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
60 IProgressIndicator progressWindow;
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 DBRefFetcher object and fetches from the currently selected
95 * set of databases, if this is null then it fetches based on feature settings
98 * fetch references for these SequenceI array
99 * @param progressIndicatorFrame
100 * the frame for progress bar monitoring
102 * array of DbSourceProxy to query references form
103 * @param featureSettings
104 * FeatureSettings to get alternative DbSourceProxy from
105 * @param isNucleotide
106 * indicates if the array of SequenceI are Nucleotides or not
108 public DBRefFetcher(SequenceI[] seqs,
109 IProgressIndicator progressIndicatorFrame,
110 DbSourceProxy[] sources, FeatureSettings featureSettings, boolean isNucleotide)
112 this.progressWindow = progressIndicatorFrame;
113 alseqs = new SequenceI[seqs.length];
114 SequenceI[] ds = new SequenceI[seqs.length];
115 for (int i = 0; i < seqs.length; i++)
118 if (seqs[i].getDatasetSequence() != null)
120 ds[i] = seqs[i].getDatasetSequence();
128 // TODO Jalview 2.5 lots of this code should be in the gui package!
129 sfetcher = jalview.gui.SequenceFetcher
130 .getSequenceFetcherSingleton(progressIndicatorFrame);
131 // set default behaviour for transferring excess sequence data to the
133 trimDsSeqs = Cache.getDefault("TRIM_FETCHED_DATASET_SEQS", true);
136 // af.featureSettings_actionPerformed(null);
137 String[] defdb = null, otherdb = sfetcher
138 .getDbInstances(jalview.ws.dbsources.das.datamodel.DasSequenceSource.class);
139 List<DbSourceProxy> selsources = new ArrayList<DbSourceProxy>();
140 Vector dasselsrc = (featureSettings != null) ? featureSettings
141 .getSelectedSources() : new jalview.gui.DasSourceBrowser()
142 .getSelectedSources();
143 Enumeration<jalviewSourceI> en = dasselsrc.elements();
144 while (en.hasMoreElements())
146 jalviewSourceI src = en.nextElement();
147 List<DbSourceProxy> sp = src.getSequenceSourceProxies();
150 selsources.addAll(sp);
153 Cache.log.debug("Added many Db Sources for :" + src.getTitle());
157 // select appropriate databases based on alignFrame context.
160 defdb = DBRefSource.DNACODINGDBS;
164 defdb = DBRefSource.PROTEINDBS;
166 List<DbSourceProxy> srces = new ArrayList<DbSourceProxy>();
167 for (String ddb : defdb)
169 List<DbSourceProxy> srcesfordb = sfetcher.getSourceProxy(ddb);
170 if (srcesfordb != null)
172 srces.addAll(srcesfordb);
176 // append the selected sequence sources to the default dbs
177 srces.addAll(selsources);
178 dbSources = srces.toArray(new DbSourceProxy[0]);
182 // we assume the caller knows what they're doing and ensured that all the
183 // db source names are valid
189 * retrieve all the das sequence sources and add them to the list of db
190 * sources to retrieve from
192 public void appendAllDasSources()
194 if (dbSources == null)
196 dbSources = new DbSourceProxy[0];
198 // append additional sources
199 DbSourceProxy[] otherdb = sfetcher
200 .getDbSourceProxyInstances(jalview.ws.dbsources.das.datamodel.DasSequenceSource.class);
201 if (otherdb != null && otherdb.length > 0)
203 DbSourceProxy[] newsrc = new DbSourceProxy[dbSources.length
205 System.arraycopy(dbSources, 0, newsrc, 0, dbSources.length);
206 System.arraycopy(otherdb, 0, newsrc, dbSources.length, otherdb.length);
212 * start the fetcher thread
214 * @param waitTillFinished
215 * true to block until the fetcher has finished
217 public void fetchDBRefs(boolean waitTillFinished)
219 Thread thread = new Thread(this);
223 if (waitTillFinished)
230 } catch (Exception ex)
238 * The sequence will be added to a vector of sequences belonging to key which
239 * could be either seq name or dbref id
246 void addSeqId(SequenceI seq, String key)
248 key = key.toUpperCase();
251 if (seqRefs.containsKey(key))
253 seqs = (Vector) seqRefs.get(key);
255 if (seqs != null && !seqs.contains(seq))
257 seqs.addElement(seq);
259 else if (seqs == null)
262 seqs.addElement(seq);
269 seqs.addElement(seq);
272 seqRefs.put(key, seqs);
281 if (dbSources == null)
285 .getString("error.implementation_error_must_init_dbsources"));
288 long startTime = System.currentTimeMillis();
289 if (progressWindow != null)
291 progressWindow.setProgressBar(
292 MessageManager.getString("status.fetching_db_refs"),
297 if (Cache.getDefault("DBREFFETCH_USEPICR", false))
299 picrClient = new uk.ac.ebi.www.picr.AccessionMappingService.AccessionMapperServiceLocator()
300 .getAccessionMapperPort();
302 } catch (Exception e)
304 System.err.println("Couldn't locate PICR service instance.\n");
308 Vector sdataset = new Vector();
309 for (int s = 0; s < dataset.length; s++)
311 sdataset.addElement(dataset[s]);
313 while (sdataset.size() > 0 && db < dbSources.length)
315 int maxqlen = 1; // default number of queries made to at one time
316 System.err.println("Verifying against " + dbSources[db].getDbName());
319 // iterate through db for each remaining un-verified sequence
320 SequenceI[] currSeqs = new SequenceI[sdataset.size()];
321 sdataset.copyInto(currSeqs);// seqs that are to be validated against
323 Vector queries = new Vector(); // generated queries curSeq
324 seqRefs = new Hashtable();
328 jalview.ws.seqfetcher.DbSourceProxy dbsource = dbSources[db];
330 // for moment, we dumbly iterate over all retrieval sources for a
331 // particular database
332 // TODO: introduce multithread multisource queries and logic to remove a
333 // query from other sources if any source for a database returns a
335 maxqlen = dbsource.getMaximumQueryCount();
337 while (queries.size() > 0 || seqIndex < currSeqs.length)
339 if (queries.size() > 0)
341 // Still queries to make for current seqIndex
342 StringBuffer queryString = new StringBuffer("");
343 int numq = 0, nqSize = (maxqlen > queries.size()) ? queries
346 while (queries.size() > 0 && numq < nqSize)
348 String query = (String) queries.elementAt(0);
349 if (dbsource.isValidReference(query))
351 queryString.append((numq == 0) ? "" : dbsource
352 .getAccessionSeparator());
353 queryString.append(query);
356 // remove the extracted query string
357 queries.removeElementAt(0);
359 // make the queries and process the response
360 AlignmentI retrieved = null;
363 if (jalview.bin.Cache.log.isDebugEnabled())
365 jalview.bin.Cache.log.debug("Querying "
366 + dbsource.getDbName() + " with : '"
367 + queryString.toString() + "'");
369 retrieved = dbsource.getSequenceRecords(queryString
371 } catch (Exception ex)
373 ex.printStackTrace();
374 } catch (OutOfMemoryError err)
376 new OOMWarning("retrieving database references ("
377 + queryString.toString() + ")", err);
379 if (retrieved != null)
381 transferReferences(sdataset, dbsource.getDbSource(),
382 retrieved, trimDsSeqs);
387 // make some more strings for use as queries
388 for (int i = 0; (seqIndex < dataset.length) && (i < 50); seqIndex++, i++)
390 SequenceI sequence = dataset[seqIndex];
391 DBRefEntry[] uprefs = jalview.util.DBRefUtils.selectRefs(
392 sequence.getDBRefs(),
393 new String[] { dbsource.getDbSource() }); // jalview.datamodel.DBRefSource.UNIPROT
395 // check for existing dbrefs to use
396 if (uprefs != null && uprefs.length > 0)
398 for (int j = 0; j < uprefs.length; j++)
400 addSeqId(sequence, uprefs[j].getAccessionId());
401 queries.addElement(uprefs[j].getAccessionId()
407 // generate queries from sequence ID string
408 StringTokenizer st = new StringTokenizer(
409 sequence.getName(), "|");
410 while (st.hasMoreTokens())
412 String token = st.nextToken();
413 UPEntry[] presp = null;
414 if (picrClient != null)
416 // resolve the string against PICR to recover valid IDs
419 presp = picrClient.getUPIForAccession(token, null,
420 picrClient.getMappedDatabaseNames(), null,
422 } catch (Exception e)
424 System.err.println("Exception with Picr for '"
429 if (presp != null && presp.length > 0)
431 for (int id = 0; id < presp.length; id++)
433 // construct sequences from response if sequences are
434 // present, and do a transferReferences
435 // otherwise transfer non sequence x-references directly.
438 .println("Validated ID against PICR... (for what its worth):"
440 addSeqId(sequence, token);
441 queries.addElement(token.toUpperCase());
446 // System.out.println("Not querying source with token="+token+"\n");
447 addSeqId(sequence, token);
448 queries.addElement(token.toUpperCase());
456 // advance to next database
458 } // all databases have been queries.
459 if (sbuffer.length() > 0)
461 output.setText(MessageManager
462 .getString("label.your_sequences_have_been_verified")
463 + sbuffer.toString());
464 Desktop.addInternalFrame(output,
465 MessageManager.getString("label.sequence_names_updated"),
467 // The above is the dataset, we must now find out the index
468 // of the viewed sequence
471 if (progressWindow != null)
473 progressWindow.setProgressBar(
474 MessageManager.getString("label.dbref_search_completed"),
476 // promptBeforeBlast();
485 * Verify local sequences in seqRefs against the retrieved sequence database
488 * @param trimDatasetSeqs
491 void transferReferences(Vector sdataset, String dbSource,
492 AlignmentI retrievedAl, boolean trimDatasetSeqs) // File
495 System.out.println("trimming ? " + trimDatasetSeqs);
496 if (retrievedAl == null || retrievedAl.getHeight() == 0)
500 SequenceI[] retrieved = recoverDbSequences(retrievedAl
501 .getSequencesArray());
502 SequenceI sequence = null;
503 boolean transferred = false;
504 StringBuffer messages = new StringBuffer();
506 // Vector entries = new Uniprot().getUniprotEntries(file);
508 int i, iSize = retrieved.length; // entries == null ? 0 : entries.size();
509 // UniprotEntry entry;
510 for (i = 0; i < iSize; i++)
512 SequenceI entry = retrieved[i]; // (UniprotEntry) entries.elementAt(i);
514 // Work out which sequences this sequence matches,
515 // taking into account all accessionIds and names in the file
516 Vector sequenceMatches = new Vector();
517 // look for corresponding accession ids
518 DBRefEntry[] entryRefs = jalview.util.DBRefUtils.selectRefs(
519 entry.getDBRefs(), new String[] { dbSource });
520 if (entryRefs == null)
523 .println("Dud dbSource string ? no entryrefs selected for "
524 + dbSource + " on " + entry.getName());
527 for (int j = 0; j < entryRefs.length; j++)
529 String accessionId = entryRefs[j].getAccessionId(); // .getAccession().elementAt(j).toString();
530 // match up on accessionId
531 if (seqRefs.containsKey(accessionId.toUpperCase()))
533 Vector seqs = (Vector) seqRefs.get(accessionId);
534 for (int jj = 0; jj < seqs.size(); jj++)
536 sequence = (SequenceI) seqs.elementAt(jj);
537 if (!sequenceMatches.contains(sequence))
539 sequenceMatches.addElement(sequence);
544 if (sequenceMatches.size() == 0)
546 // failed to match directly on accessionId==query so just compare all
547 // sequences to entry
548 Enumeration e = seqRefs.keys();
549 while (e.hasMoreElements())
551 Vector sqs = (Vector) seqRefs.get(e.nextElement());
552 if (sqs != null && sqs.size() > 0)
554 Enumeration sqe = sqs.elements();
555 while (sqe.hasMoreElements())
557 sequenceMatches.addElement(sqe.nextElement());
562 // look for corresponding names
563 // this is uniprot specific ?
564 // could be useful to extend this so we try to find any 'significant'
565 // information in common between two sequence objects.
567 * DBRefEntry[] entryRefs =
568 * jalview.util.DBRefUtils.selectRefs(entry.getDBRef(), new String[] {
569 * dbSource }); for (int j = 0; j < entry.getName().size(); j++) { String
570 * name = entry.getName().elementAt(j).toString(); if
571 * (seqRefs.containsKey(name)) { Vector seqs = (Vector) seqRefs.get(name);
572 * for (int jj = 0; jj < seqs.size(); jj++) { sequence = (SequenceI)
573 * seqs.elementAt(jj); if (!sequenceMatches.contains(sequence)) {
574 * sequenceMatches.addElement(sequence); } } } }
576 // sequenceMatches now contains the set of all sequences associated with
577 // the returned db record
578 String entrySeq = entry.getSequenceAsString().toUpperCase();
579 for (int m = 0; m < sequenceMatches.size(); m++)
581 sequence = (SequenceI) sequenceMatches.elementAt(m);
582 // only update start and end positions and shift features if there are
583 // no existing references
584 // TODO: test for legacy where uniprot or EMBL refs exist but no
585 // mappings are made (but content matches retrieved set)
586 boolean updateRefFrame = sequence.getDBRefs() == null
587 || sequence.getDBRefs().length == 0;
589 // verify sequence against the entry sequence
591 String nonGapped = AlignSeq.extractGaps("-. ",
592 sequence.getSequenceAsString()).toUpperCase();
594 int absStart = entrySeq.indexOf(nonGapped);
597 final int sequenceStart = sequence.getStart();
600 // Is local sequence contained in dataset sequence?
601 absStart = nonGapped.indexOf(entrySeq);
603 { // verification failed.
604 messages.append(sequence.getName()
605 + " SEQUENCE NOT %100 MATCH \n");
609 sbuffer.append(sequence.getName() + " HAS " + absStart
610 + " PREFIXED RESIDUES COMPARED TO " + dbSource + "\n");
612 // + " - ANY SEQUENCE FEATURES"
613 // + " HAVE BEEN ADJUSTED ACCORDINGLY \n");
615 // create valid mapping between matching region of local sequence and
616 // the mapped sequence
617 mp = new Mapping(null, new int[] { sequenceStart + absStart,
618 sequenceStart + absStart + entrySeq.length() - 1 }, new int[]
619 { entry.getStart(), entry.getStart() + entrySeq.length() - 1 },
621 updateRefFrame = false; // mapping is based on current start/end so
622 // don't modify start and end
627 // update start and end of local sequence to place it in entry's
629 // apply identity map map from whole of local sequence to matching
630 // region of database
632 mp = null; // Mapping.getIdentityMap();
634 // new int[] { absStart+sequence.getStart(),
635 // absStart+sequence.getStart()+entrySeq.length()-1},
636 // new int[] { entry.getStart(), entry.getEnd() }, 1, 1);
637 // relocate local features for updated start
640 if (sequence.getSequenceFeatures() != null)
642 SequenceFeature[] sf = sequence.getSequenceFeatures();
643 int start = sequenceStart;
644 int end = sequence.getEnd();
645 int startShift = 1 - absStart - start; // how much the features
648 for (int sfi = 0; sfi < sf.length; sfi++)
650 if (sf[sfi].getBegin() >= start && sf[sfi].getEnd() <= end)
652 // shift feature along by absstart
653 sf[sfi].setBegin(sf[sfi].getBegin() + startShift);
654 sf[sfi].setEnd(sf[sfi].getEnd() + startShift);
661 System.out.println("Adding dbrefs to " + sequence.getName()
662 + " from " + dbSource + " sequence : " + entry.getName());
663 sequence.transferAnnotation(entry, mp);
664 // unknownSequences.remove(sequence);
665 absStart += entry.getStart();
666 int absEnd = absStart + nonGapped.length() - 1;
667 if (!trimDatasetSeqs)
669 // insert full length sequence from record
670 sequence.setSequence(entry.getSequenceAsString());
671 sequence.setStart(entry.getStart());
675 // finally, update local sequence reference frame if we're allowed
678 // just fix start/end
679 sequence.setStart(absStart);
680 sequence.setEnd(absEnd);
682 // search for alignment sequences to update coordinate frame for
683 for (int alsq = 0; alsq < alseqs.length; alsq++)
685 if (alseqs[alsq].getDatasetSequence() == sequence)
687 String ngAlsq = AlignSeq.extractGaps("-. ",
688 alseqs[alsq].getSequenceAsString()).toUpperCase();
689 int oldstrt = alseqs[alsq].getStart();
690 alseqs[alsq].setStart(sequence.getSequenceAsString()
691 .toUpperCase().indexOf(ngAlsq)
692 + sequence.getStart());
693 if (oldstrt != alseqs[alsq].getStart())
695 alseqs[alsq].setEnd(ngAlsq.length()
696 + alseqs[alsq].getStart() - 1);
700 // TODO: search for all other references to this dataset sequence, and
702 // TODO: update all AlCodonMappings which involve this alignment
703 // sequence (e.g. Q30167 cdna translation from exon2 product (vamsas
706 // and remove it from the rest
707 // TODO: decide if we should remove annotated sequence from set
708 sdataset.remove(sequence);
709 // TODO: should we make a note of sequences that have received new DB
710 // ids, so we can query all enabled DAS servers for them ?
715 // report the ID/sequence mismatches
716 sbuffer.append(messages);
721 * loop thru and collect additional sequences in Map.
723 * @param sequencesArray
726 private SequenceI[] recoverDbSequences(SequenceI[] sequencesArray)
728 Vector nseq = new Vector();
729 for (int i = 0; sequencesArray != null && i < sequencesArray.length; i++)
731 nseq.addElement(sequencesArray[i]);
732 DBRefEntry dbr[] = sequencesArray[i].getDBRefs();
733 jalview.datamodel.Mapping map = null;
734 for (int r = 0; (dbr != null) && r < dbr.length; r++)
736 if ((map = dbr[r].getMap()) != null)
738 if (map.getTo() != null && !nseq.contains(map.getTo()))
740 nseq.addElement(map.getTo());
747 sequencesArray = new SequenceI[nseq.size()];
748 nseq.toArray(sequencesArray);
750 return sequencesArray;