2 * Jalview - A Sequence Alignment Editor and Viewer (Version 2.8.1)
3 * Copyright (C) 2014 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 of the License, or (at your option) any later version.
11 * Jalview is distributed in the hope that it will be useful, but
12 * WITHOUT ANY WARRANTY; without even the implied warranty
13 * of MERCHANTABILITY or FITNESS FOR A PARTICULAR
14 * PURPOSE. See the GNU General Public License for more details.
16 * You should have received a copy of the GNU General Public License along with Jalview. If not, see <http://www.gnu.org/licenses/>.
17 * The Jalview Authors are detailed in the 'AUTHORS' file.
21 import jalview.analysis.AlignSeq;
22 import jalview.bin.Cache;
23 import jalview.datamodel.AlignmentI;
24 import jalview.datamodel.DBRefEntry;
25 import jalview.datamodel.DBRefSource;
26 import jalview.datamodel.Mapping;
27 import jalview.datamodel.SequenceFeature;
28 import jalview.datamodel.SequenceI;
29 import jalview.gui.AlignFrame;
30 import jalview.gui.CutAndPasteTransfer;
31 import jalview.gui.Desktop;
32 import jalview.gui.IProgressIndicator;
33 import jalview.gui.OOMWarning;
34 import jalview.util.MessageManager;
35 import jalview.ws.dbsources.das.api.jalviewSourceI;
36 import jalview.ws.seqfetcher.DbSourceProxy;
38 import java.util.ArrayList;
39 import java.util.Enumeration;
40 import java.util.Hashtable;
41 import java.util.List;
42 import java.util.StringTokenizer;
43 import java.util.Vector;
45 import uk.ac.ebi.picr.model.UPEntry;
48 * Implements a runnable for validating a sequence against external databases
49 * and then propagating references and features onto the sequence(s)
54 public class DBRefFetcher implements Runnable
58 IProgressIndicator af;
60 CutAndPasteTransfer output = new CutAndPasteTransfer();
62 StringBuffer sbuffer = new StringBuffer();
64 boolean running = false;
67 * picr client instance
69 uk.ac.ebi.www.picr.AccessionMappingService.AccessionMapperInterface picrClient = null;
71 // /This will be a collection of Vectors of sequenceI refs.
72 // The key will be the seq name or accession id of the seq
75 DbSourceProxy[] dbSources;
77 SequenceFetcher sfetcher;
79 private SequenceI[] alseqs;
82 * when true - retrieved sequences will be trimmed to cover longest derived alignment sequence
84 private boolean trimDsSeqs=true;
91 * Creates a new SequenceFeatureFetcher object and fetches from the currently
92 * selected set of databases.
95 * fetch references for these sequences
97 * the parent alignframe for progress bar monitoring.
99 public DBRefFetcher(SequenceI[] seqs, AlignFrame af)
101 this(seqs, af, null);
105 * Creates a new SequenceFeatureFetcher object and fetches from the currently
106 * selected set of databases.
109 * fetch references for these sequences
111 * the parent alignframe for progress bar monitoring.
113 * array of database source strings to query references from
115 public DBRefFetcher(SequenceI[] seqs, AlignFrame af,
116 DbSourceProxy[] sources)
119 alseqs = new SequenceI[seqs.length];
120 SequenceI[] ds = new SequenceI[seqs.length];
121 for (int i = 0; i < seqs.length; i++)
124 if (seqs[i].getDatasetSequence() != null)
125 ds[i] = seqs[i].getDatasetSequence();
130 // TODO Jalview 2.5 lots of this code should be in the gui package!
131 sfetcher = jalview.gui.SequenceFetcher.getSequenceFetcherSingleton(af);
132 // set default behaviour for transferring excess sequence data to the dataset
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 = (af.featureSettings != null) ? af.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.
158 if (af.getViewport().getAlignment().isNucleotide())
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);
280 if (dbSources == null)
282 throw new Error("Implementation error. Must initialise dbSources");
285 long startTime = System.currentTimeMillis();
286 af.setProgressBar("Fetching db refs", startTime);
289 if (Cache.getDefault("DBREFFETCH_USEPICR", false))
291 picrClient = new uk.ac.ebi.www.picr.AccessionMappingService.AccessionMapperServiceLocator()
292 .getAccessionMapperPort();
294 } catch (Exception e)
296 System.err.println("Couldn't locate PICR service instance.\n");
300 Vector sdataset = new Vector();
301 for (int s = 0; s < dataset.length; s++)
303 sdataset.addElement(dataset[s]);
305 while (sdataset.size() > 0 && db < dbSources.length)
307 int maxqlen = 1; // default number of queries made to at one time
308 System.err.println("Verifying against " + dbSources[db].getDbName());
311 // iterate through db for each remaining un-verified sequence
312 SequenceI[] currSeqs = new SequenceI[sdataset.size()];
313 sdataset.copyInto(currSeqs);// seqs that are to be validated against
315 Vector queries = new Vector(); // generated queries curSeq
316 seqRefs = new Hashtable();
320 jalview.ws.seqfetcher.DbSourceProxy dbsource = dbSources[db];
322 // for moment, we dumbly iterate over all retrieval sources for a
323 // particular database
324 // TODO: introduce multithread multisource queries and logic to remove a
325 // query from other sources if any source for a database returns a
327 if (dbsource.getDbSourceProperties().containsKey(
328 DBRefSource.MULTIACC))
330 maxqlen = ((Integer) dbsource.getDbSourceProperties().get(
331 DBRefSource.MULTIACC)).intValue();
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.getDBRef(), new String[]
393 { 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.getString("label.your_sequences_have_been_verified")
462 + sbuffer.toString());
463 Desktop.addInternalFrame(output, MessageManager.getString("label.sequence_names_updated"), 600, 300);
464 // The above is the dataset, we must now find out the index
465 // of the viewed sequence
469 af.setProgressBar(MessageManager.getString("label.dbref_search_completed"), startTime);
470 // promptBeforeBlast();
477 * Verify local sequences in seqRefs against the retrieved sequence database
479 * @param trimDatasetSeqs
482 void transferReferences(Vector sdataset, String dbSource,
483 AlignmentI retrievedAl, boolean trimDatasetSeqs) // File
486 System.out.println("trimming ? "+trimDatasetSeqs);
487 if (retrievedAl == null || retrievedAl.getHeight() == 0)
491 SequenceI[] retrieved = recoverDbSequences(retrievedAl
492 .getSequencesArray());
493 SequenceI sequence = null;
494 boolean transferred = false;
495 StringBuffer messages = new StringBuffer();
497 // Vector entries = new Uniprot().getUniprotEntries(file);
499 int i, iSize = retrieved.length; // entries == null ? 0 : entries.size();
500 // UniprotEntry entry;
501 for (i = 0; i < iSize; i++)
503 SequenceI entry = retrieved[i]; // (UniprotEntry) entries.elementAt(i);
505 // Work out which sequences this sequence matches,
506 // taking into account all accessionIds and names in the file
507 Vector sequenceMatches = new Vector();
508 // look for corresponding accession ids
509 DBRefEntry[] entryRefs = jalview.util.DBRefUtils.selectRefs(
510 entry.getDBRef(), new String[]
512 if (entryRefs == null)
515 .println("Dud dbSource string ? no entryrefs selected for "
516 + dbSource + " on " + entry.getName());
519 for (int j = 0; j < entryRefs.length; j++)
521 String accessionId = entryRefs[j].getAccessionId(); // .getAccession().elementAt(j).toString();
522 // match up on accessionId
523 if (seqRefs.containsKey(accessionId.toUpperCase()))
525 Vector seqs = (Vector) seqRefs.get(accessionId);
526 for (int jj = 0; jj < seqs.size(); jj++)
528 sequence = (SequenceI) seqs.elementAt(jj);
529 if (!sequenceMatches.contains(sequence))
531 sequenceMatches.addElement(sequence);
536 if (sequenceMatches.size() == 0)
538 // failed to match directly on accessionId==query so just compare all
539 // sequences to entry
540 Enumeration e = seqRefs.keys();
541 while (e.hasMoreElements())
543 Vector sqs = (Vector) seqRefs.get(e.nextElement());
544 if (sqs != null && sqs.size() > 0)
546 Enumeration sqe = sqs.elements();
547 while (sqe.hasMoreElements())
549 sequenceMatches.addElement(sqe.nextElement());
554 // look for corresponding names
555 // this is uniprot specific ?
556 // could be useful to extend this so we try to find any 'significant'
557 // information in common between two sequence objects.
559 * DBRefEntry[] entryRefs =
560 * jalview.util.DBRefUtils.selectRefs(entry.getDBRef(), new String[] {
561 * dbSource }); for (int j = 0; j < entry.getName().size(); j++) { String
562 * name = entry.getName().elementAt(j).toString(); if
563 * (seqRefs.containsKey(name)) { Vector seqs = (Vector) seqRefs.get(name);
564 * for (int jj = 0; jj < seqs.size(); jj++) { sequence = (SequenceI)
565 * seqs.elementAt(jj); if (!sequenceMatches.contains(sequence)) {
566 * sequenceMatches.addElement(sequence); } } } }
568 // sequenceMatches now contains the set of all sequences associated with
569 // the returned db record
570 String entrySeq = entry.getSequenceAsString().toUpperCase();
571 for (int m = 0; m < sequenceMatches.size(); m++)
573 sequence = (SequenceI) sequenceMatches.elementAt(m);
574 // only update start and end positions and shift features if there are
575 // no existing references
576 // TODO: test for legacy where uniprot or EMBL refs exist but no
577 // mappings are made (but content matches retrieved set)
578 boolean updateRefFrame = sequence.getDBRef() == null
579 || sequence.getDBRef().length == 0;
581 // verify sequence against the entry sequence
583 String nonGapped = AlignSeq.extractGaps("-. ",
584 sequence.getSequenceAsString()).toUpperCase();
586 int absStart = entrySeq.indexOf(nonGapped);
587 int mapStart = entry.getStart();
588 jalview.datamodel.Mapping mp;
592 // Is local sequence contained in dataset sequence?
593 absStart = nonGapped.indexOf(entrySeq);
595 { // verification failed.
596 messages.append(sequence.getName()
597 + " SEQUENCE NOT %100 MATCH \n");
601 sbuffer.append(sequence.getName() + " HAS " + absStart
602 + " PREFIXED RESIDUES COMPARED TO " + dbSource + "\n");
604 // + " - ANY SEQUENCE FEATURES"
605 // + " HAVE BEEN ADJUSTED ACCORDINGLY \n");
607 // create valid mapping between matching region of local sequence and
608 // the mapped sequence
609 mp = new Mapping(null, new int[]
610 { sequence.getStart() + absStart,
611 sequence.getStart() + absStart + entrySeq.length() - 1 },
614 entry.getStart() + entrySeq.length() - 1 }, 1, 1);
615 updateRefFrame = false; // mapping is based on current start/end so
616 // don't modify start and end
621 // update start and end of local sequence to place it in entry's
623 // apply identity map map from whole of local sequence to matching
624 // region of database
626 mp = null; // Mapping.getIdentityMap();
628 // new int[] { absStart+sequence.getStart(),
629 // absStart+sequence.getStart()+entrySeq.length()-1},
630 // new int[] { entry.getStart(), entry.getEnd() }, 1, 1);
631 // relocate local features for updated start
634 if (sequence.getSequenceFeatures() != null)
636 SequenceFeature[] sf = sequence.getSequenceFeatures();
637 int start = sequence.getStart();
638 int end = sequence.getEnd();
639 int startShift = 1 - absStart - start; // how much the features
642 for (int sfi = 0; sfi < sf.length; sfi++)
644 if (sf[sfi].getBegin() >= start && sf[sfi].getEnd() <= end)
646 // shift feature along by absstart
647 sf[sfi].setBegin(sf[sfi].getBegin() + startShift);
648 sf[sfi].setEnd(sf[sfi].getEnd() + startShift);
655 System.out.println("Adding dbrefs to " + sequence.getName()
656 + " from " + dbSource + " sequence : " + entry.getName());
657 sequence.transferAnnotation(entry, mp);
658 // unknownSequences.remove(sequence);
659 int absEnd = absStart + nonGapped.length();
661 if (!trimDatasetSeqs) {
662 // insert full length sequence from record
663 sequence.setSequence(entry.getSequenceAsString());
664 sequence.setStart(entry.getStart());
668 // finally, update local sequence reference frame if we're allowed
669 if (trimDatasetSeqs) {
670 // just fix start/end
671 sequence.setStart(absStart);
672 sequence.setEnd(absEnd);
674 // search for alignment sequences to update coordinate frame for
675 for (int alsq = 0; alsq < alseqs.length; alsq++)
677 if (alseqs[alsq].getDatasetSequence() == sequence)
679 String ngAlsq = AlignSeq.extractGaps("-. ",
680 alseqs[alsq].getSequenceAsString()).toUpperCase();
681 int oldstrt = alseqs[alsq].getStart();
682 alseqs[alsq].setStart(sequence.getSequenceAsString()
683 .toUpperCase().indexOf(ngAlsq)
684 + sequence.getStart());
685 if (oldstrt != alseqs[alsq].getStart())
687 alseqs[alsq].setEnd(ngAlsq.length()
688 + alseqs[alsq].getStart() - 1);
692 // TODO: search for all other references to this dataset sequence, and
694 // TODO: update all AlCodonMappings which involve this alignment
695 // sequence (e.g. Q30167 cdna translation from exon2 product (vamsas
698 // and remove it from the rest
699 // TODO: decide if we should remove annotated sequence from set
700 sdataset.remove(sequence);
701 // TODO: should we make a note of sequences that have received new DB
702 // ids, so we can query all enabled DAS servers for them ?
707 // report the ID/sequence mismatches
708 sbuffer.append(messages);
713 * loop thru and collect additional sequences in Map.
715 * @param sequencesArray
718 private SequenceI[] recoverDbSequences(SequenceI[] sequencesArray)
720 Vector nseq = new Vector();
721 for (int i = 0; sequencesArray != null && i < sequencesArray.length; i++)
723 nseq.addElement(sequencesArray[i]);
724 DBRefEntry dbr[] = sequencesArray[i].getDBRef();
725 jalview.datamodel.Mapping map = null;
726 for (int r = 0; (dbr != null) && r < dbr.length; r++)
728 if ((map = dbr[r].getMap()) != null)
730 if (map.getTo() != null && !nseq.contains(map.getTo()))
732 nseq.addElement(map.getTo());
739 sequencesArray = new SequenceI[nseq.size()];
740 nseq.toArray(sequencesArray);
742 return sequencesArray;