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.DasSourceBrowser;
33 import jalview.gui.Desktop;
34 import jalview.gui.FeatureSettings;
35 import jalview.gui.IProgressIndicator;
36 import jalview.gui.OOMWarning;
37 import jalview.util.DBRefUtils;
38 import jalview.util.MessageManager;
39 import jalview.ws.dbsources.das.api.jalviewSourceI;
40 import jalview.ws.dbsources.das.datamodel.DasSequenceSource;
41 import jalview.ws.seqfetcher.DbSourceProxy;
43 import java.util.ArrayList;
44 import java.util.Arrays;
45 import java.util.Enumeration;
46 import java.util.Hashtable;
47 import java.util.List;
48 import java.util.StringTokenizer;
49 import java.util.Vector;
51 import uk.ac.ebi.picr.model.UPEntry;
52 import uk.ac.ebi.www.picr.AccessionMappingService.AccessionMapperServiceLocator;
55 * Implements a runnable for validating a sequence against external databases
56 * and then propagating references and features onto the sequence(s)
61 public class DBRefFetcher implements Runnable
63 private static final String NEWLINE = System.lineSeparator();
65 public interface FetchFinishedListenerI
72 IProgressIndicator progressWindow;
74 CutAndPasteTransfer output = new CutAndPasteTransfer();
76 boolean running = false;
79 * picr client instance
81 uk.ac.ebi.www.picr.AccessionMappingService.AccessionMapperInterface picrClient = null;
83 // This will be a collection of Vectors of sequenceI refs.
84 // The key will be the seq name or accession id of the seq
85 Hashtable<String, Vector<SequenceI>> seqRefs;
87 DbSourceProxy[] dbSources;
89 SequenceFetcher sfetcher;
91 private List<FetchFinishedListenerI> listeners;
93 private SequenceI[] alseqs;
96 * when true - retrieved sequences will be trimmed to cover longest derived
99 private boolean trimDsSeqs = true;
102 * Creates a new DBRefFetcher object and fetches from the currently selected
103 * set of databases, if this is null then it fetches based on feature settings
106 * fetch references for these SequenceI array
107 * @param progressIndicatorFrame
108 * the frame for progress bar monitoring
110 * array of DbSourceProxy to query references form
111 * @param featureSettings
112 * FeatureSettings to get alternative DbSourceProxy from
113 * @param isNucleotide
114 * indicates if the array of SequenceI are Nucleotides or not
116 public DBRefFetcher(SequenceI[] seqs,
117 IProgressIndicator progressIndicatorFrame,
118 DbSourceProxy[] sources, FeatureSettings featureSettings,
119 boolean isNucleotide)
121 listeners = new ArrayList<FetchFinishedListenerI>();
122 this.progressWindow = progressIndicatorFrame;
123 alseqs = new SequenceI[seqs.length];
124 SequenceI[] ds = new SequenceI[seqs.length];
125 for (int i = 0; i < seqs.length; i++)
128 if (seqs[i].getDatasetSequence() != null)
130 ds[i] = seqs[i].getDatasetSequence();
138 // TODO Jalview 2.5 lots of this code should be in the gui package!
139 sfetcher = jalview.gui.SequenceFetcher
140 .getSequenceFetcherSingleton(progressIndicatorFrame);
141 // set default behaviour for transferring excess sequence data to the
143 trimDsSeqs = Cache.getDefault("TRIM_FETCHED_DATASET_SEQS", true);
146 setDatabaseSources(featureSettings, isNucleotide);
150 // we assume the caller knows what they're doing and ensured that all the
151 // db source names are valid
157 * Helper method to configure the list of database sources to query
159 * @param featureSettings
160 * @param forNucleotide
162 void setDatabaseSources(FeatureSettings featureSettings,
163 boolean forNucleotide)
165 // af.featureSettings_actionPerformed(null);
166 String[] defdb = null;
167 List<DbSourceProxy> selsources = new ArrayList<DbSourceProxy>();
168 Vector<jalviewSourceI> dasselsrc = (featureSettings != null)
169 ? featureSettings.getSelectedSources()
170 : new DasSourceBrowser().getSelectedSources();
172 for (jalviewSourceI src : dasselsrc)
174 List<DbSourceProxy> sp = src.getSequenceSourceProxies();
177 selsources.addAll(sp);
180 Cache.log.debug("Added many Db Sources for :" + src.getTitle());
184 // select appropriate databases based on alignFrame context.
187 defdb = DBRefSource.DNACODINGDBS;
191 defdb = DBRefSource.PROTEINDBS;
193 List<DbSourceProxy> srces = new ArrayList<DbSourceProxy>();
194 for (String ddb : defdb)
196 List<DbSourceProxy> srcesfordb = sfetcher.getSourceProxy(ddb);
197 if (srcesfordb != null)
199 for (DbSourceProxy src : srcesfordb)
201 if (!srces.contains(src))
203 srces.addAll(srcesfordb);
208 // append the PDB data source, since it is 'special', catering for both
209 // nucleotide and protein
210 // srces.addAll(sfetcher.getSourceProxy(DBRefSource.PDB));
212 srces.addAll(selsources);
213 dbSources = srces.toArray(new DbSourceProxy[srces.size()]);
217 * Constructor with only sequences provided
221 public DBRefFetcher(SequenceI[] sequences)
223 this(sequences, null, null, null, false);
227 * Add a listener to be notified when sequence fetching is complete
231 public void addListener(FetchFinishedListenerI l)
237 * retrieve all the das sequence sources and add them to the list of db
238 * sources to retrieve from
240 public void appendAllDasSources()
242 if (dbSources == null)
244 dbSources = new DbSourceProxy[0];
246 // append additional sources
247 DbSourceProxy[] otherdb = sfetcher
248 .getDbSourceProxyInstances(DasSequenceSource.class);
249 if (otherdb != null && otherdb.length > 0)
251 DbSourceProxy[] newsrc = new DbSourceProxy[dbSources.length
253 System.arraycopy(dbSources, 0, newsrc, 0, dbSources.length);
254 System.arraycopy(otherdb, 0, newsrc, dbSources.length,
261 * start the fetcher thread
263 * @param waitTillFinished
264 * true to block until the fetcher has finished
266 public void fetchDBRefs(boolean waitTillFinished)
268 // TODO can we not simply write
269 // if (waitTillFinished) { run(); } else { new Thread(this).start(); }
271 Thread thread = new Thread(this);
275 if (waitTillFinished)
282 } catch (Exception ex)
290 * The sequence will be added to a vector of sequences belonging to key which
291 * could be either seq name or dbref id
298 void addSeqId(SequenceI seq, String key)
300 key = key.toUpperCase();
302 Vector<SequenceI> seqs;
303 if (seqRefs.containsKey(key))
305 seqs = seqRefs.get(key);
307 if (seqs != null && !seqs.contains(seq))
309 seqs.addElement(seq);
311 else if (seqs == null)
313 seqs = new Vector<SequenceI>();
314 seqs.addElement(seq);
320 seqs = new Vector<SequenceI>();
321 seqs.addElement(seq);
324 seqRefs.put(key, seqs);
333 if (dbSources == null)
335 throw new Error(MessageManager
336 .getString("error.implementation_error_must_init_dbsources"));
339 long startTime = System.currentTimeMillis();
340 if (progressWindow != null)
342 progressWindow.setProgressBar(
343 MessageManager.getString("status.fetching_db_refs"),
348 if (Cache.getDefault("DBREFFETCH_USEPICR", false))
350 picrClient = new AccessionMapperServiceLocator()
351 .getAccessionMapperPort();
353 } catch (Exception e)
355 System.err.println("Couldn't locate PICR service instance.\n");
359 Vector<SequenceI> sdataset = new Vector<SequenceI>(
360 Arrays.asList(dataset));
361 List<String> warningMessages = new ArrayList<String>();
364 while (sdataset.size() > 0 && db < dbSources.length)
366 int maxqlen = 1; // default number of queries made at one time
367 System.out.println("Verifying against " + dbSources[db].getDbName());
369 // iterate through db for each remaining un-verified sequence
370 SequenceI[] currSeqs = new SequenceI[sdataset.size()];
371 sdataset.copyInto(currSeqs);// seqs that are to be validated against
373 Vector<String> queries = new Vector<String>(); // generated queries curSeq
374 seqRefs = new Hashtable<String, Vector<SequenceI>>();
378 DbSourceProxy dbsource = dbSources[db];
379 // for moment, we dumbly iterate over all retrieval sources for a
380 // particular database
381 // TODO: introduce multithread multisource queries and logic to remove a
382 // query from other sources if any source for a database returns a
384 maxqlen = dbsource.getMaximumQueryCount();
386 while (queries.size() > 0 || seqIndex < currSeqs.length)
388 if (queries.size() > 0)
390 // Still queries to make for current seqIndex
391 StringBuffer queryString = new StringBuffer("");
393 int nqSize = (maxqlen > queries.size()) ? queries.size()
396 while (queries.size() > 0 && numq < nqSize)
398 String query = queries.elementAt(0);
399 if (dbsource.isValidReference(query))
402 (numq == 0) ? "" : dbsource.getAccessionSeparator());
403 queryString.append(query);
406 // remove the extracted query string
407 queries.removeElementAt(0);
409 // make the queries and process the response
410 AlignmentI retrieved = null;
413 if (Cache.log.isDebugEnabled())
415 Cache.log.debug("Querying " + dbsource.getDbName()
416 + " with : '" + queryString.toString() + "'");
418 retrieved = dbsource.getSequenceRecords(queryString.toString());
419 } catch (Exception ex)
421 ex.printStackTrace();
422 } catch (OutOfMemoryError err)
424 new OOMWarning("retrieving database references ("
425 + queryString.toString() + ")", err);
427 if (retrieved != null)
429 transferReferences(sdataset, dbsource.getDbSource(), retrieved,
430 trimDsSeqs, warningMessages);
435 // make some more strings for use as queries
436 for (int i = 0; (seqIndex < dataset.length)
437 && (i < 50); seqIndex++, i++)
439 SequenceI sequence = dataset[seqIndex];
440 DBRefEntry[] uprefs = DBRefUtils
441 .selectRefs(sequence.getDBRefs(), new String[]
442 { dbsource.getDbSource() }); // jalview.datamodel.DBRefSource.UNIPROT
444 // check for existing dbrefs to use
445 if (uprefs != null && uprefs.length > 0)
447 for (int j = 0; j < uprefs.length; j++)
449 addSeqId(sequence, uprefs[j].getAccessionId());
451 uprefs[j].getAccessionId().toUpperCase());
456 // generate queries from sequence ID string
457 StringTokenizer st = new StringTokenizer(sequence.getName(),
459 while (st.hasMoreTokens())
461 String token = st.nextToken();
462 UPEntry[] presp = null;
463 if (picrClient != null)
465 // resolve the string against PICR to recover valid IDs
468 presp = picrClient.getUPIForAccession(token, null,
469 picrClient.getMappedDatabaseNames(), null,
471 } catch (Exception e)
474 "Exception with Picr for '" + token + "'\n");
478 if (presp != null && presp.length > 0)
480 for (int id = 0; id < presp.length; id++)
482 // construct sequences from response if sequences are
483 // present, and do a transferReferences
484 // otherwise transfer non sequence x-references directly.
487 "Validated ID against PICR... (for what its worth):"
489 addSeqId(sequence, token);
490 queries.addElement(token.toUpperCase());
495 // System.out.println("Not querying source with
496 // token="+token+"\n");
497 addSeqId(sequence, token);
498 queries.addElement(token.toUpperCase());
505 // advance to next database
507 } // all databases have been queried
508 if (!warningMessages.isEmpty())
510 StringBuilder sb = new StringBuilder(warningMessages.size() * 30);
511 sb.append(MessageManager
512 .getString("label.your_sequences_have_been_verified"));
513 for (String msg : warningMessages)
515 sb.append(msg).append(NEWLINE);
517 output.setText(sb.toString());
519 Desktop.addInternalFrame(output,
520 MessageManager.getString("label.sequences_updated"), 600,
522 // The above is the dataset, we must now find out the index
523 // of the viewed sequence
526 if (progressWindow != null)
528 progressWindow.setProgressBar(
529 MessageManager.getString("label.dbref_search_completed"),
533 for (FetchFinishedListenerI listener : listeners)
541 * Verify local sequences in seqRefs against the retrieved sequence database
542 * records. Returns true if any sequence was modified as a result (start/end
543 * changed and/or sequence enlarged), else false.
546 * dataset sequences we are retrieving for
548 * database source we are retrieving from
550 * retrieved sequences as alignment
551 * @param trimDatasetSeqs
552 * if true, sequences will not be enlarged to match longer retrieved
553 * sequences, only their start/end adjusted
554 * @param warningMessages
555 * a list of messages to add to
557 boolean transferReferences(Vector<SequenceI> sdataset, String dbSource,
558 AlignmentI retrievedAl, boolean trimDatasetSeqs,
559 List<String> warningMessages)
561 // System.out.println("trimming ? " + trimDatasetSeqs);
562 if (retrievedAl == null || retrievedAl.getHeight() == 0)
567 boolean modified = false;
568 SequenceI[] retrieved = recoverDbSequences(
569 retrievedAl.getSequencesArray());
570 SequenceI sequence = null;
572 for (SequenceI retrievedSeq : retrieved)
574 // Work out which sequences this sequence matches,
575 // taking into account all accessionIds and names in the file
576 Vector<SequenceI> sequenceMatches = new Vector<SequenceI>();
577 // look for corresponding accession ids
578 DBRefEntry[] entryRefs = DBRefUtils
579 .selectRefs(retrievedSeq.getDBRefs(), new String[]
581 if (entryRefs == null)
584 .println("Dud dbSource string ? no entryrefs selected for "
585 + dbSource + " on " + retrievedSeq.getName());
588 for (int j = 0; j < entryRefs.length; j++)
590 String accessionId = entryRefs[j].getAccessionId();
591 // match up on accessionId
592 if (seqRefs.containsKey(accessionId.toUpperCase()))
594 Vector<SequenceI> seqs = seqRefs.get(accessionId);
595 for (int jj = 0; jj < seqs.size(); jj++)
597 sequence = seqs.elementAt(jj);
598 if (!sequenceMatches.contains(sequence))
600 sequenceMatches.addElement(sequence);
605 if (sequenceMatches.isEmpty())
607 // failed to match directly on accessionId==query so just compare all
608 // sequences to entry
609 Enumeration<String> e = seqRefs.keys();
610 while (e.hasMoreElements())
612 Vector<SequenceI> sqs = seqRefs.get(e.nextElement());
613 if (sqs != null && sqs.size() > 0)
615 Enumeration<SequenceI> sqe = sqs.elements();
616 while (sqe.hasMoreElements())
618 sequenceMatches.addElement(sqe.nextElement());
623 // look for corresponding names
624 // this is uniprot specific ?
625 // could be useful to extend this so we try to find any 'significant'
626 // information in common between two sequence objects.
628 * DBRefEntry[] entryRefs =
629 * jalview.util.DBRefUtils.selectRefs(entry.getDBRef(), new String[] {
630 * dbSource }); for (int j = 0; j < entry.getName().size(); j++) { String
631 * name = entry.getName().elementAt(j).toString(); if
632 * (seqRefs.containsKey(name)) { Vector seqs = (Vector) seqRefs.get(name);
633 * for (int jj = 0; jj < seqs.size(); jj++) { sequence = (SequenceI)
634 * seqs.elementAt(jj); if (!sequenceMatches.contains(sequence)) {
635 * sequenceMatches.addElement(sequence); } } } }
637 // sequenceMatches now contains the set of all sequences associated with
638 // the returned db record
639 final String retrievedSeqString = retrievedSeq.getSequenceAsString();
640 String entrySeq = retrievedSeqString.toUpperCase();
641 for (int m = 0; m < sequenceMatches.size(); m++)
643 sequence = sequenceMatches.elementAt(m);
644 // only update start and end positions and shift features if there are
645 // no existing references
646 // TODO: test for legacy where uniprot or EMBL refs exist but no
647 // mappings are made (but content matches retrieved set)
648 boolean updateRefFrame = sequence.getDBRefs() == null
649 || sequence.getDBRefs().length == 0;
651 // verify sequence against the entry sequence
654 final int sequenceStart = sequence.getStart();
656 boolean remoteEnclosesLocal = false;
657 String nonGapped = AlignSeq
658 .extractGaps("-. ", sequence.getSequenceAsString())
660 int absStart = entrySeq.indexOf(nonGapped);
663 // couldn't find local sequence in sequence from database, so check if
664 // the database sequence is a subsequence of local sequence
665 absStart = nonGapped.indexOf(entrySeq);
668 // verification failed. couldn't find any relationship between
669 // entrySeq and local sequence
670 // messages suppressed as many-to-many matches are confusing
671 // String msg = sequence.getName()
672 // + " Sequence not 100% match with "
673 // + retrievedSeq.getName();
674 // addWarningMessage(warningMessages, msg);
678 * retrieved sequence is a proper subsequence of local sequence
680 String msg = sequence.getName() + " has " + absStart
681 + " prefixed residues compared to "
682 + retrievedSeq.getName();
683 addWarningMessage(warningMessages, msg);
686 * So create a mapping to the external entry from the matching region of
687 * the local sequence, and leave local start/end untouched.
689 mp = new Mapping(null,
691 { sequenceStart + absStart,
692 sequenceStart + absStart + entrySeq.length() - 1 },
694 { retrievedSeq.getStart(),
695 retrievedSeq.getStart() + entrySeq.length() - 1 },
697 updateRefFrame = false;
702 * local sequence is a subsequence of (or matches) retrieved sequence
704 remoteEnclosesLocal = true;
709 SequenceFeature[] sfs = sequence.getSequenceFeatures();
713 * relocate existing sequence features by offset
715 int start = sequenceStart;
716 int end = sequence.getEnd();
717 int startShift = 1 - absStart - start;
721 for (SequenceFeature sf : sfs)
723 if (sf.getBegin() >= start && sf.getEnd() <= end)
725 sf.setBegin(sf.getBegin() + startShift);
726 sf.setEnd(sf.getEnd() + startShift);
735 System.out.println("Adding dbrefs to " + sequence.getName()
736 + " from " + dbSource + " sequence : "
737 + retrievedSeq.getName());
738 sequence.transferAnnotation(retrievedSeq, mp);
740 absStart += retrievedSeq.getStart();
741 int absEnd = absStart + nonGapped.length() - 1;
742 if (!trimDatasetSeqs)
745 * update start position and/or expand to longer retrieved sequence
747 if (!retrievedSeqString.equals(sequence.getSequenceAsString())
748 && remoteEnclosesLocal)
750 sequence.setSequence(retrievedSeqString);
752 addWarningMessage(warningMessages,
753 "Sequence for " + sequence.getName() + " expanded from "
754 + retrievedSeq.getName());
756 if (sequence.getStart() != retrievedSeq.getStart())
758 sequence.setStart(retrievedSeq.getStart());
760 if (absStart != sequenceStart)
762 addWarningMessage(warningMessages,
763 "Start/end position for " + sequence.getName()
764 + " updated from " + retrievedSeq.getName());
770 // finally, update local sequence reference frame if we're allowed
773 // just fix start/end
774 if (sequence.getStart() != absStart
775 || sequence.getEnd() != absEnd)
777 sequence.setStart(absStart);
778 sequence.setEnd(absEnd);
780 addWarningMessage(warningMessages,
781 "Start/end for " + sequence.getName()
782 + " updated from " + retrievedSeq.getName());
785 // search for alignment sequences to update coordinate frame for
786 for (int alsq = 0; alsq < alseqs.length; alsq++)
788 if (alseqs[alsq].getDatasetSequence() == sequence)
790 String ngAlsq = AlignSeq
792 alseqs[alsq].getSequenceAsString())
794 int oldstrt = alseqs[alsq].getStart();
795 alseqs[alsq].setStart(sequence.getSequenceAsString()
796 .toUpperCase().indexOf(ngAlsq) + sequence.getStart());
797 if (oldstrt != alseqs[alsq].getStart())
800 ngAlsq.length() + alseqs[alsq].getStart() - 1);
805 // TODO: search for all other references to this dataset sequence, and
807 // TODO: update all AlCodonMappings which involve this alignment
808 // sequence (e.g. Q30167 cdna translation from exon2 product (vamsas
811 // and remove it from the rest
812 // TODO: decide if we should remove annotated sequence from set
813 sdataset.remove(sequence);
814 // TODO: should we make a note of sequences that have received new DB
815 // ids, so we can query all enabled DAS servers for them ?
822 * Adds the message to the list unless it already contains it
827 void addWarningMessage(List<String> messageList, String msg)
829 if (!messageList.contains(msg))
831 messageList.add(msg);
836 * loop thru and collect additional sequences in Map.
838 * @param sequencesArray
841 private SequenceI[] recoverDbSequences(SequenceI[] sequencesArray)
843 Vector<SequenceI> nseq = new Vector<SequenceI>();
844 for (int i = 0; sequencesArray != null
845 && i < sequencesArray.length; i++)
847 nseq.addElement(sequencesArray[i]);
848 DBRefEntry[] dbr = sequencesArray[i].getDBRefs();
850 for (int r = 0; (dbr != null) && r < dbr.length; r++)
852 if ((map = dbr[r].getMap()) != null)
854 if (map.getTo() != null && !nseq.contains(map.getTo()))
856 nseq.addElement(map.getTo());
863 sequencesArray = new SequenceI[nseq.size()];
864 nseq.toArray(sequencesArray);
866 return sequencesArray;