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 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 * Add a listener to be notified when sequence fetching is complete
197 public void addListener(FetchFinishedListenerI l)
203 * retrieve all the das sequence sources and add them to the list of db
204 * sources to retrieve from
206 public void appendAllDasSources()
208 if (dbSources == null)
210 dbSources = new DbSourceProxy[0];
212 // append additional sources
213 DbSourceProxy[] otherdb = sfetcher
214 .getDbSourceProxyInstances(jalview.ws.dbsources.das.datamodel.DasSequenceSource.class);
215 if (otherdb != null && otherdb.length > 0)
217 DbSourceProxy[] newsrc = new DbSourceProxy[dbSources.length
219 System.arraycopy(dbSources, 0, newsrc, 0, dbSources.length);
220 System.arraycopy(otherdb, 0, newsrc, dbSources.length, otherdb.length);
226 * start the fetcher thread
228 * @param waitTillFinished
229 * true to block until the fetcher has finished
231 public void fetchDBRefs(boolean waitTillFinished)
233 Thread thread = new Thread(this);
237 if (waitTillFinished)
244 } catch (Exception ex)
252 * The sequence will be added to a vector of sequences belonging to key which
253 * could be either seq name or dbref id
260 void addSeqId(SequenceI seq, String key)
262 key = key.toUpperCase();
265 if (seqRefs.containsKey(key))
267 seqs = (Vector) seqRefs.get(key);
269 if (seqs != null && !seqs.contains(seq))
271 seqs.addElement(seq);
273 else if (seqs == null)
276 seqs.addElement(seq);
283 seqs.addElement(seq);
286 seqRefs.put(key, seqs);
295 if (dbSources == null)
299 .getString("error.implementation_error_must_init_dbsources"));
302 long startTime = System.currentTimeMillis();
303 if (progressWindow != null)
305 progressWindow.setProgressBar(
306 MessageManager.getString("status.fetching_db_refs"),
311 if (Cache.getDefault("DBREFFETCH_USEPICR", false))
313 picrClient = new uk.ac.ebi.www.picr.AccessionMappingService.AccessionMapperServiceLocator()
314 .getAccessionMapperPort();
316 } catch (Exception e)
318 System.err.println("Couldn't locate PICR service instance.\n");
322 Vector sdataset = new Vector();
323 for (int s = 0; s < dataset.length; s++)
325 sdataset.addElement(dataset[s]);
327 while (sdataset.size() > 0 && db < dbSources.length)
329 int maxqlen = 1; // default number of queries made to at one time
330 System.err.println("Verifying against " + dbSources[db].getDbName());
333 // iterate through db for each remaining un-verified sequence
334 SequenceI[] currSeqs = new SequenceI[sdataset.size()];
335 sdataset.copyInto(currSeqs);// seqs that are to be validated against
337 Vector queries = new Vector(); // generated queries curSeq
338 seqRefs = new Hashtable();
342 jalview.ws.seqfetcher.DbSourceProxy dbsource = dbSources[db];
344 // for moment, we dumbly iterate over all retrieval sources for a
345 // particular database
346 // TODO: introduce multithread multisource queries and logic to remove a
347 // query from other sources if any source for a database returns a
349 maxqlen = dbsource.getMaximumQueryCount();
351 while (queries.size() > 0 || seqIndex < currSeqs.length)
353 if (queries.size() > 0)
355 // Still queries to make for current seqIndex
356 StringBuffer queryString = new StringBuffer("");
357 int numq = 0, nqSize = (maxqlen > queries.size()) ? queries
360 while (queries.size() > 0 && numq < nqSize)
362 String query = (String) queries.elementAt(0);
363 if (dbsource.isValidReference(query))
365 queryString.append((numq == 0) ? "" : dbsource
366 .getAccessionSeparator());
367 queryString.append(query);
370 // remove the extracted query string
371 queries.removeElementAt(0);
373 // make the queries and process the response
374 AlignmentI retrieved = null;
377 if (jalview.bin.Cache.log.isDebugEnabled())
379 jalview.bin.Cache.log.debug("Querying "
380 + dbsource.getDbName() + " with : '"
381 + queryString.toString() + "'");
383 retrieved = dbsource.getSequenceRecords(queryString
385 } catch (Exception ex)
387 ex.printStackTrace();
388 } catch (OutOfMemoryError err)
390 new OOMWarning("retrieving database references ("
391 + queryString.toString() + ")", err);
393 if (retrieved != null)
395 transferReferences(sdataset, dbsource.getDbSource(),
396 retrieved, trimDsSeqs);
401 // make some more strings for use as queries
402 for (int i = 0; (seqIndex < dataset.length) && (i < 50); seqIndex++, i++)
404 SequenceI sequence = dataset[seqIndex];
405 DBRefEntry[] uprefs = jalview.util.DBRefUtils.selectRefs(
406 sequence.getDBRefs(),
407 new String[] { dbsource.getDbSource() }); // jalview.datamodel.DBRefSource.UNIPROT
409 // check for existing dbrefs to use
410 if (uprefs != null && uprefs.length > 0)
412 for (int j = 0; j < uprefs.length; j++)
414 addSeqId(sequence, uprefs[j].getAccessionId());
415 queries.addElement(uprefs[j].getAccessionId()
421 // generate queries from sequence ID string
422 StringTokenizer st = new StringTokenizer(
423 sequence.getName(), "|");
424 while (st.hasMoreTokens())
426 String token = st.nextToken();
427 UPEntry[] presp = null;
428 if (picrClient != null)
430 // resolve the string against PICR to recover valid IDs
433 presp = picrClient.getUPIForAccession(token, null,
434 picrClient.getMappedDatabaseNames(), null,
436 } catch (Exception e)
438 System.err.println("Exception with Picr for '"
443 if (presp != null && presp.length > 0)
445 for (int id = 0; id < presp.length; id++)
447 // construct sequences from response if sequences are
448 // present, and do a transferReferences
449 // otherwise transfer non sequence x-references directly.
452 .println("Validated ID against PICR... (for what its worth):"
454 addSeqId(sequence, token);
455 queries.addElement(token.toUpperCase());
460 // System.out.println("Not querying source with token="+token+"\n");
461 addSeqId(sequence, token);
462 queries.addElement(token.toUpperCase());
470 // advance to next database
472 } // all databases have been queries.
473 if (sbuffer.length() > 0)
475 output.setText(MessageManager
476 .getString("label.your_sequences_have_been_verified")
477 + sbuffer.toString());
478 Desktop.addInternalFrame(output,
479 MessageManager.getString("label.sequence_names_updated"),
481 // The above is the dataset, we must now find out the index
482 // of the viewed sequence
485 if (progressWindow != null)
487 progressWindow.setProgressBar(
488 MessageManager.getString("label.dbref_search_completed"),
492 for (FetchFinishedListenerI listener : listeners)
500 * Verify local sequences in seqRefs against the retrieved sequence database
503 * @param trimDatasetSeqs
506 void transferReferences(Vector sdataset, String dbSource,
507 AlignmentI retrievedAl, boolean trimDatasetSeqs) // File
510 System.out.println("trimming ? " + trimDatasetSeqs);
511 if (retrievedAl == null || retrievedAl.getHeight() == 0)
515 SequenceI[] retrieved = recoverDbSequences(retrievedAl
516 .getSequencesArray());
517 SequenceI sequence = null;
518 boolean transferred = false;
519 StringBuffer messages = new StringBuffer();
521 // Vector entries = new Uniprot().getUniprotEntries(file);
523 int i, iSize = retrieved.length; // entries == null ? 0 : entries.size();
524 // UniprotEntry entry;
525 for (i = 0; i < iSize; i++)
527 SequenceI entry = retrieved[i]; // (UniprotEntry) entries.elementAt(i);
529 // Work out which sequences this sequence matches,
530 // taking into account all accessionIds and names in the file
531 Vector sequenceMatches = new Vector();
532 // look for corresponding accession ids
533 DBRefEntry[] entryRefs = jalview.util.DBRefUtils.selectRefs(
534 entry.getDBRefs(), new String[] { dbSource });
535 if (entryRefs == null)
538 .println("Dud dbSource string ? no entryrefs selected for "
539 + dbSource + " on " + entry.getName());
542 for (int j = 0; j < entryRefs.length; j++)
544 String accessionId = entryRefs[j].getAccessionId(); // .getAccession().elementAt(j).toString();
545 // match up on accessionId
546 if (seqRefs.containsKey(accessionId.toUpperCase()))
548 Vector seqs = (Vector) seqRefs.get(accessionId);
549 for (int jj = 0; jj < seqs.size(); jj++)
551 sequence = (SequenceI) seqs.elementAt(jj);
552 if (!sequenceMatches.contains(sequence))
554 sequenceMatches.addElement(sequence);
559 if (sequenceMatches.size() == 0)
561 // failed to match directly on accessionId==query so just compare all
562 // sequences to entry
563 Enumeration e = seqRefs.keys();
564 while (e.hasMoreElements())
566 Vector sqs = (Vector) seqRefs.get(e.nextElement());
567 if (sqs != null && sqs.size() > 0)
569 Enumeration sqe = sqs.elements();
570 while (sqe.hasMoreElements())
572 sequenceMatches.addElement(sqe.nextElement());
577 // look for corresponding names
578 // this is uniprot specific ?
579 // could be useful to extend this so we try to find any 'significant'
580 // information in common between two sequence objects.
582 * DBRefEntry[] entryRefs =
583 * jalview.util.DBRefUtils.selectRefs(entry.getDBRef(), new String[] {
584 * dbSource }); for (int j = 0; j < entry.getName().size(); j++) { String
585 * name = entry.getName().elementAt(j).toString(); if
586 * (seqRefs.containsKey(name)) { Vector seqs = (Vector) seqRefs.get(name);
587 * for (int jj = 0; jj < seqs.size(); jj++) { sequence = (SequenceI)
588 * seqs.elementAt(jj); if (!sequenceMatches.contains(sequence)) {
589 * sequenceMatches.addElement(sequence); } } } }
591 // sequenceMatches now contains the set of all sequences associated with
592 // the returned db record
593 String entrySeq = entry.getSequenceAsString().toUpperCase();
594 for (int m = 0; m < sequenceMatches.size(); m++)
596 sequence = (SequenceI) sequenceMatches.elementAt(m);
597 // only update start and end positions and shift features if there are
598 // no existing references
599 // TODO: test for legacy where uniprot or EMBL refs exist but no
600 // mappings are made (but content matches retrieved set)
601 boolean updateRefFrame = sequence.getDBRefs() == null
602 || sequence.getDBRefs().length == 0;
604 // verify sequence against the entry sequence
606 String nonGapped = AlignSeq.extractGaps("-. ",
607 sequence.getSequenceAsString()).toUpperCase();
609 int absStart = entrySeq.indexOf(nonGapped);
612 final int sequenceStart = sequence.getStart();
615 // Is local sequence contained in dataset sequence?
616 absStart = nonGapped.indexOf(entrySeq);
618 { // verification failed.
619 messages.append(sequence.getName()
620 + " SEQUENCE NOT %100 MATCH \n");
624 sbuffer.append(sequence.getName() + " HAS " + absStart
625 + " PREFIXED RESIDUES COMPARED TO " + dbSource + "\n");
627 // + " - ANY SEQUENCE FEATURES"
628 // + " HAVE BEEN ADJUSTED ACCORDINGLY \n");
630 // create valid mapping between matching region of local sequence and
631 // the mapped sequence
632 mp = new Mapping(null, new int[] { sequenceStart + absStart,
633 sequenceStart + absStart + entrySeq.length() - 1 }, new int[]
634 { entry.getStart(), entry.getStart() + entrySeq.length() - 1 },
636 updateRefFrame = false; // mapping is based on current start/end so
637 // don't modify start and end
642 // update start and end of local sequence to place it in entry's
644 // apply identity map map from whole of local sequence to matching
645 // region of database
647 mp = null; // Mapping.getIdentityMap();
649 // new int[] { absStart+sequence.getStart(),
650 // absStart+sequence.getStart()+entrySeq.length()-1},
651 // new int[] { entry.getStart(), entry.getEnd() }, 1, 1);
652 // relocate local features for updated start
655 if (sequence.getSequenceFeatures() != null)
657 SequenceFeature[] sf = sequence.getSequenceFeatures();
658 int start = sequenceStart;
659 int end = sequence.getEnd();
660 int startShift = 1 - absStart - start; // how much the features
663 for (int sfi = 0; sfi < sf.length; sfi++)
665 if (sf[sfi].getBegin() >= start && sf[sfi].getEnd() <= end)
667 // shift feature along by absstart
668 sf[sfi].setBegin(sf[sfi].getBegin() + startShift);
669 sf[sfi].setEnd(sf[sfi].getEnd() + startShift);
676 System.out.println("Adding dbrefs to " + sequence.getName()
677 + " from " + dbSource + " sequence : " + entry.getName());
678 sequence.transferAnnotation(entry, mp);
679 // unknownSequences.remove(sequence);
680 absStart += entry.getStart();
681 int absEnd = absStart + nonGapped.length() - 1;
682 if (!trimDatasetSeqs)
684 // insert full length sequence from record
685 sequence.setSequence(entry.getSequenceAsString());
686 sequence.setStart(entry.getStart());
690 // finally, update local sequence reference frame if we're allowed
693 // just fix start/end
694 sequence.setStart(absStart);
695 sequence.setEnd(absEnd);
697 // search for alignment sequences to update coordinate frame for
698 for (int alsq = 0; alsq < alseqs.length; alsq++)
700 if (alseqs[alsq].getDatasetSequence() == sequence)
702 String ngAlsq = AlignSeq.extractGaps("-. ",
703 alseqs[alsq].getSequenceAsString()).toUpperCase();
704 int oldstrt = alseqs[alsq].getStart();
705 alseqs[alsq].setStart(sequence.getSequenceAsString()
706 .toUpperCase().indexOf(ngAlsq)
707 + sequence.getStart());
708 if (oldstrt != alseqs[alsq].getStart())
710 alseqs[alsq].setEnd(ngAlsq.length()
711 + alseqs[alsq].getStart() - 1);
715 // TODO: search for all other references to this dataset sequence, and
717 // TODO: update all AlCodonMappings which involve this alignment
718 // sequence (e.g. Q30167 cdna translation from exon2 product (vamsas
721 // and remove it from the rest
722 // TODO: decide if we should remove annotated sequence from set
723 sdataset.remove(sequence);
724 // TODO: should we make a note of sequences that have received new DB
725 // ids, so we can query all enabled DAS servers for them ?
730 // report the ID/sequence mismatches
731 sbuffer.append(messages);
736 * loop thru and collect additional sequences in Map.
738 * @param sequencesArray
741 private SequenceI[] recoverDbSequences(SequenceI[] sequencesArray)
743 Vector nseq = new Vector();
744 for (int i = 0; sequencesArray != null && i < sequencesArray.length; i++)
746 nseq.addElement(sequencesArray[i]);
747 DBRefEntry dbr[] = sequencesArray[i].getDBRefs();
748 jalview.datamodel.Mapping map = null;
749 for (int r = 0; (dbr != null) && r < dbr.length; r++)
751 if ((map = dbr[r].getMap()) != null)
753 if (map.getTo() != null && !nseq.contains(map.getTo()))
755 nseq.addElement(map.getTo());
762 sequencesArray = new SequenceI[nseq.size()];
763 nseq.toArray(sequencesArray);
765 return sequencesArray;