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 java.util.ArrayList;
24 import java.util.Arrays;
25 import java.util.Enumeration;
26 import java.util.HashMap;
27 import java.util.Hashtable;
28 import java.util.List;
29 import java.util.Locale;
31 import java.util.Vector;
32 import java.util.regex.Matcher;
33 import java.util.regex.Pattern;
35 import jalview.analysis.AlignSeq;
36 import jalview.api.FeatureSettingsModelI;
37 import jalview.bin.Cache;
38 import jalview.bin.Console;
39 import jalview.datamodel.AlignmentI;
40 import jalview.datamodel.DBRefEntry;
41 import jalview.datamodel.DBRefSource;
42 import jalview.datamodel.Mapping;
43 import jalview.datamodel.SequenceI;
44 import jalview.gui.CutAndPasteTransfer;
45 import jalview.gui.Desktop;
46 import jalview.gui.FeatureSettings;
47 import jalview.gui.IProgressIndicator;
48 import jalview.gui.OOMWarning;
49 import jalview.util.DBRefUtils;
50 import jalview.util.IdUtils;
51 import jalview.util.IdUtils.IdType;
52 import jalview.util.MessageManager;
53 import jalview.ws.seqfetcher.DbSourceProxy;
54 import uk.ac.ebi.picr.model.UPEntry;
55 import uk.ac.ebi.www.picr.AccessionMappingService.AccessionMapperServiceLocator;
58 * Implements a runnable for validating a sequence against external databases
59 * and then propagating references and features onto the sequence(s)
64 public class DBRefFetcher implements Runnable
66 private static final String NEWLINE = System.lineSeparator();
68 public static final String TRIM_RETRIEVED_SEQUENCES = "TRIM_FETCHED_DATASET_SEQS";
70 public interface FetchFinishedListenerI
77 IProgressIndicator progressWindow;
79 CutAndPasteTransfer output = new CutAndPasteTransfer();
82 * picr client instance
84 uk.ac.ebi.www.picr.AccessionMappingService.AccessionMapperInterface picrClient = null;
86 // This will be a collection of Vectors of sequenceI refs.
87 // The key will be the seq name or accession id of the seq
88 Hashtable<String, Vector<SequenceI>> seqRefs;
90 DbSourceProxy[] dbSources;
92 SequenceFetcher sfetcher;
94 private List<FetchFinishedListenerI> listeners;
96 private SequenceI[] alseqs;
99 * when true - retrieved sequences will be trimmed to cover longest derived
102 private boolean trimDsSeqs = true;
105 * Creates a new DBRefFetcher object and fetches from the currently selected
106 * set of databases, if this is null then it fetches based on feature settings
109 * fetch references for these SequenceI array
110 * @param progressIndicatorFrame
111 * the frame for progress bar monitoring
113 * array of DbSourceProxy to query references form
114 * @param featureSettings
115 * FeatureSettings to get alternative DbSourceProxy from
116 * @param isNucleotide
117 * indicates if the array of SequenceI are Nucleotides or not
119 public DBRefFetcher(SequenceI[] seqs,
120 IProgressIndicator progressIndicatorFrame,
121 DbSourceProxy[] sources, FeatureSettings featureSettings,
122 boolean isNucleotide)
124 listeners = new ArrayList<>();
125 this.progressWindow = progressIndicatorFrame;
126 alseqs = new SequenceI[seqs.length];
127 SequenceI[] ds = new SequenceI[seqs.length];
128 for (int i = 0; i < seqs.length; i++)
131 if (seqs[i].getDatasetSequence() != null)
133 ds[i] = seqs[i].getDatasetSequence();
141 // TODO Jalview 2.5 lots of this code should be in the gui package!
142 sfetcher = jalview.gui.SequenceFetcher.getSequenceFetcherSingleton();
143 // set default behaviour for transferring excess sequence data to the
145 trimDsSeqs = Cache.getDefault(TRIM_RETRIEVED_SEQUENCES, true);
148 setDatabaseSources(featureSettings, isNucleotide);
152 // we assume the caller knows what they're doing and ensured that all the
153 // db source names are valid
159 * Helper method to configure the list of database sources to query
161 * @param featureSettings
162 * @param forNucleotide
164 void setDatabaseSources(FeatureSettings featureSettings,
165 boolean forNucleotide)
167 // af.featureSettings_actionPerformed(null);
168 String[] defdb = null;
169 List<DbSourceProxy> selsources = new ArrayList<>();
170 // select appropriate databases based on alignFrame context.
173 defdb = DBRefSource.DNACODINGDBS;
177 defdb = DBRefSource.PROTEINDBS;
179 List<DbSourceProxy> srces = new ArrayList<>();
180 for (String ddb : defdb)
182 List<DbSourceProxy> srcesfordb = sfetcher.getSourceProxy(ddb);
183 if (srcesfordb != null)
185 for (DbSourceProxy src : srcesfordb)
187 if (!srces.contains(src))
189 srces.addAll(srcesfordb);
194 // append the PDB data source, since it is 'special', catering for both
195 // nucleotide and protein
196 // srces.addAll(sfetcher.getSourceProxy(DBRefSource.PDB));
198 srces.addAll(selsources);
199 dbSources = srces.toArray(new DbSourceProxy[srces.size()]);
203 * Constructor with only sequences provided
207 public DBRefFetcher(SequenceI[] sequences)
209 this(sequences, null, null, null, false);
213 * Add a listener to be notified when sequence fetching is complete
217 public void addListener(FetchFinishedListenerI l)
223 * start the fetcher thread
225 * @param waitTillFinished
226 * true to block until the fetcher has finished
228 public void fetchDBRefs(boolean waitTillFinished)
230 if (waitTillFinished)
236 new Thread(this).start();
241 * The sequence will be added to a vector of sequences belonging to key which
242 * could be either seq name or dbref id
249 void addSeqId(SequenceI seq, String key)
251 key = key.toUpperCase(Locale.ROOT);
253 Vector<SequenceI> seqs;
254 if (seqRefs.containsKey(key))
256 seqs = seqRefs.get(key);
258 if (seqs != null && !seqs.contains(seq))
260 seqs.addElement(seq);
262 else if (seqs == null)
264 seqs = new Vector<>();
265 seqs.addElement(seq);
271 seqs = new Vector<>();
272 seqs.addElement(seq);
275 seqRefs.put(key, seqs);
284 if (dbSources == null)
286 throw new Error(MessageManager
287 .getString("error.implementation_error_must_init_dbsources"));
289 long progressId = IdUtils.newId(IdType.PROGRESS);
290 if (progressWindow != null)
292 progressWindow.setProgressBar(
293 MessageManager.getString("status.fetching_db_refs"),
298 if (Cache.getDefault("DBREFFETCH_USEPICR", false))
300 picrClient = new AccessionMapperServiceLocator()
301 .getAccessionMapperPort();
303 } catch (Exception e)
306 .errPrintln("Couldn't locate PICR service instance.\n");
310 Vector<SequenceI> sdataset = new Vector<>(Arrays.asList(dataset));
311 List<String> warningMessages = new ArrayList<>();
313 // clear any old feature display settings recorded from past sessions
314 featureDisplaySettings = null;
317 while (sdataset.size() > 0 && db < dbSources.length)
319 int maxqlen = 1; // default number of queries made at one time
321 .outPrintln("Verifying against " + dbSources[db].getDbName());
323 // iterate through db for each remaining un-verified sequence
324 SequenceI[] currSeqs = new SequenceI[sdataset.size()];
325 sdataset.copyInto(currSeqs);// seqs that are to be validated against
327 Vector<String> queries = new Vector<>(); // generated queries curSeq
328 seqRefs = new Hashtable<>();
332 DbSourceProxy dbsource = dbSources[db];
333 // for moment, we dumbly iterate over all retrieval sources for a
334 // particular database
335 // TODO: introduce multithread multisource queries and logic to remove a
336 // query from other sources if any source for a database returns a
338 maxqlen = dbsource.getMaximumQueryCount();
340 while (queries.size() > 0 || seqIndex < currSeqs.length)
342 if (queries.size() > 0)
344 // Still queries to make for current seqIndex
345 StringBuffer queryString = new StringBuffer("");
347 int nqSize = (maxqlen > queries.size()) ? queries.size()
350 while (queries.size() > 0 && numq < nqSize)
352 String query = queries.elementAt(0);
353 if (dbsource.isValidReference(query))
356 (numq == 0) ? "" : dbsource.getAccessionSeparator());
357 queryString.append(query);
360 // remove the extracted query string
361 queries.removeElementAt(0);
363 // make the queries and process the response
364 AlignmentI retrieved = null;
367 if (Console.isDebugEnabled())
369 Console.debug("Querying " + dbsource.getDbName() + " with : '"
370 + queryString.toString() + "'");
372 retrieved = dbsource.getSequenceRecords(queryString.toString());
373 } catch (Exception ex)
375 ex.printStackTrace();
376 } catch (OutOfMemoryError err)
378 new OOMWarning("retrieving database references ("
379 + queryString.toString() + ")", err);
381 if (retrieved != null)
383 transferReferences(sdataset, dbsource, retrieved, trimDsSeqs,
389 // make some more strings for use as queries
390 for (int i = 0; (seqIndex < dataset.length)
391 && (i < 50); seqIndex++, i++)
393 SequenceI sequence = dataset[seqIndex];
394 List<DBRefEntry> uprefs = DBRefUtils
395 .selectRefs(sequence.getDBRefs(), new String[]
396 { dbsource.getDbSource() }); // jalview.datamodel.DBRefSource.UNIPROT
398 // check for existing dbrefs to use
399 if (uprefs != null && uprefs.size() > 0)
401 for (int j = 0, n = uprefs.size(); j < n; j++)
403 DBRefEntry upref = uprefs.get(j);
404 addSeqId(sequence, upref.getAccessionId());
406 upref.getAccessionId().toUpperCase(Locale.ROOT));
411 Pattern possibleIds = Pattern.compile("[A-Za-z0-9_]+");
412 // generate queries from sequence ID string
413 Matcher tokens = possibleIds.matcher(sequence.getName());
415 while (tokens.find(p))
417 String token = tokens.group();
419 UPEntry[] presp = null;
420 if (picrClient != null)
422 // resolve the string against PICR to recover valid IDs
425 presp = picrClient.getUPIForAccession(token, null,
426 picrClient.getMappedDatabaseNames(), null,
428 } catch (Exception e)
430 jalview.bin.Console.errPrintln(
431 "Exception with Picr for '" + token + "'\n");
435 if (presp != null && presp.length > 0)
437 for (int id = 0; id < presp.length; id++)
439 // construct sequences from response if sequences are
440 // present, and do a transferReferences
441 // otherwise transfer non sequence x-references directly.
443 jalview.bin.Console.outPrintln(
444 "Validated ID against PICR... (for what its worth):"
446 addSeqId(sequence, token);
447 queries.addElement(token.toUpperCase(Locale.ROOT));
452 // jalview.bin.Console.outPrintln("Not querying source with
453 // token="+token+"\n");
454 addSeqId(sequence, token);
455 queries.addElement(token.toUpperCase(Locale.ROOT));
462 // advance to next database
464 } // all databases have been queried
465 if (!warningMessages.isEmpty())
467 StringBuilder sb = new StringBuilder(warningMessages.size() * 30);
468 sb.append(MessageManager
469 .getString("label.your_sequences_have_been_verified"));
470 for (String msg : warningMessages)
472 sb.append(msg).append(NEWLINE);
474 output.setText(sb.toString());
476 Desktop.addInternalFrame(output,
477 MessageManager.getString("label.sequences_updated"), 600,
479 // The above is the dataset, we must now find out the index
480 // of the viewed sequence
483 if (progressWindow != null)
485 progressWindow.setProgressBar(
486 MessageManager.getString("label.dbref_search_completed"),
490 for (FetchFinishedListenerI listener : listeners)
497 * Verify local sequences in seqRefs against the retrieved sequence database
498 * records. Returns true if any sequence was modified as a result (start/end
499 * changed and/or sequence enlarged), else false.
502 * dataset sequences we are retrieving for
504 * database source we are retrieving from
506 * retrieved sequences as alignment
507 * @param trimDatasetSeqs
508 * if true, sequences will not be enlarged to match longer retrieved
509 * sequences, only their start/end adjusted
510 * @param warningMessages
511 * a list of messages to add to
513 boolean transferReferences(Vector<SequenceI> sdataset,
514 DbSourceProxy dbSourceProxy, AlignmentI retrievedAl,
515 boolean trimDatasetSeqs, List<String> warningMessages)
517 // jalview.bin.Console.outPrintln("trimming ? " + trimDatasetSeqs);
518 if (retrievedAl == null || retrievedAl.getHeight() == 0)
523 String dbSource = dbSourceProxy.getDbName();
524 boolean modified = false;
525 SequenceI[] retrieved = recoverDbSequences(
526 retrievedAl.getSequencesArray());
527 SequenceI sequence = null;
529 for (SequenceI retrievedSeq : retrieved)
531 // Work out which sequences this sequence matches,
532 // taking into account all accessionIds and names in the file
533 Vector<SequenceI> sequenceMatches = new Vector<>();
534 // look for corresponding accession ids
535 List<DBRefEntry> entryRefs = DBRefUtils
536 .selectRefs(retrievedSeq.getDBRefs(), new String[]
538 if (entryRefs == null)
541 .println("Dud dbSource string ? no entryrefs selected for "
542 + dbSource + " on " + retrievedSeq.getName());
545 for (int j = 0, n = entryRefs.size(); j < n; j++)
547 DBRefEntry ref = entryRefs.get(j);
548 String accessionId = ref.getAccessionId();
549 // match up on accessionId
550 if (seqRefs.containsKey(accessionId.toUpperCase(Locale.ROOT)))
552 Vector<SequenceI> seqs = seqRefs.get(accessionId);
553 for (int jj = 0; jj < seqs.size(); jj++)
555 sequence = seqs.elementAt(jj);
556 if (!sequenceMatches.contains(sequence))
558 sequenceMatches.addElement(sequence);
563 if (sequenceMatches.isEmpty())
565 // failed to match directly on accessionId==query so just compare all
566 // sequences to entry
567 Enumeration<String> e = seqRefs.keys();
568 while (e.hasMoreElements())
570 Vector<SequenceI> sqs = seqRefs.get(e.nextElement());
571 if (sqs != null && sqs.size() > 0)
573 Enumeration<SequenceI> sqe = sqs.elements();
574 while (sqe.hasMoreElements())
576 sequenceMatches.addElement(sqe.nextElement());
581 // look for corresponding names
582 // this is uniprot specific ?
583 // could be useful to extend this so we try to find any 'significant'
584 // information in common between two sequence objects.
586 * List<DBRefEntry> entryRefs =
587 * jalview.util.DBRefUtils.selectRefs(entry.getDBRef(), new String[] {
588 * dbSource }); for (int j = 0; j < entry.getName().size(); j++) { String
589 * name = entry.getName().elementAt(j).toString(); if
590 * (seqRefs.containsKey(name)) { Vector seqs = (Vector) seqRefs.get(name);
591 * for (int jj = 0; jj < seqs.size(); jj++) { sequence = (SequenceI)
592 * seqs.elementAt(jj); if (!sequenceMatches.contains(sequence)) {
593 * sequenceMatches.addElement(sequence); } } } }
595 if (sequenceMatches.size() > 0)
597 addFeatureSettings(dbSourceProxy);
599 // sequenceMatches now contains the set of all sequences associated with
600 // the returned db record
601 final String retrievedSeqString = retrievedSeq.getSequenceAsString();
602 String entrySeq = retrievedSeqString.toUpperCase(Locale.ROOT);
603 for (int m = 0; m < sequenceMatches.size(); m++)
605 sequence = sequenceMatches.elementAt(m);
606 // only update start and end positions and shift features if there are
607 // no existing references
608 // TODO: test for legacy where uniprot or EMBL refs exist but no
609 // mappings are made (but content matches retrieved set)
610 boolean updateRefFrame = sequence.getDBRefs() == null
611 || sequence.getDBRefs().size() == 0;
613 // verify sequence against the entry sequence
616 final int sequenceStart = sequence.getStart();
618 boolean remoteEnclosesLocal = false;
619 String nonGapped = AlignSeq
620 .extractGaps("-. ", sequence.getSequenceAsString())
621 .toUpperCase(Locale.ROOT);
622 int absStart = entrySeq.indexOf(nonGapped);
625 // couldn't find local sequence in sequence from database, so check if
626 // the database sequence is a subsequence of local sequence
627 absStart = nonGapped.indexOf(entrySeq);
630 // verification failed. couldn't find any relationship between
631 // entrySeq and local sequence
632 // messages suppressed as many-to-many matches are confusing
633 // String msg = sequence.getName()
634 // + " Sequence not 100% match with "
635 // + retrievedSeq.getName();
636 // addWarningMessage(warningMessages, msg);
640 * retrieved sequence is a proper subsequence of local sequence
642 String msg = sequence.getName() + " has " + absStart
643 + " prefixed residues compared to "
644 + retrievedSeq.getName();
645 addWarningMessage(warningMessages, msg);
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,
653 { sequenceStart + absStart,
654 sequenceStart + absStart + entrySeq.length() - 1 },
656 { retrievedSeq.getStart(),
657 retrievedSeq.getStart() + entrySeq.length() - 1 },
659 updateRefFrame = false;
664 * local sequence is a subsequence of (or matches) retrieved sequence
666 remoteEnclosesLocal = true;
672 * relocate existing sequence features by offset
674 int startShift = absStart - sequenceStart + 1;
677 modified |= sequence.getFeatures().shiftFeatures(1,
683 jalview.bin.Console.outPrintln("Adding dbrefs to "
684 + sequence.getName() + " from " + dbSource + " sequence : "
685 + retrievedSeq.getName());
686 sequence.transferAnnotation(retrievedSeq, mp);
688 absStart += retrievedSeq.getStart();
689 int absEnd = absStart + nonGapped.length() - 1;
690 if (!trimDatasetSeqs)
693 * update start position and/or expand to longer retrieved sequence
695 if (!retrievedSeqString.equals(sequence.getSequenceAsString())
696 && remoteEnclosesLocal)
698 sequence.setSequence(retrievedSeqString);
700 addWarningMessage(warningMessages,
701 "Sequence for " + sequence.getName() + " expanded from "
702 + retrievedSeq.getName());
704 if (sequence.getStart() != retrievedSeq.getStart())
706 sequence.setStart(retrievedSeq.getStart());
708 if (absStart != sequenceStart)
710 addWarningMessage(warningMessages,
711 "Start/end position for " + sequence.getName()
712 + " updated from " + retrievedSeq.getName());
718 // finally, update local sequence reference frame if we're allowed
721 // just fix start/end
722 if (sequence.getStart() != absStart
723 || sequence.getEnd() != absEnd)
725 sequence.setStart(absStart);
726 sequence.setEnd(absEnd);
728 addWarningMessage(warningMessages,
729 "Start/end for " + sequence.getName()
730 + " updated from " + retrievedSeq.getName());
733 // search for alignment sequences to update coordinate frame for
734 for (int alsq = 0; alsq < alseqs.length; alsq++)
736 if (alseqs[alsq].getDatasetSequence() == sequence)
738 String ngAlsq = AlignSeq
740 alseqs[alsq].getSequenceAsString())
741 .toUpperCase(Locale.ROOT);
742 int oldstrt = alseqs[alsq].getStart();
743 alseqs[alsq].setStart(sequence.getSequenceAsString()
744 .toUpperCase(Locale.ROOT).indexOf(ngAlsq)
745 + sequence.getStart());
746 if (oldstrt != alseqs[alsq].getStart())
749 ngAlsq.length() + alseqs[alsq].getStart() - 1);
754 // TODO: search for all other references to this dataset sequence, and
756 // TODO: update all AlCodonMappings which involve this alignment
757 // sequence (e.g. Q30167 cdna translation from exon2 product (vamsas
760 // and remove it from the rest
761 // TODO: decide if we should remove annotated sequence from set
762 sdataset.remove(sequence);
768 Map<String, FeatureSettingsModelI> featureDisplaySettings = null;
770 private void addFeatureSettings(DbSourceProxy dbSourceProxy)
772 FeatureSettingsModelI fsettings = dbSourceProxy
773 .getFeatureColourScheme();
774 if (fsettings != null)
776 if (featureDisplaySettings == null)
778 featureDisplaySettings = new HashMap<>();
780 featureDisplaySettings.put(dbSourceProxy.getDbName(), fsettings);
786 * @return any feature settings associated with sources that have provided
789 public List<FeatureSettingsModelI> getFeatureSettingsModels()
791 return featureDisplaySettings == null
792 ? Arrays.asList(new FeatureSettingsModelI[0])
793 : Arrays.asList(featureDisplaySettings.values()
794 .toArray(new FeatureSettingsModelI[1]));
798 * Adds the message to the list unless it already contains it
803 void addWarningMessage(List<String> messageList, String msg)
805 if (!messageList.contains(msg))
807 messageList.add(msg);
812 * loop thru and collect additional sequences in Map.
814 * @param sequencesArray
817 private SequenceI[] recoverDbSequences(SequenceI[] sequencesArray)
820 if (sequencesArray == null || (n = sequencesArray.length) == 0)
821 return sequencesArray;
822 ArrayList<SequenceI> nseq = new ArrayList<>();
823 for (int i = 0; i < n; i++)
825 nseq.add(sequencesArray[i]);
826 List<DBRefEntry> dbr = sequencesArray[i].getDBRefs();
830 for (int r = 0, rn = dbr.size(); r < rn; r++)
832 if ((map = dbr.get(r).getMap()) != null)
834 if (map.getTo() != null && !nseq.contains(map.getTo()))
836 nseq.add(map.getTo());
842 // BH 2019.01.25 question here if this is the right logic. Return the
843 // original if nothing found?
846 return nseq.toArray(new SequenceI[nseq.size()]);
848 return sequencesArray;