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.AlignFrame;
32 import jalview.gui.CutAndPasteTransfer;
33 import jalview.gui.Desktop;
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
60 IProgressIndicator af;
62 CutAndPasteTransfer output = new CutAndPasteTransfer();
64 StringBuffer sbuffer = new StringBuffer();
66 boolean running = false;
69 * picr client instance
71 uk.ac.ebi.www.picr.AccessionMappingService.AccessionMapperInterface picrClient = null;
73 // /This will be a collection of Vectors of sequenceI refs.
74 // The key will be the seq name or accession id of the seq
77 DbSourceProxy[] dbSources;
79 SequenceFetcher sfetcher;
81 private SequenceI[] alseqs;
84 * when true - retrieved sequences will be trimmed to cover longest derived
87 private boolean trimDsSeqs = true;
94 * Creates a new SequenceFeatureFetcher object and fetches from the currently
95 * selected set of databases.
98 * fetch references for these sequences
100 * the parent alignframe for progress bar monitoring.
102 public DBRefFetcher(SequenceI[] seqs, AlignFrame af)
104 this(seqs, af, null);
108 * Creates a new SequenceFeatureFetcher object and fetches from the currently
109 * selected set of databases.
112 * fetch references for these sequences
114 * the parent alignframe for progress bar monitoring.
116 * array of database source strings to query references from
118 public DBRefFetcher(SequenceI[] seqs, AlignFrame af,
119 DbSourceProxy[] sources)
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)
128 ds[i] = seqs[i].getDatasetSequence();
133 // TODO Jalview 2.5 lots of this code should be in the gui package!
134 sfetcher = jalview.gui.SequenceFetcher.getSequenceFetcherSingleton(af);
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 dasselsrc = (af.featureSettings != null) ? af.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.
162 if (af.getViewport().getAlignment().isNucleotide())
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 * retrieve all the das sequence sources and add them to the list of db
194 * sources to retrieve from
196 public void appendAllDasSources()
198 if (dbSources == null)
200 dbSources = new DbSourceProxy[0];
202 // append additional sources
203 DbSourceProxy[] otherdb = sfetcher
204 .getDbSourceProxyInstances(jalview.ws.dbsources.das.datamodel.DasSequenceSource.class);
205 if (otherdb != null && otherdb.length > 0)
207 DbSourceProxy[] newsrc = new DbSourceProxy[dbSources.length
209 System.arraycopy(dbSources, 0, newsrc, 0, dbSources.length);
210 System.arraycopy(otherdb, 0, newsrc, dbSources.length, otherdb.length);
216 * start the fetcher thread
218 * @param waitTillFinished
219 * true to block until the fetcher has finished
221 public void fetchDBRefs(boolean waitTillFinished)
223 Thread thread = new Thread(this);
227 if (waitTillFinished)
234 } catch (Exception ex)
242 * The sequence will be added to a vector of sequences belonging to key which
243 * could be either seq name or dbref id
250 void addSeqId(SequenceI seq, String key)
252 key = key.toUpperCase();
255 if (seqRefs.containsKey(key))
257 seqs = (Vector) seqRefs.get(key);
259 if (seqs != null && !seqs.contains(seq))
261 seqs.addElement(seq);
263 else if (seqs == null)
266 seqs.addElement(seq);
273 seqs.addElement(seq);
276 seqRefs.put(key, seqs);
284 if (dbSources == null)
288 .getString("error.implementation_error_must_init_dbsources"));
291 long startTime = System.currentTimeMillis();
292 af.setProgressBar(MessageManager.getString("status.fetching_db_refs"),
296 if (Cache.getDefault("DBREFFETCH_USEPICR", false))
298 picrClient = new uk.ac.ebi.www.picr.AccessionMappingService.AccessionMapperServiceLocator()
299 .getAccessionMapperPort();
301 } catch (Exception e)
303 System.err.println("Couldn't locate PICR service instance.\n");
307 Vector sdataset = new Vector();
308 for (int s = 0; s < dataset.length; s++)
310 sdataset.addElement(dataset[s]);
312 while (sdataset.size() > 0 && db < dbSources.length)
314 int maxqlen = 1; // default number of queries made to at one time
315 System.err.println("Verifying against " + dbSources[db].getDbName());
318 // iterate through db for each remaining un-verified sequence
319 SequenceI[] currSeqs = new SequenceI[sdataset.size()];
320 sdataset.copyInto(currSeqs);// seqs that are to be validated against
322 Vector queries = new Vector(); // generated queries curSeq
323 seqRefs = new Hashtable();
327 jalview.ws.seqfetcher.DbSourceProxy dbsource = dbSources[db];
329 // for moment, we dumbly iterate over all retrieval sources for a
330 // particular database
331 // TODO: introduce multithread multisource queries and logic to remove a
332 // query from other sources if any source for a database returns a
334 if (dbsource.getDbSourceProperties().containsKey(
335 DBRefSource.MULTIACC))
337 maxqlen = ((Integer) dbsource.getDbSourceProperties().get(
338 DBRefSource.MULTIACC)).intValue();
344 while (queries.size() > 0 || seqIndex < currSeqs.length)
346 if (queries.size() > 0)
348 // Still queries to make for current seqIndex
349 StringBuffer queryString = new StringBuffer("");
350 int numq = 0, nqSize = (maxqlen > queries.size()) ? queries
353 while (queries.size() > 0 && numq < nqSize)
355 String query = (String) queries.elementAt(0);
356 if (dbsource.isValidReference(query))
358 queryString.append((numq == 0) ? "" : dbsource
359 .getAccessionSeparator());
360 queryString.append(query);
363 // remove the extracted query string
364 queries.removeElementAt(0);
366 // make the queries and process the response
367 AlignmentI retrieved = null;
370 if (jalview.bin.Cache.log.isDebugEnabled())
372 jalview.bin.Cache.log.debug("Querying "
373 + dbsource.getDbName() + " with : '"
374 + queryString.toString() + "'");
376 retrieved = dbsource.getSequenceRecords(queryString
378 } catch (Exception ex)
380 ex.printStackTrace();
381 } catch (OutOfMemoryError err)
383 new OOMWarning("retrieving database references ("
384 + queryString.toString() + ")", err);
386 if (retrieved != null)
388 transferReferences(sdataset, dbsource.getDbSource(),
389 retrieved, trimDsSeqs);
394 // make some more strings for use as queries
395 for (int i = 0; (seqIndex < dataset.length) && (i < 50); seqIndex++, i++)
397 SequenceI sequence = dataset[seqIndex];
398 DBRefEntry[] uprefs = jalview.util.DBRefUtils.selectRefs(
400 new String[] { dbsource.getDbSource() }); // jalview.datamodel.DBRefSource.UNIPROT
402 // check for existing dbrefs to use
403 if (uprefs != null && uprefs.length > 0)
405 for (int j = 0; j < uprefs.length; j++)
407 addSeqId(sequence, uprefs[j].getAccessionId());
408 queries.addElement(uprefs[j].getAccessionId()
414 // generate queries from sequence ID string
415 StringTokenizer st = new StringTokenizer(
416 sequence.getName(), "|");
417 while (st.hasMoreTokens())
419 String token = st.nextToken();
420 UPEntry[] presp = null;
421 if (picrClient != null)
423 // resolve the string against PICR to recover valid IDs
426 presp = picrClient.getUPIForAccession(token, null,
427 picrClient.getMappedDatabaseNames(), null,
429 } catch (Exception e)
431 System.err.println("Exception with Picr for '"
436 if (presp != null && presp.length > 0)
438 for (int id = 0; id < presp.length; id++)
440 // construct sequences from response if sequences are
441 // present, and do a transferReferences
442 // otherwise transfer non sequence x-references directly.
445 .println("Validated ID against PICR... (for what its worth):"
447 addSeqId(sequence, token);
448 queries.addElement(token.toUpperCase());
453 // System.out.println("Not querying source with token="+token+"\n");
454 addSeqId(sequence, token);
455 queries.addElement(token.toUpperCase());
463 // advance to next database
465 } // all databases have been queries.
466 if (sbuffer.length() > 0)
468 output.setText(MessageManager
469 .getString("label.your_sequences_have_been_verified")
470 + sbuffer.toString());
471 Desktop.addInternalFrame(output,
472 MessageManager.getString("label.sequence_names_updated"),
474 // The above is the dataset, we must now find out the index
475 // of the viewed sequence
480 MessageManager.getString("label.dbref_search_completed"),
482 // promptBeforeBlast();
489 * Verify local sequences in seqRefs against the retrieved sequence database
492 * @param trimDatasetSeqs
495 void transferReferences(Vector sdataset, String dbSource,
496 AlignmentI retrievedAl, boolean trimDatasetSeqs) // File
499 System.out.println("trimming ? " + trimDatasetSeqs);
500 if (retrievedAl == null || retrievedAl.getHeight() == 0)
504 SequenceI[] retrieved = recoverDbSequences(retrievedAl
505 .getSequencesArray());
506 SequenceI sequence = null;
507 boolean transferred = false;
508 StringBuffer messages = new StringBuffer();
510 // Vector entries = new Uniprot().getUniprotEntries(file);
512 int i, iSize = retrieved.length; // entries == null ? 0 : entries.size();
513 // UniprotEntry entry;
514 for (i = 0; i < iSize; i++)
516 SequenceI entry = retrieved[i]; // (UniprotEntry) entries.elementAt(i);
518 // Work out which sequences this sequence matches,
519 // taking into account all accessionIds and names in the file
520 Vector sequenceMatches = new Vector();
521 // look for corresponding accession ids
522 DBRefEntry[] entryRefs = jalview.util.DBRefUtils.selectRefs(
523 entry.getDBRef(), new String[] { dbSource });
524 if (entryRefs == null)
527 .println("Dud dbSource string ? no entryrefs selected for "
528 + dbSource + " on " + entry.getName());
531 for (int j = 0; j < entryRefs.length; j++)
533 String accessionId = entryRefs[j].getAccessionId(); // .getAccession().elementAt(j).toString();
534 // match up on accessionId
535 if (seqRefs.containsKey(accessionId.toUpperCase()))
537 Vector seqs = (Vector) seqRefs.get(accessionId);
538 for (int jj = 0; jj < seqs.size(); jj++)
540 sequence = (SequenceI) seqs.elementAt(jj);
541 if (!sequenceMatches.contains(sequence))
543 sequenceMatches.addElement(sequence);
548 if (sequenceMatches.size() == 0)
550 // failed to match directly on accessionId==query so just compare all
551 // sequences to entry
552 Enumeration e = seqRefs.keys();
553 while (e.hasMoreElements())
555 Vector sqs = (Vector) seqRefs.get(e.nextElement());
556 if (sqs != null && sqs.size() > 0)
558 Enumeration sqe = sqs.elements();
559 while (sqe.hasMoreElements())
561 sequenceMatches.addElement(sqe.nextElement());
566 // look for corresponding names
567 // this is uniprot specific ?
568 // could be useful to extend this so we try to find any 'significant'
569 // information in common between two sequence objects.
571 * DBRefEntry[] entryRefs =
572 * jalview.util.DBRefUtils.selectRefs(entry.getDBRef(), new String[] {
573 * dbSource }); for (int j = 0; j < entry.getName().size(); j++) { String
574 * name = entry.getName().elementAt(j).toString(); if
575 * (seqRefs.containsKey(name)) { Vector seqs = (Vector) seqRefs.get(name);
576 * for (int jj = 0; jj < seqs.size(); jj++) { sequence = (SequenceI)
577 * seqs.elementAt(jj); if (!sequenceMatches.contains(sequence)) {
578 * sequenceMatches.addElement(sequence); } } } }
580 // sequenceMatches now contains the set of all sequences associated with
581 // the returned db record
582 String entrySeq = entry.getSequenceAsString().toUpperCase();
583 for (int m = 0; m < sequenceMatches.size(); m++)
585 sequence = (SequenceI) sequenceMatches.elementAt(m);
586 // only update start and end positions and shift features if there are
587 // no existing references
588 // TODO: test for legacy where uniprot or EMBL refs exist but no
589 // mappings are made (but content matches retrieved set)
590 boolean updateRefFrame = sequence.getDBRef() == null
591 || sequence.getDBRef().length == 0;
593 // verify sequence against the entry sequence
595 String nonGapped = AlignSeq.extractGaps("-. ",
596 sequence.getSequenceAsString()).toUpperCase();
598 int absStart = entrySeq.indexOf(nonGapped);
599 int mapStart = entry.getStart();
600 jalview.datamodel.Mapping mp;
604 // Is local sequence contained in dataset sequence?
605 absStart = nonGapped.indexOf(entrySeq);
607 { // verification failed.
608 messages.append(sequence.getName()
609 + " SEQUENCE NOT %100 MATCH \n");
613 sbuffer.append(sequence.getName() + " HAS " + absStart
614 + " PREFIXED RESIDUES COMPARED TO " + dbSource + "\n");
616 // + " - ANY SEQUENCE FEATURES"
617 // + " HAVE BEEN ADJUSTED ACCORDINGLY \n");
619 // create valid mapping between matching region of local sequence and
620 // the mapped sequence
621 mp = new Mapping(null, new int[] {
622 sequence.getStart() + absStart,
623 sequence.getStart() + absStart + entrySeq.length() - 1 },
624 new int[] { entry.getStart(),
625 entry.getStart() + entrySeq.length() - 1 }, 1, 1);
626 updateRefFrame = false; // mapping is based on current start/end so
627 // don't modify start and end
632 // update start and end of local sequence to place it in entry's
634 // apply identity map map from whole of local sequence to matching
635 // region of database
637 mp = null; // Mapping.getIdentityMap();
639 // new int[] { absStart+sequence.getStart(),
640 // absStart+sequence.getStart()+entrySeq.length()-1},
641 // new int[] { entry.getStart(), entry.getEnd() }, 1, 1);
642 // relocate local features for updated start
645 if (sequence.getSequenceFeatures() != null)
647 SequenceFeature[] sf = sequence.getSequenceFeatures();
648 int start = sequence.getStart();
649 int end = sequence.getEnd();
650 int startShift = 1 - absStart - start; // how much the features
653 for (int sfi = 0; sfi < sf.length; sfi++)
655 if (sf[sfi].getBegin() >= start && sf[sfi].getEnd() <= end)
657 // shift feature along by absstart
658 sf[sfi].setBegin(sf[sfi].getBegin() + startShift);
659 sf[sfi].setEnd(sf[sfi].getEnd() + startShift);
666 System.out.println("Adding dbrefs to " + sequence.getName()
667 + " from " + dbSource + " sequence : " + entry.getName());
668 sequence.transferAnnotation(entry, mp);
669 // unknownSequences.remove(sequence);
670 int absEnd = absStart + nonGapped.length();
672 if (!trimDatasetSeqs)
674 // insert full length sequence from record
675 sequence.setSequence(entry.getSequenceAsString());
676 sequence.setStart(entry.getStart());
680 // finally, update local sequence reference frame if we're allowed
683 // just fix start/end
684 sequence.setStart(absStart);
685 sequence.setEnd(absEnd);
687 // search for alignment sequences to update coordinate frame for
688 for (int alsq = 0; alsq < alseqs.length; alsq++)
690 if (alseqs[alsq].getDatasetSequence() == sequence)
692 String ngAlsq = AlignSeq.extractGaps("-. ",
693 alseqs[alsq].getSequenceAsString()).toUpperCase();
694 int oldstrt = alseqs[alsq].getStart();
695 alseqs[alsq].setStart(sequence.getSequenceAsString()
696 .toUpperCase().indexOf(ngAlsq)
697 + sequence.getStart());
698 if (oldstrt != alseqs[alsq].getStart())
700 alseqs[alsq].setEnd(ngAlsq.length()
701 + alseqs[alsq].getStart() - 1);
705 // TODO: search for all other references to this dataset sequence, and
707 // TODO: update all AlCodonMappings which involve this alignment
708 // sequence (e.g. Q30167 cdna translation from exon2 product (vamsas
711 // and remove it from the rest
712 // TODO: decide if we should remove annotated sequence from set
713 sdataset.remove(sequence);
714 // TODO: should we make a note of sequences that have received new DB
715 // ids, so we can query all enabled DAS servers for them ?
720 // report the ID/sequence mismatches
721 sbuffer.append(messages);
726 * loop thru and collect additional sequences in Map.
728 * @param sequencesArray
731 private SequenceI[] recoverDbSequences(SequenceI[] sequencesArray)
733 Vector nseq = new Vector();
734 for (int i = 0; sequencesArray != null && i < sequencesArray.length; i++)
736 nseq.addElement(sequencesArray[i]);
737 DBRefEntry dbr[] = sequencesArray[i].getDBRef();
738 jalview.datamodel.Mapping map = null;
739 for (int r = 0; (dbr != null) && r < dbr.length; r++)
741 if ((map = dbr[r].getMap()) != null)
743 if (map.getTo() != null && !nseq.contains(map.getTo()))
745 nseq.addElement(map.getTo());
752 sequencesArray = new SequenceI[nseq.size()];
753 nseq.toArray(sequencesArray);
755 return sequencesArray;