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);
179 // append the PDB data source, since it is 'special', catering for both
180 // nucleotide and protein
181 srces.addAll(sfetcher.getSourceProxy(DBRefSource.PDB));
183 // append the selected sequence sources to the default dbs
184 srces.addAll(selsources);
185 dbSources = srces.toArray(new DbSourceProxy[0]);
189 // we assume the caller knows what they're doing and ensured that all the
190 // db source names are valid
196 * Constructor with only sequences provided
200 public DBRefFetcher(SequenceI[] sequences)
202 this(sequences, null, null, null, false);
206 * Add a listener to be notified when sequence fetching is complete
210 public void addListener(FetchFinishedListenerI l)
216 * retrieve all the das sequence sources and add them to the list of db
217 * sources to retrieve from
219 public void appendAllDasSources()
221 if (dbSources == null)
223 dbSources = new DbSourceProxy[0];
225 // append additional sources
226 DbSourceProxy[] otherdb = sfetcher
227 .getDbSourceProxyInstances(jalview.ws.dbsources.das.datamodel.DasSequenceSource.class);
228 if (otherdb != null && otherdb.length > 0)
230 DbSourceProxy[] newsrc = new DbSourceProxy[dbSources.length
232 System.arraycopy(dbSources, 0, newsrc, 0, dbSources.length);
233 System.arraycopy(otherdb, 0, newsrc, dbSources.length, otherdb.length);
239 * start the fetcher thread
241 * @param waitTillFinished
242 * true to block until the fetcher has finished
244 public void fetchDBRefs(boolean waitTillFinished)
246 Thread thread = new Thread(this);
250 if (waitTillFinished)
257 } catch (Exception ex)
265 * The sequence will be added to a vector of sequences belonging to key which
266 * could be either seq name or dbref id
273 void addSeqId(SequenceI seq, String key)
275 key = key.toUpperCase();
278 if (seqRefs.containsKey(key))
280 seqs = (Vector) seqRefs.get(key);
282 if (seqs != null && !seqs.contains(seq))
284 seqs.addElement(seq);
286 else if (seqs == null)
289 seqs.addElement(seq);
296 seqs.addElement(seq);
299 seqRefs.put(key, seqs);
308 if (dbSources == null)
312 .getString("error.implementation_error_must_init_dbsources"));
315 long startTime = System.currentTimeMillis();
316 if (progressWindow != null)
318 progressWindow.setProgressBar(
319 MessageManager.getString("status.fetching_db_refs"),
324 if (Cache.getDefault("DBREFFETCH_USEPICR", false))
326 picrClient = new uk.ac.ebi.www.picr.AccessionMappingService.AccessionMapperServiceLocator()
327 .getAccessionMapperPort();
329 } catch (Exception e)
331 System.err.println("Couldn't locate PICR service instance.\n");
335 Vector sdataset = new Vector();
336 for (int s = 0; s < dataset.length; s++)
338 sdataset.addElement(dataset[s]);
340 while (sdataset.size() > 0 && db < dbSources.length)
342 int maxqlen = 1; // default number of queries made to at one time
343 System.err.println("Verifying against " + dbSources[db].getDbName());
346 // iterate through db for each remaining un-verified sequence
347 SequenceI[] currSeqs = new SequenceI[sdataset.size()];
348 sdataset.copyInto(currSeqs);// seqs that are to be validated against
350 Vector queries = new Vector(); // generated queries curSeq
351 seqRefs = new Hashtable();
355 jalview.ws.seqfetcher.DbSourceProxy dbsource = dbSources[db];
357 // for moment, we dumbly iterate over all retrieval sources for a
358 // particular database
359 // TODO: introduce multithread multisource queries and logic to remove a
360 // query from other sources if any source for a database returns a
362 maxqlen = dbsource.getMaximumQueryCount();
364 while (queries.size() > 0 || seqIndex < currSeqs.length)
366 if (queries.size() > 0)
368 // Still queries to make for current seqIndex
369 StringBuffer queryString = new StringBuffer("");
370 int numq = 0, nqSize = (maxqlen > queries.size()) ? queries
373 while (queries.size() > 0 && numq < nqSize)
375 String query = (String) queries.elementAt(0);
376 if (dbsource.isValidReference(query))
378 queryString.append((numq == 0) ? "" : dbsource
379 .getAccessionSeparator());
380 queryString.append(query);
383 // remove the extracted query string
384 queries.removeElementAt(0);
386 // make the queries and process the response
387 AlignmentI retrieved = null;
390 if (jalview.bin.Cache.log.isDebugEnabled())
392 jalview.bin.Cache.log.debug("Querying "
393 + dbsource.getDbName() + " with : '"
394 + queryString.toString() + "'");
396 retrieved = dbsource.getSequenceRecords(queryString
398 } catch (Exception ex)
400 ex.printStackTrace();
401 } catch (OutOfMemoryError err)
403 new OOMWarning("retrieving database references ("
404 + queryString.toString() + ")", err);
406 if (retrieved != null)
408 transferReferences(sdataset, dbsource.getDbSource(),
409 retrieved, trimDsSeqs);
414 // make some more strings for use as queries
415 for (int i = 0; (seqIndex < dataset.length) && (i < 50); seqIndex++, i++)
417 SequenceI sequence = dataset[seqIndex];
418 DBRefEntry[] uprefs = jalview.util.DBRefUtils.selectRefs(
419 sequence.getDBRefs(),
420 new String[] { dbsource.getDbSource() }); // jalview.datamodel.DBRefSource.UNIPROT
422 // check for existing dbrefs to use
423 if (uprefs != null && uprefs.length > 0)
425 for (int j = 0; j < uprefs.length; j++)
427 addSeqId(sequence, uprefs[j].getAccessionId());
428 queries.addElement(uprefs[j].getAccessionId()
434 // generate queries from sequence ID string
435 StringTokenizer st = new StringTokenizer(
436 sequence.getName(), "|");
437 while (st.hasMoreTokens())
439 String token = st.nextToken();
440 UPEntry[] presp = null;
441 if (picrClient != null)
443 // resolve the string against PICR to recover valid IDs
446 presp = picrClient.getUPIForAccession(token, null,
447 picrClient.getMappedDatabaseNames(), null,
449 } catch (Exception e)
451 System.err.println("Exception with Picr for '"
456 if (presp != null && presp.length > 0)
458 for (int id = 0; id < presp.length; id++)
460 // construct sequences from response if sequences are
461 // present, and do a transferReferences
462 // otherwise transfer non sequence x-references directly.
465 .println("Validated ID against PICR... (for what its worth):"
467 addSeqId(sequence, token);
468 queries.addElement(token.toUpperCase());
473 // System.out.println("Not querying source with token="+token+"\n");
474 addSeqId(sequence, token);
475 queries.addElement(token.toUpperCase());
483 // advance to next database
485 } // all databases have been queries.
486 if (sbuffer.length() > 0)
488 output.setText(MessageManager
489 .getString("label.your_sequences_have_been_verified")
490 + sbuffer.toString());
491 Desktop.addInternalFrame(output,
492 MessageManager.getString("label.sequence_names_updated"),
494 // The above is the dataset, we must now find out the index
495 // of the viewed sequence
498 if (progressWindow != null)
500 progressWindow.setProgressBar(
501 MessageManager.getString("label.dbref_search_completed"),
505 for (FetchFinishedListenerI listener : listeners)
513 * Verify local sequences in seqRefs against the retrieved sequence database
516 * @param trimDatasetSeqs
519 void transferReferences(Vector sdataset, String dbSource,
520 AlignmentI retrievedAl, boolean trimDatasetSeqs) // File
523 System.out.println("trimming ? " + trimDatasetSeqs);
524 if (retrievedAl == null || retrievedAl.getHeight() == 0)
528 SequenceI[] retrieved = recoverDbSequences(retrievedAl
529 .getSequencesArray());
530 SequenceI sequence = null;
531 boolean transferred = false;
532 StringBuffer messages = new StringBuffer();
534 // Vector entries = new Uniprot().getUniprotEntries(file);
536 int i, iSize = retrieved.length; // entries == null ? 0 : entries.size();
537 // UniprotEntry entry;
538 for (i = 0; i < iSize; i++)
540 SequenceI entry = retrieved[i]; // (UniprotEntry) entries.elementAt(i);
542 // Work out which sequences this sequence matches,
543 // taking into account all accessionIds and names in the file
544 Vector sequenceMatches = new Vector();
545 // look for corresponding accession ids
546 DBRefEntry[] entryRefs = jalview.util.DBRefUtils.selectRefs(
547 entry.getDBRefs(), new String[] { dbSource });
548 if (entryRefs == null)
551 .println("Dud dbSource string ? no entryrefs selected for "
552 + dbSource + " on " + entry.getName());
555 for (int j = 0; j < entryRefs.length; j++)
557 String accessionId = entryRefs[j].getAccessionId(); // .getAccession().elementAt(j).toString();
558 // match up on accessionId
559 if (seqRefs.containsKey(accessionId.toUpperCase()))
561 Vector seqs = (Vector) seqRefs.get(accessionId);
562 for (int jj = 0; jj < seqs.size(); jj++)
564 sequence = (SequenceI) seqs.elementAt(jj);
565 if (!sequenceMatches.contains(sequence))
567 sequenceMatches.addElement(sequence);
572 if (sequenceMatches.size() == 0)
574 // failed to match directly on accessionId==query so just compare all
575 // sequences to entry
576 Enumeration e = seqRefs.keys();
577 while (e.hasMoreElements())
579 Vector sqs = (Vector) seqRefs.get(e.nextElement());
580 if (sqs != null && sqs.size() > 0)
582 Enumeration sqe = sqs.elements();
583 while (sqe.hasMoreElements())
585 sequenceMatches.addElement(sqe.nextElement());
590 // look for corresponding names
591 // this is uniprot specific ?
592 // could be useful to extend this so we try to find any 'significant'
593 // information in common between two sequence objects.
595 * DBRefEntry[] entryRefs =
596 * jalview.util.DBRefUtils.selectRefs(entry.getDBRef(), new String[] {
597 * dbSource }); for (int j = 0; j < entry.getName().size(); j++) { String
598 * name = entry.getName().elementAt(j).toString(); if
599 * (seqRefs.containsKey(name)) { Vector seqs = (Vector) seqRefs.get(name);
600 * for (int jj = 0; jj < seqs.size(); jj++) { sequence = (SequenceI)
601 * seqs.elementAt(jj); if (!sequenceMatches.contains(sequence)) {
602 * sequenceMatches.addElement(sequence); } } } }
604 // sequenceMatches now contains the set of all sequences associated with
605 // the returned db record
606 String entrySeq = entry.getSequenceAsString().toUpperCase();
607 for (int m = 0; m < sequenceMatches.size(); m++)
609 sequence = (SequenceI) sequenceMatches.elementAt(m);
610 // only update start and end positions and shift features if there are
611 // no existing references
612 // TODO: test for legacy where uniprot or EMBL refs exist but no
613 // mappings are made (but content matches retrieved set)
614 boolean updateRefFrame = sequence.getDBRefs() == null
615 || sequence.getDBRefs().length == 0;
617 // verify sequence against the entry sequence
619 String nonGapped = AlignSeq.extractGaps("-. ",
620 sequence.getSequenceAsString()).toUpperCase();
622 int absStart = entrySeq.indexOf(nonGapped);
625 final int sequenceStart = sequence.getStart();
628 // couldn't find local sequence in sequence from database, so check if
629 // the database sequence is a subsequence of local sequence
630 absStart = nonGapped.indexOf(entrySeq);
633 // verification failed. couldn't find any relationship between
634 // entrySeq and local sequence
635 messages.append(sequence.getName()
636 + " SEQUENCE NOT %100 MATCH \n");
640 * found match for the whole of the database sequence within the local
641 * sequence's reference frame.
644 sbuffer.append(sequence.getName() + " HAS " + absStart
645 + " PREFIXED RESIDUES COMPARED TO " + dbSource + "\n");
648 * So create a mapping to the external entry from the matching region of
649 * the local sequence, and leave local start/end untouched.
651 mp = new Mapping(null, new int[] { sequenceStart + absStart,
652 sequenceStart + absStart + entrySeq.length() - 1 }, new int[]
653 { entry.getStart(), entry.getStart() + entrySeq.length() - 1 },
655 updateRefFrame = false;
660 * found a match for the local sequence within sequence from
661 * the external database
665 // update start and end of local sequence to place it in entry's
667 // apply identity map map from whole of local sequence to matching
668 // region of database
670 mp = null; // Mapping.getIdentityMap();
672 // new int[] { absStart+sequence.getStart(),
673 // absStart+sequence.getStart()+entrySeq.length()-1},
674 // new int[] { entry.getStart(), entry.getEnd() }, 1, 1);
675 // relocate local features for updated start
679 if (sequence.getSequenceFeatures() != null)
682 * relocate existing sequence features by offset
684 SequenceFeature[] sf = sequence.getSequenceFeatures();
685 int start = sequenceStart;
686 int end = sequence.getEnd();
687 int startShift = 1 - absStart - start; // how much the features
690 for (int sfi = 0; sfi < sf.length; sfi++)
692 if (sf[sfi].getBegin() >= start && sf[sfi].getEnd() <= end)
694 // shift feature along by absstart
695 sf[sfi].setBegin(sf[sfi].getBegin() + startShift);
696 sf[sfi].setEnd(sf[sfi].getEnd() + startShift);
703 System.out.println("Adding dbrefs to " + sequence.getName()
704 + " from " + dbSource + " sequence : " + entry.getName());
705 sequence.transferAnnotation(entry, mp);
707 absStart += entry.getStart();
708 int absEnd = absStart + nonGapped.length() - 1;
709 if (!trimDatasetSeqs)
711 // insert full length sequence from record
712 sequence.setSequence(entry.getSequenceAsString());
713 sequence.setStart(entry.getStart());
717 // finally, update local sequence reference frame if we're allowed
720 // just fix start/end
721 sequence.setStart(absStart);
722 sequence.setEnd(absEnd);
724 // search for alignment sequences to update coordinate frame for
725 for (int alsq = 0; alsq < alseqs.length; alsq++)
727 if (alseqs[alsq].getDatasetSequence() == sequence)
729 String ngAlsq = AlignSeq.extractGaps("-. ",
730 alseqs[alsq].getSequenceAsString()).toUpperCase();
731 int oldstrt = alseqs[alsq].getStart();
732 alseqs[alsq].setStart(sequence.getSequenceAsString()
733 .toUpperCase().indexOf(ngAlsq)
734 + sequence.getStart());
735 if (oldstrt != alseqs[alsq].getStart())
737 alseqs[alsq].setEnd(ngAlsq.length()
738 + alseqs[alsq].getStart() - 1);
742 // TODO: search for all other references to this dataset sequence, and
744 // TODO: update all AlCodonMappings which involve this alignment
745 // sequence (e.g. Q30167 cdna translation from exon2 product (vamsas
748 // and remove it from the rest
749 // TODO: decide if we should remove annotated sequence from set
750 sdataset.remove(sequence);
751 // TODO: should we make a note of sequences that have received new DB
752 // ids, so we can query all enabled DAS servers for them ?
757 // report the ID/sequence mismatches
758 sbuffer.append(messages);
763 * loop thru and collect additional sequences in Map.
765 * @param sequencesArray
768 private SequenceI[] recoverDbSequences(SequenceI[] sequencesArray)
770 Vector nseq = new Vector();
771 for (int i = 0; sequencesArray != null && i < sequencesArray.length; i++)
773 nseq.addElement(sequencesArray[i]);
774 DBRefEntry dbr[] = sequencesArray[i].getDBRefs();
775 jalview.datamodel.Mapping map = null;
776 for (int r = 0; (dbr != null) && r < dbr.length; r++)
778 if ((map = dbr[r].getMap()) != null)
780 if (map.getTo() != null && !nseq.contains(map.getTo()))
782 nseq.addElement(map.getTo());
789 sequencesArray = new SequenceI[nseq.size()];
790 nseq.toArray(sequencesArray);
792 return sequencesArray;