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.SequenceFeature;
30 import jalview.datamodel.SequenceI;
31 import jalview.gui.CutAndPasteTransfer;
32 import jalview.gui.Desktop;
33 import jalview.gui.FeatureSettings;
34 import jalview.gui.IProgressIndicator;
35 import jalview.gui.OOMWarning;
36 import jalview.util.MessageManager;
37 import jalview.ws.dbsources.das.api.jalviewSourceI;
38 import jalview.ws.seqfetcher.DbSourceProxy;
40 import java.util.ArrayList;
41 import java.util.Enumeration;
42 import java.util.Hashtable;
43 import java.util.List;
44 import java.util.StringTokenizer;
45 import java.util.Vector;
47 import uk.ac.ebi.picr.model.UPEntry;
50 * Implements a runnable for validating a sequence against external databases
51 * and then propagating references and features onto the sequence(s)
56 public class DBRefFetcher implements Runnable
58 public interface FetchFinishedListenerI
63 private List<FetchFinishedListenerI> listeners;
67 IProgressIndicator progressWindow;
69 CutAndPasteTransfer output = new CutAndPasteTransfer();
71 StringBuffer sbuffer = new StringBuffer();
73 boolean running = false;
76 * picr client instance
78 uk.ac.ebi.www.picr.AccessionMappingService.AccessionMapperInterface picrClient = null;
80 // /This will be a collection of Vectors of sequenceI refs.
81 // The key will be the seq name or accession id of the seq
84 DbSourceProxy[] dbSources;
86 SequenceFetcher sfetcher;
88 private SequenceI[] alseqs;
91 * when true - retrieved sequences will be trimmed to cover longest derived
94 private boolean trimDsSeqs = true;
97 * Creates a new DBRefFetcher object and fetches from the currently selected
98 * set of databases, if this is null then it fetches based on feature settings
101 * fetch references for these SequenceI array
102 * @param progressIndicatorFrame
103 * the frame for progress bar monitoring
105 * array of DbSourceProxy to query references form
106 * @param featureSettings
107 * FeatureSettings to get alternative DbSourceProxy from
108 * @param isNucleotide
109 * indicates if the array of SequenceI are Nucleotides or not
111 public DBRefFetcher(SequenceI[] seqs,
112 IProgressIndicator progressIndicatorFrame,
113 DbSourceProxy[] sources, FeatureSettings featureSettings, boolean isNucleotide)
115 listeners = new ArrayList<FetchFinishedListenerI>();
116 this.progressWindow = progressIndicatorFrame;
117 alseqs = new SequenceI[seqs.length];
118 SequenceI[] ds = new SequenceI[seqs.length];
119 for (int i = 0; i < seqs.length; i++)
122 if (seqs[i].getDatasetSequence() != null)
124 ds[i] = seqs[i].getDatasetSequence();
132 // TODO Jalview 2.5 lots of this code should be in the gui package!
133 sfetcher = jalview.gui.SequenceFetcher
134 .getSequenceFetcherSingleton(progressIndicatorFrame);
135 // set default behaviour for transferring excess sequence data to the
137 trimDsSeqs = Cache.getDefault("TRIM_FETCHED_DATASET_SEQS", true);
140 // af.featureSettings_actionPerformed(null);
141 String[] defdb = null, otherdb = sfetcher
142 .getDbInstances(jalview.ws.dbsources.das.datamodel.DasSequenceSource.class);
143 List<DbSourceProxy> selsources = new ArrayList<DbSourceProxy>();
144 Vector<jalviewSourceI> dasselsrc = (featureSettings != null) ? featureSettings
145 .getSelectedSources() : new jalview.gui.DasSourceBrowser()
146 .getSelectedSources();
147 Enumeration<jalviewSourceI> en = dasselsrc.elements();
148 while (en.hasMoreElements())
150 jalviewSourceI src = en.nextElement();
151 List<DbSourceProxy> sp = src.getSequenceSourceProxies();
154 selsources.addAll(sp);
157 Cache.log.debug("Added many Db Sources for :" + src.getTitle());
161 // select appropriate databases based on alignFrame context.
164 defdb = DBRefSource.DNACODINGDBS;
168 defdb = DBRefSource.PROTEINDBS;
170 List<DbSourceProxy> srces = new ArrayList<DbSourceProxy>();
171 for (String ddb : defdb)
173 List<DbSourceProxy> srcesfordb = sfetcher.getSourceProxy(ddb);
174 if (srcesfordb != null)
176 srces.addAll(srcesfordb);
180 // append the selected sequence sources to the default dbs
181 srces.addAll(selsources);
182 dbSources = srces.toArray(new DbSourceProxy[0]);
186 // we assume the caller knows what they're doing and ensured that all the
187 // db source names are valid
193 * Constructor with only sequences provided
197 public DBRefFetcher(SequenceI[] sequences)
199 this(sequences, null, null, null, false);
203 * Add a listener to be notified when sequence fetching is complete
207 public void addListener(FetchFinishedListenerI l)
213 * retrieve all the das sequence sources and add them to the list of db
214 * sources to retrieve from
216 public void appendAllDasSources()
218 if (dbSources == null)
220 dbSources = new DbSourceProxy[0];
222 // append additional sources
223 DbSourceProxy[] otherdb = sfetcher
224 .getDbSourceProxyInstances(jalview.ws.dbsources.das.datamodel.DasSequenceSource.class);
225 if (otherdb != null && otherdb.length > 0)
227 DbSourceProxy[] newsrc = new DbSourceProxy[dbSources.length
229 System.arraycopy(dbSources, 0, newsrc, 0, dbSources.length);
230 System.arraycopy(otherdb, 0, newsrc, dbSources.length, otherdb.length);
236 * start the fetcher thread
238 * @param waitTillFinished
239 * true to block until the fetcher has finished
241 public void fetchDBRefs(boolean waitTillFinished)
243 Thread thread = new Thread(this);
247 if (waitTillFinished)
254 } catch (Exception ex)
262 * The sequence will be added to a vector of sequences belonging to key which
263 * could be either seq name or dbref id
270 void addSeqId(SequenceI seq, String key)
272 key = key.toUpperCase();
275 if (seqRefs.containsKey(key))
277 seqs = (Vector) seqRefs.get(key);
279 if (seqs != null && !seqs.contains(seq))
281 seqs.addElement(seq);
283 else if (seqs == null)
286 seqs.addElement(seq);
293 seqs.addElement(seq);
296 seqRefs.put(key, seqs);
305 if (dbSources == null)
309 .getString("error.implementation_error_must_init_dbsources"));
312 long startTime = System.currentTimeMillis();
313 if (progressWindow != null)
315 progressWindow.setProgressBar(
316 MessageManager.getString("status.fetching_db_refs"),
321 if (Cache.getDefault("DBREFFETCH_USEPICR", false))
323 picrClient = new uk.ac.ebi.www.picr.AccessionMappingService.AccessionMapperServiceLocator()
324 .getAccessionMapperPort();
326 } catch (Exception e)
328 System.err.println("Couldn't locate PICR service instance.\n");
332 Vector sdataset = new Vector();
333 for (int s = 0; s < dataset.length; s++)
335 sdataset.addElement(dataset[s]);
337 while (sdataset.size() > 0 && db < dbSources.length)
339 int maxqlen = 1; // default number of queries made to at one time
340 System.err.println("Verifying against " + dbSources[db].getDbName());
343 // iterate through db for each remaining un-verified sequence
344 SequenceI[] currSeqs = new SequenceI[sdataset.size()];
345 sdataset.copyInto(currSeqs);// seqs that are to be validated against
347 Vector queries = new Vector(); // generated queries curSeq
348 seqRefs = new Hashtable();
352 jalview.ws.seqfetcher.DbSourceProxy dbsource = dbSources[db];
354 // for moment, we dumbly iterate over all retrieval sources for a
355 // particular database
356 // TODO: introduce multithread multisource queries and logic to remove a
357 // query from other sources if any source for a database returns a
359 maxqlen = dbsource.getMaximumQueryCount();
361 while (queries.size() > 0 || seqIndex < currSeqs.length)
363 if (queries.size() > 0)
365 // Still queries to make for current seqIndex
366 StringBuffer queryString = new StringBuffer("");
367 int numq = 0, nqSize = (maxqlen > queries.size()) ? queries
370 while (queries.size() > 0 && numq < nqSize)
372 String query = (String) queries.elementAt(0);
373 if (dbsource.isValidReference(query))
375 queryString.append((numq == 0) ? "" : dbsource
376 .getAccessionSeparator());
377 queryString.append(query);
380 // remove the extracted query string
381 queries.removeElementAt(0);
383 // make the queries and process the response
384 AlignmentI retrieved = null;
387 if (jalview.bin.Cache.log.isDebugEnabled())
389 jalview.bin.Cache.log.debug("Querying "
390 + dbsource.getDbName() + " with : '"
391 + queryString.toString() + "'");
393 retrieved = dbsource.getSequenceRecords(queryString
395 } catch (Exception ex)
397 ex.printStackTrace();
398 } catch (OutOfMemoryError err)
400 new OOMWarning("retrieving database references ("
401 + queryString.toString() + ")", err);
403 if (retrieved != null)
405 transferReferences(sdataset, dbsource.getDbSource(),
406 retrieved, trimDsSeqs);
411 // make some more strings for use as queries
412 for (int i = 0; (seqIndex < dataset.length) && (i < 50); seqIndex++, i++)
414 SequenceI sequence = dataset[seqIndex];
415 DBRefEntry[] uprefs = jalview.util.DBRefUtils.selectRefs(
416 sequence.getDBRefs(),
417 new String[] { dbsource.getDbSource() }); // jalview.datamodel.DBRefSource.UNIPROT
419 // check for existing dbrefs to use
420 if (uprefs != null && uprefs.length > 0)
422 for (int j = 0; j < uprefs.length; j++)
424 addSeqId(sequence, uprefs[j].getAccessionId());
425 queries.addElement(uprefs[j].getAccessionId()
431 // generate queries from sequence ID string
432 StringTokenizer st = new StringTokenizer(
433 sequence.getName(), "|");
434 while (st.hasMoreTokens())
436 String token = st.nextToken();
437 UPEntry[] presp = null;
438 if (picrClient != null)
440 // resolve the string against PICR to recover valid IDs
443 presp = picrClient.getUPIForAccession(token, null,
444 picrClient.getMappedDatabaseNames(), null,
446 } catch (Exception e)
448 System.err.println("Exception with Picr for '"
453 if (presp != null && presp.length > 0)
455 for (int id = 0; id < presp.length; id++)
457 // construct sequences from response if sequences are
458 // present, and do a transferReferences
459 // otherwise transfer non sequence x-references directly.
462 .println("Validated ID against PICR... (for what its worth):"
464 addSeqId(sequence, token);
465 queries.addElement(token.toUpperCase());
470 // System.out.println("Not querying source with token="+token+"\n");
471 addSeqId(sequence, token);
472 queries.addElement(token.toUpperCase());
480 // advance to next database
482 } // all databases have been queries.
483 if (sbuffer.length() > 0)
485 output.setText(MessageManager
486 .getString("label.your_sequences_have_been_verified")
487 + sbuffer.toString());
488 Desktop.addInternalFrame(output,
489 MessageManager.getString("label.sequence_names_updated"),
491 // The above is the dataset, we must now find out the index
492 // of the viewed sequence
495 if (progressWindow != null)
497 progressWindow.setProgressBar(
498 MessageManager.getString("label.dbref_search_completed"),
502 for (FetchFinishedListenerI listener : listeners)
510 * Verify local sequences in seqRefs against the retrieved sequence database
513 * @param trimDatasetSeqs
516 void transferReferences(Vector sdataset, String dbSource,
517 AlignmentI retrievedAl, boolean trimDatasetSeqs) // File
520 System.out.println("trimming ? " + trimDatasetSeqs);
521 if (retrievedAl == null || retrievedAl.getHeight() == 0)
525 SequenceI[] retrieved = recoverDbSequences(retrievedAl
526 .getSequencesArray());
527 SequenceI sequence = null;
528 boolean transferred = false;
529 StringBuffer messages = new StringBuffer();
531 // Vector entries = new Uniprot().getUniprotEntries(file);
533 int i, iSize = retrieved.length; // entries == null ? 0 : entries.size();
534 // UniprotEntry entry;
535 for (i = 0; i < iSize; i++)
537 SequenceI entry = retrieved[i]; // (UniprotEntry) entries.elementAt(i);
539 // Work out which sequences this sequence matches,
540 // taking into account all accessionIds and names in the file
541 Vector sequenceMatches = new Vector();
542 // look for corresponding accession ids
543 DBRefEntry[] entryRefs = jalview.util.DBRefUtils.selectRefs(
544 entry.getDBRefs(), new String[] { dbSource });
545 if (entryRefs == null)
548 .println("Dud dbSource string ? no entryrefs selected for "
549 + dbSource + " on " + entry.getName());
552 for (int j = 0; j < entryRefs.length; j++)
554 String accessionId = entryRefs[j].getAccessionId(); // .getAccession().elementAt(j).toString();
555 // match up on accessionId
556 if (seqRefs.containsKey(accessionId.toUpperCase()))
558 Vector seqs = (Vector) seqRefs.get(accessionId);
559 for (int jj = 0; jj < seqs.size(); jj++)
561 sequence = (SequenceI) seqs.elementAt(jj);
562 if (!sequenceMatches.contains(sequence))
564 sequenceMatches.addElement(sequence);
569 if (sequenceMatches.size() == 0)
571 // failed to match directly on accessionId==query so just compare all
572 // sequences to entry
573 Enumeration e = seqRefs.keys();
574 while (e.hasMoreElements())
576 Vector sqs = (Vector) seqRefs.get(e.nextElement());
577 if (sqs != null && sqs.size() > 0)
579 Enumeration sqe = sqs.elements();
580 while (sqe.hasMoreElements())
582 sequenceMatches.addElement(sqe.nextElement());
587 // look for corresponding names
588 // this is uniprot specific ?
589 // could be useful to extend this so we try to find any 'significant'
590 // information in common between two sequence objects.
592 * DBRefEntry[] entryRefs =
593 * jalview.util.DBRefUtils.selectRefs(entry.getDBRef(), new String[] {
594 * dbSource }); for (int j = 0; j < entry.getName().size(); j++) { String
595 * name = entry.getName().elementAt(j).toString(); if
596 * (seqRefs.containsKey(name)) { Vector seqs = (Vector) seqRefs.get(name);
597 * for (int jj = 0; jj < seqs.size(); jj++) { sequence = (SequenceI)
598 * seqs.elementAt(jj); if (!sequenceMatches.contains(sequence)) {
599 * sequenceMatches.addElement(sequence); } } } }
601 // sequenceMatches now contains the set of all sequences associated with
602 // the returned db record
603 String entrySeq = entry.getSequenceAsString().toUpperCase();
604 for (int m = 0; m < sequenceMatches.size(); m++)
606 sequence = (SequenceI) sequenceMatches.elementAt(m);
607 // only update start and end positions and shift features if there are
608 // no existing references
609 // TODO: test for legacy where uniprot or EMBL refs exist but no
610 // mappings are made (but content matches retrieved set)
611 boolean updateRefFrame = sequence.getDBRefs() == null
612 || sequence.getDBRefs().length == 0;
614 // verify sequence against the entry sequence
616 String nonGapped = AlignSeq.extractGaps("-. ",
617 sequence.getSequenceAsString()).toUpperCase();
619 int absStart = entrySeq.indexOf(nonGapped);
622 final int sequenceStart = sequence.getStart();
625 // Is local sequence contained in dataset sequence?
626 absStart = nonGapped.indexOf(entrySeq);
628 { // verification failed.
629 messages.append(sequence.getName()
630 + " SEQUENCE NOT %100 MATCH \n");
634 sbuffer.append(sequence.getName() + " HAS " + absStart
635 + " PREFIXED RESIDUES COMPARED TO " + dbSource + "\n");
637 // + " - ANY SEQUENCE FEATURES"
638 // + " HAVE BEEN ADJUSTED ACCORDINGLY \n");
640 // create valid mapping between matching region of local sequence and
641 // the mapped sequence
642 mp = new Mapping(null, new int[] { sequenceStart + absStart,
643 sequenceStart + absStart + entrySeq.length() - 1 }, new int[]
644 { entry.getStart(), entry.getStart() + entrySeq.length() - 1 },
646 updateRefFrame = false; // mapping is based on current start/end so
647 // don't modify start and end
652 // update start and end of local sequence to place it in entry's
654 // apply identity map map from whole of local sequence to matching
655 // region of database
657 mp = null; // Mapping.getIdentityMap();
659 // new int[] { absStart+sequence.getStart(),
660 // absStart+sequence.getStart()+entrySeq.length()-1},
661 // new int[] { entry.getStart(), entry.getEnd() }, 1, 1);
662 // relocate local features for updated start
665 if (sequence.getSequenceFeatures() != null)
667 SequenceFeature[] sf = sequence.getSequenceFeatures();
668 int start = sequenceStart;
669 int end = sequence.getEnd();
670 int startShift = 1 - absStart - start; // how much the features
673 for (int sfi = 0; sfi < sf.length; sfi++)
675 if (sf[sfi].getBegin() >= start && sf[sfi].getEnd() <= end)
677 // shift feature along by absstart
678 sf[sfi].setBegin(sf[sfi].getBegin() + startShift);
679 sf[sfi].setEnd(sf[sfi].getEnd() + startShift);
686 System.out.println("Adding dbrefs to " + sequence.getName()
687 + " from " + dbSource + " sequence : " + entry.getName());
688 sequence.transferAnnotation(entry, mp);
689 // unknownSequences.remove(sequence);
690 absStart += entry.getStart();
691 int absEnd = absStart + nonGapped.length() - 1;
692 if (!trimDatasetSeqs)
694 // insert full length sequence from record
695 sequence.setSequence(entry.getSequenceAsString());
696 sequence.setStart(entry.getStart());
700 // finally, update local sequence reference frame if we're allowed
703 // just fix start/end
704 sequence.setStart(absStart);
705 sequence.setEnd(absEnd);
707 // search for alignment sequences to update coordinate frame for
708 for (int alsq = 0; alsq < alseqs.length; alsq++)
710 if (alseqs[alsq].getDatasetSequence() == sequence)
712 String ngAlsq = AlignSeq.extractGaps("-. ",
713 alseqs[alsq].getSequenceAsString()).toUpperCase();
714 int oldstrt = alseqs[alsq].getStart();
715 alseqs[alsq].setStart(sequence.getSequenceAsString()
716 .toUpperCase().indexOf(ngAlsq)
717 + sequence.getStart());
718 if (oldstrt != alseqs[alsq].getStart())
720 alseqs[alsq].setEnd(ngAlsq.length()
721 + alseqs[alsq].getStart() - 1);
725 // TODO: search for all other references to this dataset sequence, and
727 // TODO: update all AlCodonMappings which involve this alignment
728 // sequence (e.g. Q30167 cdna translation from exon2 product (vamsas
731 // and remove it from the rest
732 // TODO: decide if we should remove annotated sequence from set
733 sdataset.remove(sequence);
734 // TODO: should we make a note of sequences that have received new DB
735 // ids, so we can query all enabled DAS servers for them ?
740 // report the ID/sequence mismatches
741 sbuffer.append(messages);
746 * loop thru and collect additional sequences in Map.
748 * @param sequencesArray
751 private SequenceI[] recoverDbSequences(SequenceI[] sequencesArray)
753 Vector nseq = new Vector();
754 for (int i = 0; sequencesArray != null && i < sequencesArray.length; i++)
756 nseq.addElement(sequencesArray[i]);
757 DBRefEntry dbr[] = sequencesArray[i].getDBRefs();
758 jalview.datamodel.Mapping map = null;
759 for (int r = 0; (dbr != null) && r < dbr.length; r++)
761 if ((map = dbr[r].getMap()) != null)
763 if (map.getTo() != null && !nseq.contains(map.getTo()))
765 nseq.addElement(map.getTo());
772 sequencesArray = new SequenceI[nseq.size()];
773 nseq.toArray(sequencesArray);
775 return sequencesArray;