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.
21 package jalview.datamodel;
23 import jalview.analysis.AlignSeq;
24 import jalview.datamodel.features.SequenceFeatures;
25 import jalview.datamodel.features.SequenceFeaturesI;
26 import jalview.util.Comparison;
27 import jalview.util.DBRefUtils;
28 import jalview.util.MapList;
29 import jalview.util.StringUtils;
30 import jalview.workers.InformationThread;
32 import java.util.ArrayList;
33 import java.util.Arrays;
34 import java.util.BitSet;
35 import java.util.Collections;
36 import java.util.Enumeration;
37 import java.util.Iterator;
38 import java.util.List;
39 import java.util.ListIterator;
40 import java.util.Vector;
42 import fr.orsay.lri.varna.models.rna.RNA;
46 * Implements the SequenceI interface for a char[] based sequence object
48 public class Sequence extends ASequence implements SequenceI
52 * A subclass that gives us access to modCount, which tracks whether there
53 * have been any changes. We use this to update
59 @SuppressWarnings("serial")
60 public class DBModList<T> extends ArrayList<DBRefEntry>
63 protected int getModCount()
70 SequenceI datasetSequence;
74 private char[] sequence;
76 private String description;
82 private Vector<PDBEntry> pdbIds;
84 private String vamsasId;
86 HiddenMarkovModel hmm;
88 boolean isHMMConsensusSequence = false;
89 private DBModList<DBRefEntry> dbrefs; // controlled access
92 * a flag to let us know that elements have changed in dbrefs
96 private int refModCount = 0;
101 * This annotation is displayed below the alignment but the positions are tied
102 * to the residues of this sequence
104 * TODO: change to List<>
106 private Vector<AlignmentAnnotation> annotation;
108 private SequenceFeaturesI sequenceFeatureStore;
111 * A cursor holding the approximate current view position to the sequence,
112 * as determined by findIndex or findPosition or findPositions.
113 * Using a cursor as a hint allows these methods to be more performant for
116 private SequenceCursor cursor;
119 * A number that should be incremented whenever the sequence is edited.
120 * If the value matches the cursor token, then we can trust the cursor,
121 * if not then it should be recomputed.
123 private int changeCount;
126 * Creates a new Sequence object.
129 * display name string
131 * string to form a possibly gapped sequence out of
133 * first position of non-gap residue in the sequence
135 * last position of ungapped residues (nearly always only used for
138 public Sequence(String name, String sequence, int start, int end)
141 initSeqAndName(name, sequence.toCharArray(), start, end);
144 public Sequence(String name, char[] sequence, int start, int end)
147 initSeqAndName(name, sequence, start, end);
151 * Stage 1 constructor - assign name, sequence, and set start and end fields.
152 * start and end are updated values from name2 if it ends with /start-end
159 protected void initSeqAndName(String name2, char[] sequence2, int start2,
163 this.sequence = sequence2;
171 * If 'name' ends in /i-j, where i >= j > 0 are integers, extracts i and j as
172 * start and end respectively and removes the suffix from the name
179 "POSSIBLE IMPLEMENTATION ERROR: null sequence name passed to constructor.");
182 int slashPos = name.lastIndexOf('/');
183 if (slashPos > -1 && slashPos < name.length() - 1)
185 String suffix = name.substring(slashPos + 1);
186 String[] range = suffix.split("-");
187 if (range.length == 2)
191 int from = Integer.valueOf(range[0]);
192 int to = Integer.valueOf(range[1]);
193 if (from > 0 && to >= from)
195 name = name.substring(0, slashPos);
200 } catch (NumberFormatException e)
202 // leave name unchanged if suffix is invalid
209 * Ensures that 'end' is not before the end of the sequence, that is,
210 * (end-start+1) is at least as long as the count of ungapped positions. Note
211 * that end is permitted to be beyond the end of the sequence data.
213 void checkValidRange()
216 // http://issues.jalview.org/browse/JAL-774?focusedCommentId=11239&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-11239
219 for (int j = 0; j < sequence.length; j++)
221 if (!Comparison.isGap(sequence[j]))
240 * default constructor
244 sequenceFeatureStore = new SequenceFeatures();
248 * Creates a new Sequence object.
255 public Sequence(String name, String sequence)
257 this(name, sequence, 1, -1);
261 * Creates a new Sequence object with new AlignmentAnnotations but inherits
262 * any existing dataset sequence reference. If non exists, everything is
266 * if seq is a dataset sequence, behaves like a plain old copy
269 public Sequence(SequenceI seq)
271 this(seq, seq.getAnnotation());
275 * Create a new sequence object with new features, DBRefEntries, and PDBIds
276 * but inherits any existing dataset sequence reference, and duplicate of any
277 * annotation that is present in the given annotation array.
280 * the sequence to be copied
281 * @param alAnnotation
282 * an array of annotation including some associated with seq
284 public Sequence(SequenceI seq, AlignmentAnnotation[] alAnnotation)
287 initSeqFrom(seq, alAnnotation);
291 * does the heavy lifting when cloning a dataset sequence, or coping data from
292 * dataset to a new derived sequence.
295 * - source of attributes.
296 * @param alAnnotation
297 * - alignment annotation present on seq that should be copied onto
300 protected void initSeqFrom(SequenceI seq,
301 AlignmentAnnotation[] alAnnotation)
303 char[] oseq = seq.getSequence(); // returns a copy of the array
304 initSeqAndName(seq.getName(), oseq, seq.getStart(), seq.getEnd());
306 description = seq.getDescription();
307 if (seq != datasetSequence)
309 setDatasetSequence(seq.getDatasetSequence());
313 * only copy DBRefs and seqfeatures if we really are a dataset sequence
315 if (datasetSequence == null)
317 List<DBRefEntry> dbr = seq.getDBRefs();
320 for (int i = 0, n = dbr.size(); i < n; i++)
322 addDBRef(new DBRefEntry(dbr.get(i)));
327 * make copies of any sequence features
329 for (SequenceFeature sf : seq.getSequenceFeatures())
331 addSequenceFeature(new SequenceFeature(sf));
335 if (seq.getAnnotation() != null)
337 AlignmentAnnotation[] sqann = seq.getAnnotation();
338 for (int i = 0; i < sqann.length; i++)
340 if (sqann[i] == null)
344 boolean found = (alAnnotation == null);
347 for (int apos = 0; !found && apos < alAnnotation.length; apos++)
349 found = (alAnnotation[apos] == sqann[i]);
354 // only copy the given annotation
355 AlignmentAnnotation newann = new AlignmentAnnotation(sqann[i]);
356 addAlignmentAnnotation(newann);
360 if (seq.getAllPDBEntries() != null)
362 Vector<PDBEntry> ids = seq.getAllPDBEntries();
363 for (PDBEntry pdb : ids)
365 this.addPDBId(new PDBEntry(pdb));
368 if (seq.getHMM() != null)
370 this.hmm = new HiddenMarkovModel(seq.getHMM(), this);
375 public void setSequenceFeatures(List<SequenceFeature> features)
377 if (datasetSequence != null)
379 datasetSequence.setSequenceFeatures(features);
382 sequenceFeatureStore = new SequenceFeatures(features);
386 public synchronized boolean addSequenceFeature(SequenceFeature sf)
388 if (sf.getType() == null)
391 "SequenceFeature type may not be null: " + sf.toString());
395 if (datasetSequence != null)
397 return datasetSequence.addSequenceFeature(sf);
400 return sequenceFeatureStore.add(sf);
404 public void deleteFeature(SequenceFeature sf)
406 if (datasetSequence != null)
408 datasetSequence.deleteFeature(sf);
412 sequenceFeatureStore.delete(sf);
422 public List<SequenceFeature> getSequenceFeatures()
424 if (datasetSequence != null)
426 return datasetSequence.getSequenceFeatures();
428 return sequenceFeatureStore.getAllFeatures();
432 public SequenceFeaturesI getFeatures()
434 return datasetSequence != null ? datasetSequence.getFeatures()
435 : sequenceFeatureStore;
439 public boolean addPDBId(PDBEntry entry)
443 pdbIds = new Vector<>();
448 for (PDBEntry pdbe : pdbIds)
450 if (pdbe.updateFrom(entry))
455 pdbIds.addElement(entry);
466 public void setPDBId(Vector<PDBEntry> id)
474 * @return DOCUMENT ME!
477 public Vector<PDBEntry> getAllPDBEntries()
479 return pdbIds == null ? new Vector<>() : pdbIds;
483 * Answers the sequence name, with '/start-end' appended if jvsuffix is true
488 public String getDisplayId(boolean jvsuffix)
494 StringBuilder result = new StringBuilder(name);
495 result.append("/").append(start).append("-").append(end);
497 return result.toString();
501 * Sets the sequence name. If the name ends in /start-end, then the start-end
502 * values are parsed out and set, and the suffix is removed from the name.
507 public void setName(String theName)
516 * @return DOCUMENT ME!
519 public String getName()
531 public void setStart(int start)
540 * @return DOCUMENT ME!
543 public int getStart()
555 public void setEnd(int end)
563 * @return DOCUMENT ME!
574 * @return DOCUMENT ME!
577 public int getLength()
579 return this.sequence.length;
589 public void setSequence(String seq)
591 this.sequence = seq.toCharArray();
597 public String getSequenceAsString()
599 return new String(sequence);
603 public String getSequenceAsString(int start, int end)
605 return new String(getSequence(start, end));
609 public char[] getSequence()
612 return sequence == null ? null
613 : Arrays.copyOf(sequence, sequence.length);
619 * @see jalview.datamodel.SequenceI#getSequence(int, int)
622 public char[] getSequence(int start, int end)
628 // JBPNote - left to user to pad the result here (TODO:Decide on this
630 if (start >= sequence.length)
635 if (end >= sequence.length)
637 end = sequence.length;
640 char[] reply = new char[end - start];
641 System.arraycopy(sequence, start, reply, 0, end - start);
647 public SequenceI getSubSequence(int start, int end)
653 char[] seq = getSequence(start, end);
658 int nstart = findPosition(start);
659 int nend = findPosition(end) - 1;
660 // JBPNote - this is an incomplete copy.
661 SequenceI nseq = new Sequence(this.getName(), seq, nstart, nend);
662 nseq.setDescription(description);
663 if (datasetSequence != null)
665 nseq.setDatasetSequence(datasetSequence);
669 nseq.setDatasetSequence(this);
675 * Returns the character of the aligned sequence at the given position (base
676 * zero), or space if the position is not within the sequence's bounds
681 public char getCharAt(int i)
683 if (i >= 0 && i < sequence.length)
694 * Sets the sequence description, and also parses out any special formats of
700 public void setDescription(String desc)
702 this.description = desc;
706 public void setGeneLoci(String speciesId, String assemblyId,
707 String chromosomeId, MapList map)
709 addDBRef(new GeneLocus(speciesId, assemblyId, chromosomeId,
714 * Returns the gene loci mapping for the sequence (may be null)
719 public GeneLociI getGeneLoci()
721 List<DBRefEntry> refs = getDBRefs();
724 for (final DBRefEntry ref : refs)
726 if (ref instanceof GeneLociI)
728 return (GeneLociI) ref;
736 * Answers the description
741 public String getDescription()
743 return this.description;
750 public int findIndex(int pos)
753 * use a valid, hopefully nearby, cursor if available
755 if (isValidCursor(cursor))
757 return findIndex(pos, cursor);
765 * traverse sequence from the start counting gaps; make a note of
766 * the column of the first residue to save in the cursor
768 while ((i < sequence.length) && (j <= end) && (j <= pos))
770 if (!Comparison.isGap(sequence[i]))
781 if (j == end && j < pos)
786 updateCursor(pos, i, startColumn);
791 * Updates the cursor to the latest found residue and column position
798 * column position of the first sequence residue
800 protected void updateCursor(int residuePos, int column, int startColumn)
803 * preserve end residue column provided cursor was valid
805 int endColumn = isValidCursor(cursor) ? cursor.lastColumnPosition : 0;
807 if (residuePos == this.end)
812 cursor = new SequenceCursor(this, residuePos, column, startColumn,
813 endColumn, this.changeCount);
817 * Answers the aligned column position (1..) for the given residue position
818 * (start..) given a 'hint' of a residue/column location in the neighbourhood.
819 * The hint may be left of, at, or to the right of the required position.
825 protected int findIndex(final int pos, SequenceCursor curs)
827 if (!isValidCursor(curs))
830 * wrong or invalidated cursor, compute de novo
832 return findIndex(pos);
835 if (curs.residuePosition == pos)
837 return curs.columnPosition;
841 * move left or right to find pos from hint.position
843 int col = curs.columnPosition - 1; // convert from base 1 to base 0
844 int newPos = curs.residuePosition;
845 int delta = newPos > pos ? -1 : 1;
847 while (newPos != pos)
849 col += delta; // shift one column left or right
854 if (col == sequence.length)
856 col--; // return last column if we failed to reach pos
859 if (!Comparison.isGap(sequence[col]))
865 col++; // convert back to base 1
868 * only update cursor if we found the target position
872 updateCursor(pos, col, curs.firstColumnPosition);
882 public int findPosition(final int column)
885 * use a valid, hopefully nearby, cursor if available
887 if (isValidCursor(cursor))
889 return findPosition(column + 1, cursor);
892 // TODO recode this more naturally i.e. count residues only
893 // as they are found, not 'in anticipation'
896 * traverse the sequence counting gaps; note the column position
897 * of the first residue, to save in the cursor
899 int firstResidueColumn = 0;
900 int lastPosFound = 0;
901 int lastPosFoundColumn = 0;
902 int seqlen = sequence.length;
904 if (seqlen > 0 && !Comparison.isGap(sequence[0]))
906 lastPosFound = start;
907 lastPosFoundColumn = 0;
913 while (j < column && j < seqlen)
915 if (!Comparison.isGap(sequence[j]))
918 lastPosFoundColumn = j;
919 if (pos == this.start)
921 firstResidueColumn = j;
927 if (j < seqlen && !Comparison.isGap(sequence[j]))
930 lastPosFoundColumn = j;
931 if (pos == this.start)
933 firstResidueColumn = j;
938 * update the cursor to the last residue position found (if any)
939 * (converting column position to base 1)
941 if (lastPosFound != 0)
943 updateCursor(lastPosFound, lastPosFoundColumn + 1,
944 firstResidueColumn + 1);
951 * Answers true if the given cursor is not null, is for this sequence object,
952 * and has a token value that matches this object's changeCount, else false.
953 * This allows us to ignore a cursor as 'stale' if the sequence has been
954 * modified since the cursor was created.
959 protected boolean isValidCursor(SequenceCursor curs)
961 if (curs == null || curs.sequence != this || curs.token != changeCount)
966 * sanity check against range
968 if (curs.columnPosition < 0 || curs.columnPosition > sequence.length)
972 if (curs.residuePosition < start || curs.residuePosition > end)
980 * Answers the sequence position (start..) for the given aligned column
981 * position (1..), given a hint of a cursor in the neighbourhood. The cursor
982 * may lie left of, at, or to the right of the column position.
988 protected int findPosition(final int col, SequenceCursor curs)
990 if (!isValidCursor(curs))
993 * wrong or invalidated cursor, compute de novo
995 return findPosition(col - 1);// ugh back to base 0
998 if (curs.columnPosition == col)
1000 cursor = curs; // in case this method becomes public
1001 return curs.residuePosition; // easy case :-)
1004 if (curs.lastColumnPosition > 0 && curs.lastColumnPosition < col)
1007 * sequence lies entirely to the left of col
1008 * - return last residue + 1
1013 if (curs.firstColumnPosition > 0 && curs.firstColumnPosition > col)
1016 * sequence lies entirely to the right of col
1017 * - return first residue
1022 // todo could choose closest to col out of column,
1023 // firstColumnPosition, lastColumnPosition as a start point
1026 * move left or right to find pos from cursor position
1028 int firstResidueColumn = curs.firstColumnPosition;
1029 int column = curs.columnPosition - 1; // to base 0
1030 int newPos = curs.residuePosition;
1031 int delta = curs.columnPosition > col ? -1 : 1;
1032 boolean gapped = false;
1033 int lastFoundPosition = curs.residuePosition;
1034 int lastFoundPositionColumn = curs.columnPosition;
1036 while (column != col - 1)
1038 column += delta; // shift one column left or right
1039 if (column < 0 || column == sequence.length)
1043 gapped = Comparison.isGap(sequence[column]);
1047 lastFoundPosition = newPos;
1048 lastFoundPositionColumn = column + 1;
1049 if (lastFoundPosition == this.start)
1051 firstResidueColumn = column + 1;
1056 if (cursor == null || lastFoundPosition != cursor.residuePosition)
1058 updateCursor(lastFoundPosition, lastFoundPositionColumn,
1059 firstResidueColumn);
1063 * hack to give position to the right if on a gap
1064 * or beyond the length of the sequence (see JAL-2562)
1066 if (delta > 0 && (gapped || column >= sequence.length))
1078 public ContiguousI findPositions(int fromColumn, int toColumn)
1080 fromColumn = Math.max(fromColumn, 1);
1081 if (toColumn < fromColumn)
1087 * find the first non-gapped position, if any
1089 int firstPosition = 0;
1090 int col = fromColumn - 1;
1091 int length = sequence.length;
1092 while (col < length && col < toColumn)
1094 if (!Comparison.isGap(sequence[col]))
1096 firstPosition = findPosition(col++);
1102 if (firstPosition == 0)
1108 * find the last non-gapped position
1110 int lastPosition = firstPosition;
1111 while (col < length && col < toColumn)
1113 if (!Comparison.isGap(sequence[col++]))
1119 return new Range(firstPosition, lastPosition);
1123 * Returns an int array where indices correspond to each residue in the
1124 * sequence and the element value gives its position in the alignment
1126 * @return int[SequenceI.getEnd()-SequenceI.getStart()+1] or null if no
1127 * residues in SequenceI object
1130 public int[] gapMap()
1132 String seq = jalview.analysis.AlignSeq.extractGaps(
1133 jalview.util.Comparison.GapChars, new String(sequence));
1134 int[] map = new int[seq.length()];
1138 while (j < sequence.length)
1140 if (!jalview.util.Comparison.isGap(sequence[j]))
1152 * Build a bitset corresponding to sequence gaps
1154 * @return a BitSet where set values correspond to gaps in the sequence
1157 public BitSet gapBitset()
1159 BitSet gaps = new BitSet(sequence.length);
1161 while (j < sequence.length)
1163 if (jalview.util.Comparison.isGap(sequence[j]))
1173 public int[] findPositionMap()
1175 int map[] = new int[sequence.length];
1178 int seqlen = sequence.length;
1179 while ((j < seqlen))
1182 if (!jalview.util.Comparison.isGap(sequence[j]))
1193 public List<int[]> getInsertions()
1195 ArrayList<int[]> map = new ArrayList<>();
1196 int lastj = -1, j = 0;
1198 int seqlen = sequence.length;
1199 while ((j < seqlen))
1201 if (jalview.util.Comparison.isGap(sequence[j]))
1212 map.add(new int[] { lastj, j - 1 });
1220 map.add(new int[] { lastj, j - 1 });
1227 public BitSet getInsertionsAsBits()
1229 BitSet map = new BitSet();
1230 int lastj = -1, j = 0;
1232 int seqlen = sequence.length;
1233 while ((j < seqlen))
1235 if (jalview.util.Comparison.isGap(sequence[j]))
1261 public void deleteChars(final int i, final int j)
1263 int newstart = start, newend = end;
1264 if (i >= sequence.length || i < 0)
1269 char[] tmp = StringUtils.deleteChars(sequence, i, j);
1270 boolean createNewDs = false;
1271 // TODO: take a (second look) at the dataset creation validation method for
1272 // the very large sequence case
1274 int startIndex = findIndex(start) - 1;
1275 int endIndex = findIndex(end) - 1;
1276 int startDeleteColumn = -1; // for dataset sequence deletions
1277 int deleteCount = 0;
1279 for (int s = i; s < j && s < sequence.length; s++)
1281 if (Comparison.isGap(sequence[s]))
1286 if (startDeleteColumn == -1)
1288 startDeleteColumn = findPosition(s) - start;
1296 if (startIndex == s)
1299 * deleting characters from start of sequence; new start is the
1300 * sequence position of the next column (position to the right
1301 * if the column position is gapped)
1303 newstart = findPosition(j);
1311 * deleting characters at end of sequence; new end is the sequence
1312 * position of the column before the deletion; subtract 1 if this is
1313 * gapped since findPosition returns the next sequence position
1315 newend = findPosition(i - 1);
1316 if (Comparison.isGap(sequence[i - 1]))
1331 if (createNewDs && this.datasetSequence != null)
1334 * if deletion occured in the middle of the sequence,
1335 * construct a new dataset sequence and delete the residues
1336 * that were deleted from the aligned sequence
1338 Sequence ds = new Sequence(datasetSequence);
1339 ds.deleteChars(startDeleteColumn, startDeleteColumn + deleteCount);
1340 datasetSequence = ds;
1341 // TODO: remove any non-inheritable properties ?
1342 // TODO: create a sequence mapping (since there is a relation here ?)
1351 public void insertCharAt(int i, int length, char c)
1353 char[] tmp = new char[sequence.length + length];
1355 if (i >= sequence.length)
1357 System.arraycopy(sequence, 0, tmp, 0, sequence.length);
1358 i = sequence.length;
1362 System.arraycopy(sequence, 0, tmp, 0, i);
1372 if (i < sequence.length)
1374 System.arraycopy(sequence, i, tmp, index, sequence.length - i);
1382 public void insertCharAt(int i, char c)
1384 insertCharAt(i, 1, c);
1388 public String getVamsasId()
1394 public void setVamsasId(String id)
1401 public void setDBRefs(DBModList<DBRefEntry> newDBrefs)
1403 if (dbrefs == null && datasetSequence != null
1404 && this != datasetSequence)
1406 datasetSequence.setDBRefs(newDBrefs);
1414 public DBModList<DBRefEntry> getDBRefs()
1416 if (dbrefs == null && datasetSequence != null
1417 && this != datasetSequence)
1419 return datasetSequence.getDBRefs();
1425 public void addDBRef(DBRefEntry entry)
1427 if (datasetSequence != null)
1429 datasetSequence.addDBRef(entry);
1435 dbrefs = new DBModList<>();
1438 for (int ib = 0, nb = dbrefs.size(); ib < nb; ib++)
1440 if (dbrefs.get(ib).updateFrom(entry))
1443 * found a dbref that either matched, or could be
1444 * updated from, the new entry - no need to add it
1452 // * extend the array to make room for one more
1454 // // TODO use an ArrayList instead
1455 // int j = dbrefs.length;
1456 // List<DBRefEntry> temp = new DBRefEntry[j + 1];
1457 // System.arraycopy(dbrefs, 0, temp, 0, j);
1458 // temp[temp.length - 1] = entry;
1466 public void setDatasetSequence(SequenceI seq)
1470 throw new IllegalArgumentException(
1471 "Implementation Error: self reference passed to SequenceI.setDatasetSequence");
1473 if (seq != null && seq.getDatasetSequence() != null)
1475 throw new IllegalArgumentException(
1476 "Implementation error: cascading dataset sequences are not allowed.");
1478 datasetSequence = seq;
1482 public SequenceI getDatasetSequence()
1484 return datasetSequence;
1488 public AlignmentAnnotation[] getAnnotation()
1490 return annotation == null ? null
1492 .toArray(new AlignmentAnnotation[annotation.size()]);
1496 public boolean hasAnnotation(AlignmentAnnotation ann)
1498 return annotation == null ? false : annotation.contains(ann);
1502 public void addAlignmentAnnotation(AlignmentAnnotation annotation)
1504 if (this.annotation == null)
1506 this.annotation = new Vector<>();
1508 if (!this.annotation.contains(annotation))
1510 this.annotation.addElement(annotation);
1512 annotation.setSequenceRef(this);
1516 public void removeAlignmentAnnotation(AlignmentAnnotation annotation)
1518 if (this.annotation != null)
1520 this.annotation.removeElement(annotation);
1521 if (this.annotation.size() == 0)
1523 this.annotation = null;
1529 * test if this is a valid candidate for another sequence's dataset sequence.
1532 private boolean isValidDatasetSequence()
1534 if (datasetSequence != null)
1538 for (int i = 0; i < sequence.length; i++)
1540 if (jalview.util.Comparison.isGap(sequence[i]))
1549 public SequenceI deriveSequence()
1551 Sequence seq = null;
1552 if (datasetSequence == null)
1554 if (isValidDatasetSequence())
1556 // Use this as dataset sequence
1557 seq = new Sequence(getName(), "", 1, -1);
1558 seq.setDatasetSequence(this);
1559 seq.initSeqFrom(this, getAnnotation());
1564 // Create a new, valid dataset sequence
1565 createDatasetSequence();
1568 return new Sequence(this);
1571 private boolean _isNa;
1573 private int _seqhash = 0;
1575 private List<DBRefEntry> primaryRefs;
1578 * Answers false if the sequence is more than 85% nucleotide (ACGTU), else
1582 public boolean isProtein()
1584 if (datasetSequence != null)
1586 return datasetSequence.isProtein();
1588 if (_seqhash != sequence.hashCode())
1590 _seqhash = sequence.hashCode();
1591 _isNa = Comparison.isNucleotide(this);
1599 * @see jalview.datamodel.SequenceI#createDatasetSequence()
1602 public SequenceI createDatasetSequence()
1604 if (datasetSequence == null)
1606 Sequence dsseq = new Sequence(getName(),
1607 AlignSeq.extractGaps(jalview.util.Comparison.GapChars,
1608 getSequenceAsString()),
1609 getStart(), getEnd());
1611 datasetSequence = dsseq;
1613 dsseq.setDescription(description);
1614 // move features and database references onto dataset sequence
1615 dsseq.sequenceFeatureStore = sequenceFeatureStore;
1616 sequenceFeatureStore = null;
1617 dsseq.dbrefs = dbrefs;
1619 // TODO: search and replace any references to this sequence with
1620 // references to the dataset sequence in Mappings on dbref
1621 dsseq.pdbIds = pdbIds;
1623 datasetSequence.updatePDBIds();
1624 if (annotation != null)
1626 // annotation is cloned rather than moved, to preserve what's currently
1628 for (AlignmentAnnotation aa : annotation)
1630 AlignmentAnnotation _aa = new AlignmentAnnotation(aa);
1631 _aa.sequenceRef = datasetSequence;
1632 _aa.adjustForAlignment(); // uses annotation's own record of
1633 // sequence-column mapping
1634 datasetSequence.addAlignmentAnnotation(_aa);
1638 return datasetSequence;
1645 * jalview.datamodel.SequenceI#setAlignmentAnnotation(AlignmmentAnnotation[]
1649 public void setAlignmentAnnotation(AlignmentAnnotation[] annotations)
1651 if (annotation != null)
1653 annotation.removeAllElements();
1655 if (annotations != null)
1657 for (int i = 0; i < annotations.length; i++)
1659 if (annotations[i] != null)
1661 addAlignmentAnnotation(annotations[i]);
1668 public AlignmentAnnotation[] getAnnotation(String label)
1670 if (annotation == null || annotation.size() == 0)
1675 Vector<AlignmentAnnotation> subset = new Vector<>();
1676 Enumeration<AlignmentAnnotation> e = annotation.elements();
1677 while (e.hasMoreElements())
1679 AlignmentAnnotation ann = e.nextElement();
1680 if (ann.label != null && ann.label.equals(label))
1682 subset.addElement(ann);
1685 if (subset.size() == 0)
1689 AlignmentAnnotation[] anns = new AlignmentAnnotation[subset.size()];
1691 e = subset.elements();
1692 while (e.hasMoreElements())
1694 anns[i++] = e.nextElement();
1696 subset.removeAllElements();
1701 public boolean updatePDBIds()
1703 if (datasetSequence != null)
1705 // TODO: could merge DBRefs
1706 return datasetSequence.updatePDBIds();
1708 if (dbrefs == null || dbrefs.size() == 0)
1712 boolean added = false;
1713 for (int ib = 0, nb = dbrefs.size(); ib < nb; ib++)
1715 DBRefEntry dbr = dbrefs.get(ib);
1716 if (DBRefSource.PDB.equals(dbr.getSource()))
1719 * 'Add' any PDB dbrefs as a PDBEntry - add is only performed if the
1720 * PDB id is not already present in a 'matching' PDBEntry
1721 * Constructor parses out a chain code if appended to the accession id
1722 * (a fudge used to 'store' the chain code in the DBRef)
1724 PDBEntry pdbe = new PDBEntry(dbr);
1725 added |= addPDBId(pdbe);
1732 public void transferAnnotation(SequenceI entry, Mapping mp)
1734 if (datasetSequence != null)
1736 datasetSequence.transferAnnotation(entry, mp);
1739 if (entry.getDatasetSequence() != null)
1741 transferAnnotation(entry.getDatasetSequence(), mp);
1744 // transfer any new features from entry onto sequence
1745 if (entry.getSequenceFeatures() != null)
1748 List<SequenceFeature> sfs = entry.getSequenceFeatures();
1749 for (SequenceFeature feature : sfs)
1751 SequenceFeature sf[] = (mp != null) ? mp.locateFeature(feature)
1752 : new SequenceFeature[]
1753 { new SequenceFeature(feature) };
1756 for (int sfi = 0; sfi < sf.length; sfi++)
1758 addSequenceFeature(sf[sfi]);
1764 // transfer PDB entries
1765 if (entry.getAllPDBEntries() != null)
1767 Enumeration<PDBEntry> e = entry.getAllPDBEntries().elements();
1768 while (e.hasMoreElements())
1770 PDBEntry pdb = e.nextElement();
1774 // transfer database references
1775 List<DBRefEntry> entryRefs = entry.getDBRefs();
1776 if (entryRefs != null)
1778 for (int r = 0, n = entryRefs.size(); r < n; r++)
1780 DBRefEntry newref = new DBRefEntry(entryRefs.get(r));
1781 if (newref.getMap() != null && mp != null)
1783 // remap ref using our local mapping
1785 // we also assume all version string setting is done by dbSourceProxy
1787 * if (!newref.getSource().equalsIgnoreCase(dbSource)) {
1788 * newref.setSource(dbSource); }
1796 public void setRNA(RNA r)
1808 public List<AlignmentAnnotation> getAlignmentAnnotations(String calcId,
1811 return getAlignmentAnnotations(calcId, label, null, true);
1815 public List<AlignmentAnnotation> getAlignmentAnnotations(String calcId,
1816 String label, String description)
1818 return getAlignmentAnnotations(calcId, label, description, false);
1821 private List<AlignmentAnnotation> getAlignmentAnnotations(String calcId,
1822 String label, String description, boolean ignoreDescription)
1824 List<AlignmentAnnotation> result = new ArrayList<>();
1825 if (this.annotation != null)
1827 for (AlignmentAnnotation ann : annotation)
1829 String id = ann.getCalcId();
1830 if ((id != null && id.equals(calcId))
1831 && (ann.label != null && ann.label.equals(label))
1832 && ((ignoreDescription && description == null)
1833 || (ann.description != null
1834 && ann.description.equals(description))))
1844 public String toString()
1846 return getDisplayId(false);
1850 public PDBEntry getPDBEntry(String pdbIdStr)
1852 if (getDatasetSequence() != null)
1854 return getDatasetSequence().getPDBEntry(pdbIdStr);
1860 List<PDBEntry> entries = getAllPDBEntries();
1861 for (PDBEntry entry : entries)
1863 if (entry.getId().equalsIgnoreCase(pdbIdStr))
1871 private List<DBRefEntry> tmpList;
1874 public List<DBRefEntry> getPrimaryDBRefs()
1876 if (datasetSequence != null)
1878 return datasetSequence.getPrimaryDBRefs();
1880 if (dbrefs == null || dbrefs.size() == 0)
1882 return Collections.emptyList();
1884 synchronized (dbrefs)
1886 if (refModCount == dbrefs.getModCount() && primaryRefs != null)
1888 return primaryRefs; // no changes
1890 refModCount = dbrefs.getModCount();
1891 List<DBRefEntry> primaries = (primaryRefs == null
1892 ? (primaryRefs = new ArrayList<>())
1895 if (tmpList == null)
1897 tmpList = new ArrayList<>();
1898 tmpList.add(null); // for replacement
1900 for (int i = 0, n = dbrefs.size(); i < n; i++)
1902 DBRefEntry ref = dbrefs.get(i);
1903 if (!ref.isPrimaryCandidate())
1909 MapList mp = ref.getMap().getMap();
1910 if (mp.getFromLowest() > start || mp.getFromHighest() < end)
1912 // map only involves a subsequence, so cannot be primary
1916 // whilst it looks like it is a primary ref, we also sanity check type
1917 if (DBRefSource.PDB_CANONICAL_NAME
1918 .equals(ref.getCanonicalSourceName()))
1920 // PDB dbrefs imply there should be a PDBEntry associated
1921 // TODO: tighten PDB dbrefs
1922 // formally imply Jalview has actually downloaded and
1923 // parsed the pdb file. That means there should be a cached file
1924 // handle on the PDBEntry, and a real mapping between sequence and
1925 // extracted sequence from PDB file
1926 PDBEntry pdbentry = getPDBEntry(ref.getAccessionId());
1927 if (pdbentry == null || pdbentry.getFile() == null)
1934 // check standard protein or dna sources
1935 tmpList.set(0, ref);
1936 List<DBRefEntry> res = DBRefUtils.selectDbRefs(!isProtein(),
1938 if (res == null || res.get(0) != tmpList.get(0))
1946 // version must be not null, as otherwise it will not be a candidate,
1948 DBRefUtils.ensurePrimaries(this, primaries);
1954 public HiddenMarkovModel getHMM()
1960 public void setHMM(HiddenMarkovModel hmm)
1966 public boolean hasHMMAnnotation()
1968 if (this.annotation == null) {
1971 for (AlignmentAnnotation ann : annotation)
1973 if (InformationThread.HMM_CALC_ID.equals(ann.getCalcId()))
1984 public List<SequenceFeature> findFeatures(int fromColumn, int toColumn,
1987 int startPos = findPosition(fromColumn - 1); // convert base 1 to base 0
1988 int endPos = fromColumn == toColumn ? startPos
1989 : findPosition(toColumn - 1);
1991 List<SequenceFeature> result = getFeatures().findFeatures(startPos,
1995 * if end column is gapped, endPos may be to the right,
1996 * and we may have included adjacent or enclosing features;
1997 * remove any that are not enclosing, non-contact features
1999 boolean endColumnIsGapped = toColumn > 0 && toColumn <= sequence.length
2000 && Comparison.isGap(sequence[toColumn - 1]);
2001 if (endPos > this.end || endColumnIsGapped)
2003 ListIterator<SequenceFeature> it = result.listIterator();
2004 while (it.hasNext())
2006 SequenceFeature sf = it.next();
2007 int sfBegin = sf.getBegin();
2008 int sfEnd = sf.getEnd();
2009 int featureStartColumn = findIndex(sfBegin);
2010 if (featureStartColumn > toColumn)
2014 else if (featureStartColumn < fromColumn)
2016 int featureEndColumn = sfEnd == sfBegin ? featureStartColumn
2018 if (featureEndColumn < fromColumn)
2022 else if (featureEndColumn > toColumn && sf.isContactFeature())
2025 * remove an enclosing feature if it is a contact feature
2037 * Invalidates any stale cursors (forcing recalculation) by incrementing the
2038 * token that has to match the one presented by the cursor
2041 public void sequenceChanged()
2050 public int replace(char c1, char c2)
2057 synchronized (sequence)
2059 for (int c = 0; c < sequence.length; c++)
2061 if (sequence[c] == c1)
2077 public String getSequenceStringFromIterator(Iterator<int[]> it)
2079 StringBuilder newSequence = new StringBuilder();
2080 while (it.hasNext())
2082 int[] block = it.next();
2085 newSequence.append(getSequence(block[0], block[1] + 1));
2089 newSequence.append(getSequence(block[0], block[1]));
2093 return newSequence.toString();
2097 public int firstResidueOutsideIterator(Iterator<int[]> regions)
2101 if (!regions.hasNext())
2103 return findIndex(getStart()) - 1;
2106 // Simply walk along the sequence whilst watching for region
2108 int hideStart = getLength();
2110 boolean foundStart = false;
2112 // step through the non-gapped positions of the sequence
2113 for (int i = getStart(); i <= getEnd() && (!foundStart); i++)
2115 // get alignment position of this residue in the sequence
2116 int p = findIndex(i) - 1;
2118 // update region start/end
2119 while (hideEnd < p && regions.hasNext())
2121 int[] region = regions.next();
2122 hideStart = region[0];
2123 hideEnd = region[1];
2127 hideStart = getLength();
2129 // update boundary for sequence
2141 // otherwise, sequence was completely hidden
2146 public boolean hasHMMProfile()