j++) { String
630 * name = entry.getName().elementAt(j).toString(); if
631 * (seqRefs.containsKey(name)) { Vector seqs = (Vector) seqRefs.get(name);
632 * for (int jj = 0; jj < seqs.size(); jj++) { sequence = (SequenceI)
633 * seqs.elementAt(jj); if (!sequenceMatches.contains(sequence)) {
634 * sequenceMatches.addElement(sequence); } } } }
636 // sequenceMatches now contains the set of all sequences associated with
637 // the returned db record
638 final String retrievedSeqString = retrievedSeq.getSequenceAsString();
639 String entrySeq = retrievedSeqString.toUpperCase();
640 for (int m = 0; m < sequenceMatches.size(); m++)
642 sequence = sequenceMatches.elementAt(m);
643 // only update start and end positions and shift features if there are
644 // no existing references
645 // TODO: test for legacy where uniprot or EMBL refs exist but no
646 // mappings are made (but content matches retrieved set)
647 boolean updateRefFrame = sequence.getDBRefs() == null
648 || sequence.getDBRefs().length == 0;
650 // verify sequence against the entry sequence
653 final int sequenceStart = sequence.getStart();
655 boolean remoteEnclosesLocal = false;
656 String nonGapped = AlignSeq
657 .extractGaps("-. ", sequence.getSequenceAsString())
659 int absStart = entrySeq.indexOf(nonGapped);
662 // couldn't find local sequence in sequence from database, so check if
663 // the database sequence is a subsequence of local sequence
664 absStart = nonGapped.indexOf(entrySeq);
667 // verification failed. couldn't find any relationship between
668 // entrySeq and local sequence
669 // messages suppressed as many-to-many matches are confusing
670 // String msg = sequence.getName()
671 // + " Sequence not 100% match with "
672 // + retrievedSeq.getName();
673 // addWarningMessage(warningMessages, msg);
677 * retrieved sequence is a proper subsequence of local sequence
679 String msg = sequence.getName() + " has " + absStart
680 + " prefixed residues compared to "
681 + retrievedSeq.getName();
682 addWarningMessage(warningMessages, msg);
685 * So create a mapping to the external entry from the matching region of
686 * the local sequence, and leave local start/end untouched.
688 mp = new Mapping(null,
690 { sequenceStart + absStart,
691 sequenceStart + absStart + entrySeq.length() - 1 },
693 { retrievedSeq.getStart(),
694 retrievedSeq.getStart() + entrySeq.length() - 1 },
696 updateRefFrame = false;
701 * local sequence is a subsequence of (or matches) retrieved sequence
703 remoteEnclosesLocal = true;
709 * relocate existing sequence features by offset
711 int startShift = absStart - sequenceStart + 1;
714 modified |= sequence.getFeatures().shiftFeatures(startShift);
719 System.out.println("Adding dbrefs to " + sequence.getName()
720 + " from " + dbSource + " sequence : "
721 + retrievedSeq.getName());
722 sequence.transferAnnotation(retrievedSeq, mp);
724 absStart += retrievedSeq.getStart();
725 int absEnd = absStart + nonGapped.length() - 1;
726 if (!trimDatasetSeqs)
729 * update start position and/or expand to longer retrieved sequence
731 if (!retrievedSeqString.equals(sequence.getSequenceAsString())
732 && remoteEnclosesLocal)
734 sequence.setSequence(retrievedSeqString);
736 addWarningMessage(warningMessages,
737 "Sequence for " + sequence.getName() + " expanded from "
738 + retrievedSeq.getName());
740 if (sequence.getStart() != retrievedSeq.getStart())
742 sequence.setStart(retrievedSeq.getStart());
744 if (absStart != sequenceStart)
746 addWarningMessage(warningMessages,
747 "Start/end position for " + sequence.getName()
748 + " updated from " + retrievedSeq.getName());
754 // finally, update local sequence reference frame if we're allowed
757 // just fix start/end
758 if (sequence.getStart() != absStart
759 || sequence.getEnd() != absEnd)
761 sequence.setStart(absStart);
762 sequence.setEnd(absEnd);
764 addWarningMessage(warningMessages,
765 "Start/end for " + sequence.getName()
766 + " updated from " + retrievedSeq.getName());
769 // search for alignment sequences to update coordinate frame for
770 for (int alsq = 0; alsq < alseqs.length; alsq++)
772 if (alseqs[alsq].getDatasetSequence() == sequence)
774 String ngAlsq = AlignSeq
776 alseqs[alsq].getSequenceAsString())
778 int oldstrt = alseqs[alsq].getStart();
779 alseqs[alsq].setStart(sequence.getSequenceAsString()
780 .toUpperCase().indexOf(ngAlsq) + sequence.getStart());
781 if (oldstrt != alseqs[alsq].getStart())
784 ngAlsq.length() + alseqs[alsq].getStart() - 1);
789 // TODO: search for all other references to this dataset sequence, and
791 // TODO: update all AlCodonMappings which involve this alignment
792 // sequence (e.g. Q30167 cdna translation from exon2 product (vamsas
795 // and remove it from the rest
796 // TODO: decide if we should remove annotated sequence from set
797 sdataset.remove(sequence);
798 // TODO: should we make a note of sequences that have received new DB
799 // ids, so we can query all enabled DAS servers for them ?
806 * Adds the message to the list unless it already contains it
811 void addWarningMessage(List<String> messageList, String msg)
813 if (!messageList.contains(msg))
815 messageList.add(msg);
820 * loop thru and collect additional sequences in Map.
822 * @param sequencesArray
825 private SequenceI[] recoverDbSequences(SequenceI[] sequencesArray)
827 Vector<SequenceI> nseq = new Vector<SequenceI>();
828 for (int i = 0; sequencesArray != null
829 && i < sequencesArray.length; i++)
831 nseq.addElement(sequencesArray[i]);
832 DBRefEntry[] dbr = sequencesArray[i].getDBRefs();
834 for (int r = 0; (dbr != null) && r < dbr.length; r++)
836 if ((map = dbr[r].getMap()) != null)
838 if (map.getTo() != null && !nseq.contains(map.getTo()))
840 nseq.addElement(map.getTo());
847 sequencesArray = new SequenceI[nseq.size()];
848 nseq.toArray(sequencesArray);
850 return sequencesArray;