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.util.MapList;
24 import jalview.util.MappingUtils;
26 import java.util.AbstractList;
27 import java.util.ArrayList;
28 import java.util.List;
31 * Stores mapping between the columns of a protein alignment and a DNA alignment
32 * and a list of individual codon to amino acid mappings between sequences.
34 public class AlignedCodonFrame
38 * Data bean to hold mappings from one sequence to another
40 public class SequenceToSequenceMapping
42 private SequenceI fromSeq;
44 private Mapping mapping;
46 SequenceToSequenceMapping(SequenceI from, Mapping map)
53 * Readable representation for debugging only, not guaranteed not to change
56 public String toString()
58 return String.format("From %s %s", fromSeq.getName(),
63 * Returns a hashCode derived from the hashcodes of the mappings and fromSeq
65 * @see SequenceToSequenceMapping#hashCode()
70 return (fromSeq == null ? 0 : fromSeq.hashCode() * 31)
75 * Answers true if the objects hold the same mapping between the same two
81 public boolean equals(Object obj)
83 if (!(obj instanceof SequenceToSequenceMapping))
87 SequenceToSequenceMapping that = (SequenceToSequenceMapping) obj;
88 if (this.mapping == null)
90 return that.mapping == null;
92 // TODO: can simplify by asserting fromSeq is a dataset sequence
93 return (this.fromSeq == that.fromSeq
94 || (this.fromSeq != null && that.fromSeq != null
95 && this.fromSeq.getDatasetSequence() != null
96 && this.fromSeq.getDatasetSequence() == that.fromSeq
97 .getDatasetSequence()))
98 && this.mapping.equals(that.mapping);
101 public SequenceI getFromSeq()
106 public Mapping getMapping()
112 * Returns true if the mapping covers the full length of the given sequence.
113 * This allows us to distinguish the CDS that codes for a protein from
114 * another overlapping CDS in the parent dna sequence.
119 public boolean covers(SequenceI seq)
121 return covers(seq,false,false);
126 * @param localCover - when true - compare extent of seq's dataset sequence rather than the local extent
127 * @param either - when true coverage is required for either seq or the mapped sequence
128 * @return true if mapping covers full length of given sequence (or the other if either==true)
130 public boolean covers(SequenceI seq, boolean localCover,boolean either)
132 List<int[]> mappedRanges = null,otherRanges=null;
133 MapList mapList = mapping.getMap();
134 int mstart=seq.getStart(),mend=seq.getEnd(),ostart,oend;
136 if (fromSeq == seq || fromSeq == seq.getDatasetSequence())
138 if (localCover && fromSeq !=seq)
140 mstart=fromSeq.getStart();
141 mend=fromSeq.getEnd();
143 mappedRanges = mapList.getFromRanges();
144 otherRanges=mapList.getToRanges();
145 ostart=mapping.to.getStart();
146 oend=mapping.to.getEnd();
148 else if (mapping.to == seq || mapping.to == seq.getDatasetSequence())
150 if (localCover && mapping.to !=seq)
152 mstart=mapping.to.getStart();
153 mend=mapping.to.getEnd();
155 mappedRanges = mapList.getToRanges();
156 otherRanges=mapList.getFromRanges();
157 ostart=fromSeq.getStart();
158 oend=fromSeq.getEnd();
166 * check that each mapped range lieS with the sequence range
167 * (necessary for circular CDS - example EMBL:J03321:AAA91567)
168 * and mapped length covers (at least) sequence length
170 int length = countRange(mappedRanges,mstart,mend);
174 // add 1 to mapped length to allow for a mapped stop codon
175 if (length + 1 >= (mend - mstart + 1))
182 // also check coverage of the other range
183 length = countRange(otherRanges, ostart, oend);
186 if (length + 1 >= (oend - ostart + 1))
194 private int countRange(List<int[]> mappedRanges,int mstart,int mend)
197 for (int[] range : mappedRanges)
199 int from = Math.min(range[0], range[1]);
200 int to = Math.max(range[0], range[1]);
201 if (from < mstart || to > mend)
205 length += (to - from + 1);
211 private List<SequenceToSequenceMapping> mappings;
216 public AlignedCodonFrame()
218 mappings = new ArrayList<>();
222 * Adds a mapping between the dataset sequences for the associated dna and
223 * protein sequence objects
229 public void addMap(SequenceI dnaseq, SequenceI aaseq, MapList map)
231 addMap(dnaseq, aaseq, map, null);
235 * Adds a mapping between the dataset sequences for the associated dna and
236 * protein sequence objects
243 public void addMap(SequenceI dnaseq, SequenceI aaseq, MapList map,
246 // JBPNote DEBUG! THIS !
247 // dnaseq.transferAnnotation(aaseq, mp);
248 // aaseq.transferAnnotation(dnaseq, new Mapping(map.getInverse()));
250 SequenceI fromSeq = (dnaseq.getDatasetSequence() == null) ? dnaseq
251 : dnaseq.getDatasetSequence();
252 SequenceI toSeq = (aaseq.getDatasetSequence() == null) ? aaseq
253 : aaseq.getDatasetSequence();
256 * if we already hold a mapping between these sequences, just add to it
257 * note that 'adding' a duplicate map does nothing; this protects against
258 * creating duplicate mappings in AlignedCodonFrame
260 for (SequenceToSequenceMapping ssm : mappings)
262 if (ssm.fromSeq == fromSeq && ssm.mapping.to == toSeq)
264 ssm.mapping.map.addMapList(map);
270 * otherwise, add a new sequence mapping
272 Mapping mp = new Mapping(toSeq, map);
273 mp.setMappedFromId(mapFromId);
274 mappings.add(new SequenceToSequenceMapping(fromSeq, mp));
277 public SequenceI[] getdnaSeqs()
279 // TODO return a list instead?
281 List<SequenceI> seqs = new ArrayList<>();
282 for (SequenceToSequenceMapping ssm : mappings)
284 seqs.add(ssm.fromSeq);
286 return seqs.toArray(new SequenceI[seqs.size()]);
289 public SequenceI[] getAaSeqs()
291 // TODO not used - remove?
292 List<SequenceI> seqs = new ArrayList<>();
293 for (SequenceToSequenceMapping ssm : mappings)
295 seqs.add(ssm.mapping.to);
297 return seqs.toArray(new SequenceI[seqs.size()]);
300 public MapList[] getdnaToProt()
302 List<MapList> maps = new ArrayList<>();
303 for (SequenceToSequenceMapping ssm : mappings)
305 maps.add(ssm.mapping.map);
307 return maps.toArray(new MapList[maps.size()]);
310 public Mapping[] getProtMappings()
312 List<Mapping> maps = new ArrayList<>();
313 for (SequenceToSequenceMapping ssm : mappings)
315 maps.add(ssm.mapping);
317 return maps.toArray(new Mapping[maps.size()]);
321 * Returns the first mapping found which is to or from the given sequence, or
322 * null if none is found
327 public Mapping getMappingForSequence(SequenceI seq)
329 SequenceI seqDs = seq.getDatasetSequence();
330 seqDs = seqDs != null ? seqDs : seq;
332 for (SequenceToSequenceMapping ssm : mappings)
334 if (ssm.fromSeq == seqDs || ssm.mapping.to == seqDs)
343 * Return the corresponding aligned or dataset aa sequence for given dna
344 * sequence, null if not found.
349 public SequenceI getAaForDnaSeq(SequenceI dnaSeqRef)
351 SequenceI dnads = dnaSeqRef.getDatasetSequence();
352 for (SequenceToSequenceMapping ssm : mappings)
354 if (ssm.fromSeq == dnaSeqRef || ssm.fromSeq == dnads)
356 return ssm.mapping.to;
363 * Return the corresponding aligned or dataset dna sequence for given amino
364 * acid sequence, or null if not found. returns the sequence from the first
365 * mapping found that involves the protein sequence.
370 public SequenceI getDnaForAaSeq(SequenceI aaSeqRef)
372 SequenceI aads = aaSeqRef.getDatasetSequence();
373 for (SequenceToSequenceMapping ssm : mappings)
375 if (ssm.mapping.to == aaSeqRef || ssm.mapping.to == aads)
384 * test to see if codon frame involves seq in any way
387 * a nucleotide or protein sequence
388 * @return true if a mapping exists to or from this sequence to any translated
391 public boolean involvesSequence(SequenceI seq)
393 return getAaForDnaSeq(seq) != null || getDnaForAaSeq(seq) != null;
397 * Add search results for regions in other sequences that translate or are
398 * translated from a particular position in seq
404 * where highlighted regions go
406 public void markMappedRegion(SequenceI seq, int index,
407 SearchResultsI results)
410 SequenceI ds = seq.getDatasetSequence();
411 for (SequenceToSequenceMapping ssm : mappings)
413 if (ssm.fromSeq == seq || ssm.fromSeq == ds)
415 codon = ssm.mapping.map.locateInTo(index, index);
418 for (int i = 0; i < codon.length; i += 2)
420 results.addResult(ssm.mapping.to, codon[i], codon[i + 1]);
424 else if (ssm.mapping.to == seq || ssm.mapping.to == ds)
427 codon = ssm.mapping.map.locateInFrom(index, index);
430 for (int i = 0; i < codon.length; i += 2)
432 results.addResult(ssm.fromSeq, codon[i], codon[i + 1]);
441 * Returns the DNA codon positions (base 1) for the given position (base 1) in
442 * a mapped protein sequence, or null if no mapping is found.
444 * Intended for use in aligning cDNA to match aligned protein. Only the first
445 * mapping found is returned, so not suitable for use if multiple protein
446 * sequences are mapped to the same cDNA (but aligning cDNA as protein is
447 * ill-defined for this case anyway).
450 * the DNA dataset sequence
452 * residue position (base 1) in a protein sequence
455 public int[] getDnaPosition(SequenceI seq, int aaPos)
458 * Adapted from markMappedRegion().
462 for (SequenceToSequenceMapping ssm : mappings)
464 if (ssm.fromSeq == seq)
466 ml = getdnaToProt()[i];
471 return ml == null ? null : ml.locateInFrom(aaPos, aaPos);
475 * Convenience method to return the first aligned sequence in the given
476 * alignment whose dataset has a mapping with the given (aligned or dataset)
484 public SequenceI findAlignedSequence(SequenceI seq, AlignmentI al)
486 return findAlignedSequence(seq, al, null);
489 * Convenience method to return the first aligned sequence in the given
490 * alignment whose dataset has a mapping with the given (aligned or dataset)
491 * sequence, and optionally the mapping that relates them
495 * @param map - list to add the mapping to
496 * @return sequence from al that maps to seq
498 public SequenceI findAlignedSequence(SequenceI seq, AlignmentI al,List<SequenceToSequenceMapping> map)
501 * Search mapped protein ('to') sequences first.
503 for (SequenceToSequenceMapping ssm : mappings)
505 int mStart=ssm.getMapping().getMap().getFromLowest(),mEnd=ssm.getMapping().map.getFromHighest();
506 if ((ssm.fromSeq == seq || ssm.fromSeq == seq.getDatasetSequence())
507 // here AlignmentUtilsTest. testAlignProteinAsDna_incompleteStartCodon fails because mStart/mEnd is contained by seq
508 // without this filter, we don't get the correct mapping, however
509 )// && seq.getStart()>=mStart && seq.getEnd()<=mEnd)
511 for (SequenceI sourceAligned : al.getSequences())
513 if (ssm.covers(sourceAligned,true,false))
519 return sourceAligned;
526 * Then try mapped dna sequences.
528 for (SequenceToSequenceMapping ssm : mappings)
530 int mStart=ssm.getMapping().getMap().getToLowest(),mEnd=ssm.getMapping().map.getToHighest();
531 if ((ssm.mapping.to == seq
532 || ssm.mapping.to == seq.getDatasetSequence())
533 && seq.getStart()>=mStart && seq.getEnd()<=mEnd)
535 for (SequenceI sourceAligned : al.getSequences())
537 if (ssm.covers(sourceAligned,true,true))
543 return sourceAligned;
553 * Returns the region in the target sequence's dataset that is mapped to the
554 * given position (base 1) in the query sequence's dataset. The region is a
555 * set of start/end position pairs.
562 public int[] getMappedRegion(SequenceI target, SequenceI query,
565 SequenceI targetDs = target.getDatasetSequence() == null ? target
566 : target.getDatasetSequence();
567 SequenceI queryDs = query.getDatasetSequence() == null ? query
568 : query.getDatasetSequence();
569 if (targetDs == null || queryDs == null /*|| dnaToProt == null*/)
573 for (SequenceToSequenceMapping ssm : mappings)
576 * try mapping from target to query
578 if (ssm.fromSeq == targetDs && ssm.mapping.to == queryDs)
580 int[] codon = ssm.mapping.map.locateInFrom(queryPos, queryPos);
587 * else try mapping from query to target
589 else if (ssm.fromSeq == queryDs && ssm.mapping.to == targetDs)
591 int[] codon = ssm.mapping.map.locateInTo(queryPos, queryPos);
602 * Returns the mapped DNA codons for the given position in a protein sequence,
603 * or null if no mapping is found. Returns a list of (e.g.) ['g', 'c', 't']
604 * codons. There may be more than one codon mapped to the protein if (for
605 * example), there are mappings to cDNA variants.
608 * the peptide dataset sequence
610 * residue position (base 1) in the peptide sequence
613 public List<char[]> getMappedCodons(SequenceI protein, int aaPos)
616 SequenceI dnaSeq = null;
617 List<char[]> result = new ArrayList<>();
619 for (SequenceToSequenceMapping ssm : mappings)
621 if (ssm.mapping.to == protein
622 && ssm.mapping.getMap().getFromRatio() == 3)
624 ml = ssm.mapping.map;
625 dnaSeq = ssm.fromSeq;
627 int[] codonPos = ml.locateInFrom(aaPos, aaPos);
628 if (codonPos == null)
634 * Read off the mapped nucleotides (converting to position base 0)
636 codonPos = MappingUtils.flattenRanges(codonPos);
637 int start = dnaSeq.getStart();
638 char c1 = dnaSeq.getCharAt(codonPos[0] - start);
639 char c2 = dnaSeq.getCharAt(codonPos[1] - start);
640 char c3 = dnaSeq.getCharAt(codonPos[2] - start);
641 result.add(new char[] { c1, c2, c3 });
644 return result.isEmpty() ? null : result;
648 * Returns any mappings found which are from the given sequence, and to
649 * distinct sequences.
654 public List<Mapping> getMappingsFromSequence(SequenceI seq)
656 List<Mapping> result = new ArrayList<>();
657 List<SequenceI> related = new ArrayList<>();
658 SequenceI seqDs = seq.getDatasetSequence();
659 seqDs = seqDs != null ? seqDs : seq;
661 for (SequenceToSequenceMapping ssm : mappings)
663 final Mapping mapping = ssm.mapping;
664 if (ssm.fromSeq == seqDs)
666 if (!related.contains(mapping.to))
669 related.add(mapping.to);
677 * Test whether the given sequence is substitutable for one or more dummy
678 * sequences in this mapping
684 public boolean isRealisableWith(SequenceI seq)
686 return realiseWith(seq, false) > 0;
690 * Replace any matchable mapped dummy sequences with the given real one.
691 * Returns the count of sequence mappings instantiated.
696 public int realiseWith(SequenceI seq)
698 return realiseWith(seq, true);
702 * Returns the number of mapped dummy sequences that could be replaced with
703 * the given real sequence.
708 * if true, performs replacements, else only counts
711 protected int realiseWith(SequenceI seq, boolean doUpdate)
713 SequenceI ds = seq.getDatasetSequence() != null
714 ? seq.getDatasetSequence()
719 * check for replaceable DNA ('map from') sequences
721 for (SequenceToSequenceMapping ssm : mappings)
723 SequenceI dna = ssm.fromSeq;
724 if (dna instanceof SequenceDummy
725 && dna.getName().equals(ds.getName()))
727 Mapping mapping = ssm.mapping;
728 int mapStart = mapping.getMap().getFromLowest();
729 int mapEnd = mapping.getMap().getFromHighest();
730 boolean mappable = couldRealiseSequence(dna, ds, mapStart, mapEnd);
736 // TODO: new method ? ds.realise(dna);
737 // might want to copy database refs as well
738 ds.setSequenceFeatures(dna.getSequenceFeatures());
741 System.out.println("Realised mapped sequence " + ds.getName());
747 * check for replaceable protein ('map to') sequences
749 Mapping mapping = ssm.mapping;
750 SequenceI prot = mapping.getTo();
751 int mapStart = mapping.getMap().getToLowest();
752 int mapEnd = mapping.getMap().getToHighest();
753 boolean mappable = couldRealiseSequence(prot, ds, mapStart, mapEnd);
759 // TODO: new method ? ds.realise(dna);
760 // might want to copy database refs as well
761 ds.setSequenceFeatures(dna.getSequenceFeatures());
762 ssm.mapping.setTo(ds);
770 * Helper method to test whether a 'real' sequence could replace a 'dummy'
771 * sequence in the map. The criteria are that they have the same name, and
772 * that the mapped region overlaps the candidate sequence.
780 protected static boolean couldRealiseSequence(SequenceI existing,
781 SequenceI replacement, int mapStart, int mapEnd)
783 if (existing instanceof SequenceDummy
784 && !(replacement instanceof SequenceDummy)
785 && existing.getName().equals(replacement.getName()))
787 int start = replacement.getStart();
788 int end = replacement.getEnd();
789 boolean mappingOverlapsSequence = (mapStart >= start
790 && mapStart <= end) || (mapEnd >= start && mapEnd <= end);
791 if (mappingOverlapsSequence)
800 * Change any mapping to the given sequence to be to its dataset sequence
801 * instead. For use when mappings are created before their referenced
802 * sequences are instantiated, for example when parsing GFF data.
806 public void updateToDataset(SequenceI seq)
808 if (seq == null || seq.getDatasetSequence() == null)
812 SequenceI ds = seq.getDatasetSequence();
814 for (SequenceToSequenceMapping ssm : mappings)
819 if (ssm.fromSeq == seq)
827 if (ssm.mapping.to == seq)
835 * Answers true if this object contains no mappings
839 public boolean isEmpty()
841 return mappings.isEmpty();
845 * Method for debug / inspection purposes only, may change in future
848 public String toString()
850 return mappings == null ? "null" : mappings.toString();
854 * Returns the first mapping found that is between 'fromSeq' and 'toSeq', or
858 * aligned or dataset sequence
860 * aligned or dataset sequence
863 public Mapping getMappingBetween(SequenceI fromSeq, SequenceI toSeq)
865 SequenceI dssFrom = fromSeq.getDatasetSequence() == null ? fromSeq
866 : fromSeq.getDatasetSequence();
867 SequenceI dssTo = toSeq.getDatasetSequence() == null ? toSeq
868 : toSeq.getDatasetSequence();
870 for (SequenceToSequenceMapping mapping : mappings)
872 SequenceI from = mapping.fromSeq;
873 SequenceI to = mapping.mapping.to;
874 if ((from == dssFrom && to == dssTo)
875 || (from == dssTo && to == dssFrom))
877 return mapping.mapping;
884 * Returns a hashcode derived from the list of sequence mappings
886 * @see SequenceToSequenceMapping#hashCode()
887 * @see AbstractList#hashCode()
890 public int hashCode()
892 return this.mappings.hashCode();
896 * Two AlignedCodonFrame objects are equal if they hold the same ordered list
899 * @see SequenceToSequenceMapping#equals
902 public boolean equals(Object obj)
904 if (!(obj instanceof AlignedCodonFrame))
908 return this.mappings.equals(((AlignedCodonFrame) obj).mappings);
911 public List<SequenceToSequenceMapping> getMappings()
917 * Returns the first mapping found which is between the two given sequences,
918 * and covers the full extent of both.
924 public SequenceToSequenceMapping getCoveringMapping(SequenceI seq1,
927 for (SequenceToSequenceMapping mapping : mappings)
929 if (mapping.covers(seq2) && mapping.covers(seq1))