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 java.util.AbstractList;
24 import java.util.ArrayList;
25 import java.util.List;
27 import jalview.util.MapList;
28 import jalview.util.MappingUtils;
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 within 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 * Adds any regions mapped to or from position {@code pos} in sequence
212 * {@code seq} to the given search results
218 public void markMappedRegion(SequenceI seq, int pos, SearchResultsI sr)
221 SequenceI mappedSeq = null;
222 SequenceI ds = seq.getDatasetSequence();
224 if (this.fromSeq == seq || this.fromSeq == ds)
226 codon = this.mapping.map.locateInTo(pos, pos);
227 mappedSeq = this.mapping.to;
229 else if (this.mapping.to == seq || this.mapping.to == ds)
231 codon = this.mapping.map.locateInFrom(pos, pos);
232 mappedSeq = this.fromSeq;
237 for (int i = 0; i < codon.length; i += 2)
239 sr.addResult(mappedSeq, codon[i], codon[i + 1]);
245 private List<SequenceToSequenceMapping> mappings;
250 public AlignedCodonFrame()
252 mappings = new ArrayList<>();
256 * Adds a mapping between the dataset sequences for the associated dna and
257 * protein sequence objects
263 public void addMap(SequenceI dnaseq, SequenceI aaseq, MapList map)
265 addMap(dnaseq, aaseq, map, null);
269 * Adds a mapping between the dataset sequences for the associated dna and
270 * protein sequence objects
277 public void addMap(SequenceI dnaseq, SequenceI aaseq, MapList map,
280 // JBPNote DEBUG! THIS !
281 // dnaseq.transferAnnotation(aaseq, mp);
282 // aaseq.transferAnnotation(dnaseq, new Mapping(map.getInverse()));
284 SequenceI fromSeq = (dnaseq.getDatasetSequence() == null) ? dnaseq
285 : dnaseq.getDatasetSequence();
286 SequenceI toSeq = (aaseq.getDatasetSequence() == null) ? aaseq
287 : aaseq.getDatasetSequence();
290 * if we already hold a mapping between these sequences, just add to it
291 * note that 'adding' a duplicate map does nothing; this protects against
292 * creating duplicate mappings in AlignedCodonFrame
294 for (SequenceToSequenceMapping ssm : mappings)
296 if (ssm.fromSeq == fromSeq && ssm.mapping.to == toSeq)
298 ssm.mapping.map.addMapList(map);
304 * otherwise, add a new sequence mapping
306 Mapping mp = new Mapping(toSeq, map);
307 mp.setMappedFromId(mapFromId);
308 mappings.add(new SequenceToSequenceMapping(fromSeq, mp));
311 public SequenceI[] getdnaSeqs()
313 // TODO return a list instead?
315 List<SequenceI> seqs = new ArrayList<>();
316 for (SequenceToSequenceMapping ssm : mappings)
318 seqs.add(ssm.fromSeq);
320 return seqs.toArray(new SequenceI[seqs.size()]);
323 public SequenceI[] getAaSeqs()
325 // TODO not used - remove?
326 List<SequenceI> seqs = new ArrayList<>();
327 for (SequenceToSequenceMapping ssm : mappings)
329 seqs.add(ssm.mapping.to);
331 return seqs.toArray(new SequenceI[seqs.size()]);
334 public MapList[] getdnaToProt()
336 List<MapList> maps = new ArrayList<>();
337 for (SequenceToSequenceMapping ssm : mappings)
339 maps.add(ssm.mapping.map);
341 return maps.toArray(new MapList[maps.size()]);
344 public Mapping[] getProtMappings()
346 List<Mapping> maps = new ArrayList<>();
347 for (SequenceToSequenceMapping ssm : mappings)
349 maps.add(ssm.mapping);
351 return maps.toArray(new Mapping[maps.size()]);
355 * Returns the first mapping found which is to or from the given sequence, or
356 * null if none is found
361 public Mapping getMappingForSequence(SequenceI seq)
363 SequenceI seqDs = seq.getDatasetSequence();
364 seqDs = seqDs != null ? seqDs : seq;
366 for (SequenceToSequenceMapping ssm : mappings)
368 if (ssm.fromSeq == seqDs || ssm.mapping.to == seqDs)
377 * Return the corresponding aligned or dataset aa sequence for given dna
378 * sequence, null if not found.
383 public SequenceI getAaForDnaSeq(SequenceI dnaSeqRef)
385 SequenceI dnads = dnaSeqRef.getDatasetSequence();
386 for (SequenceToSequenceMapping ssm : mappings)
388 if (ssm.fromSeq == dnaSeqRef || ssm.fromSeq == dnads)
390 return ssm.mapping.to;
397 * Return the corresponding aligned or dataset dna sequence for given amino
398 * acid sequence, or null if not found. returns the sequence from the first
399 * mapping found that involves the protein sequence.
404 public SequenceI getDnaForAaSeq(SequenceI aaSeqRef)
406 SequenceI aads = aaSeqRef.getDatasetSequence();
407 for (SequenceToSequenceMapping ssm : mappings)
409 if (ssm.mapping.to == aaSeqRef || ssm.mapping.to == aads)
418 * test to see if codon frame involves seq in any way
421 * a nucleotide or protein sequence
422 * @return true if a mapping exists to or from this sequence to any translated
425 public boolean involvesSequence(SequenceI seq)
427 return getAaForDnaSeq(seq) != null || getDnaForAaSeq(seq) != null;
431 * Add search results for regions in other sequences that translate or are
432 * translated from a particular position in seq
438 * where highlighted regions go
440 public void markMappedRegion(SequenceI seq, int index,
441 SearchResultsI results)
443 SequenceI ds = seq.getDatasetSequence();
444 for (SequenceToSequenceMapping ssm : mappings)
446 ssm.markMappedRegion(ds, index, results);
451 * Returns the DNA codon positions (base 1) for the given position (base 1) in
452 * a mapped protein sequence, or null if no mapping is found.
454 * Intended for use in aligning cDNA to match aligned protein. Only the first
455 * mapping found is returned, so not suitable for use if multiple protein
456 * sequences are mapped to the same cDNA (but aligning cDNA as protein is
457 * ill-defined for this case anyway).
460 * the DNA dataset sequence
462 * residue position (base 1) in a protein sequence
465 public int[] getDnaPosition(SequenceI seq, int aaPos)
468 * Adapted from markMappedRegion().
472 for (SequenceToSequenceMapping ssm : mappings)
474 if (ssm.fromSeq == seq)
476 ml = getdnaToProt()[i];
481 return ml == null ? null : ml.locateInFrom(aaPos, aaPos);
485 * Convenience method to return the first aligned sequence in the given
486 * alignment whose dataset has a mapping with the given (aligned or dataset)
494 public SequenceI findAlignedSequence(SequenceI seq, AlignmentI al)
496 return findAlignedSequence(seq, al, null);
499 * Convenience method to return the first aligned sequence in the given
500 * alignment whose dataset has a mapping with the given (aligned or dataset)
501 * sequence, and optionally the mapping that relates them
505 * @param map - list to add the mapping to
506 * @return sequence from al that maps to seq
508 public SequenceI findAlignedSequence(SequenceI seq, AlignmentI al,List<SequenceToSequenceMapping> map)
511 * Search mapped protein ('to') sequences first.
513 for (SequenceToSequenceMapping ssm : mappings)
515 int mStart=ssm.getMapping().getMap().getFromLowest(),mEnd=ssm.getMapping().map.getFromHighest();
516 if ((ssm.fromSeq == seq || ssm.fromSeq == seq.getDatasetSequence())
517 // here AlignmentUtilsTest. testAlignProteinAsDna_incompleteStartCodon fails because mStart/mEnd is contained by seq
518 // without this filter, we don't get the correct mapping, however
519 )// && seq.getStart()>=mStart && seq.getEnd()<=mEnd)
521 for (SequenceI sourceAligned : al.getSequences())
523 if (ssm.covers(sourceAligned,true,false))
529 return sourceAligned;
536 * Then try mapped dna sequences.
538 for (SequenceToSequenceMapping ssm : mappings)
540 int mStart=ssm.getMapping().getMap().getToLowest(),mEnd=ssm.getMapping().map.getToHighest();
541 if ((ssm.mapping.to == seq
542 || ssm.mapping.to == seq.getDatasetSequence())
543 && seq.getStart()>=mStart && seq.getEnd()<=mEnd)
545 for (SequenceI sourceAligned : al.getSequences())
547 if (ssm.covers(sourceAligned,true,true))
553 return sourceAligned;
563 * Returns the region in the target sequence's dataset that is mapped to the
564 * given position (base 1) in the query sequence's dataset. The region is a
565 * set of start/end position pairs.
572 public int[] getMappedRegion(SequenceI target, SequenceI query,
575 SequenceI targetDs = target.getDatasetSequence() == null ? target
576 : target.getDatasetSequence();
577 SequenceI queryDs = query.getDatasetSequence() == null ? query
578 : query.getDatasetSequence();
579 if (targetDs == null || queryDs == null /*|| dnaToProt == null*/)
583 for (SequenceToSequenceMapping ssm : mappings)
586 * try mapping from target to query
588 if (ssm.fromSeq == targetDs && ssm.mapping.to == queryDs)
590 int[] codon = ssm.mapping.map.locateInFrom(queryPos, queryPos);
597 * else try mapping from query to target
599 else if (ssm.fromSeq == queryDs && ssm.mapping.to == targetDs)
601 int[] codon = ssm.mapping.map.locateInTo(queryPos, queryPos);
612 * Returns the mapped DNA codons for the given position in a protein sequence,
613 * or null if no mapping is found. Returns a list of (e.g.) ['g', 'c', 't']
614 * codons. There may be more than one codon mapped to the protein if (for
615 * example), there are mappings to cDNA variants.
618 * the peptide dataset sequence
620 * residue position (base 1) in the peptide sequence
623 public List<char[]> getMappedCodons(SequenceI protein, int aaPos)
626 SequenceI dnaSeq = null;
627 List<char[]> result = new ArrayList<>();
629 for (SequenceToSequenceMapping ssm : mappings)
631 if (ssm.mapping.to == protein
632 && ssm.mapping.getMap().getFromRatio() == 3)
634 ml = ssm.mapping.map;
635 dnaSeq = ssm.fromSeq;
637 int[] codonPos = ml.locateInFrom(aaPos, aaPos);
638 if (codonPos == null)
644 * Read off the mapped nucleotides (converting to position base 0)
646 codonPos = MappingUtils.flattenRanges(codonPos);
647 int start = dnaSeq.getStart();
648 char c1 = dnaSeq.getCharAt(codonPos[0] - start);
649 char c2 = dnaSeq.getCharAt(codonPos[1] - start);
650 char c3 = dnaSeq.getCharAt(codonPos[2] - start);
651 result.add(new char[] { c1, c2, c3 });
654 return result.isEmpty() ? null : result;
658 * Returns any mappings found which are from the given sequence, and to
659 * distinct sequences.
664 public List<Mapping> getMappingsFromSequence(SequenceI seq)
666 List<Mapping> result = new ArrayList<>();
667 List<SequenceI> related = new ArrayList<>();
668 SequenceI seqDs = seq.getDatasetSequence();
669 seqDs = seqDs != null ? seqDs : seq;
671 for (SequenceToSequenceMapping ssm : mappings)
673 final Mapping mapping = ssm.mapping;
674 if (ssm.fromSeq == seqDs)
676 if (!related.contains(mapping.to))
679 related.add(mapping.to);
687 * Test whether the given sequence is substitutable for one or more dummy
688 * sequences in this mapping
694 public boolean isRealisableWith(SequenceI seq)
696 return realiseWith(seq, false) > 0;
700 * Replace any matchable mapped dummy sequences with the given real one.
701 * Returns the count of sequence mappings instantiated.
706 public int realiseWith(SequenceI seq)
708 return realiseWith(seq, true);
712 * Returns the number of mapped dummy sequences that could be replaced with
713 * the given real sequence.
718 * if true, performs replacements, else only counts
721 protected int realiseWith(SequenceI seq, boolean doUpdate)
723 SequenceI ds = seq.getDatasetSequence() != null
724 ? seq.getDatasetSequence()
729 * check for replaceable DNA ('map from') sequences
731 for (SequenceToSequenceMapping ssm : mappings)
733 SequenceI dna = ssm.fromSeq;
734 if (dna instanceof SequenceDummy
735 && dna.getName().equals(ds.getName()))
737 Mapping mapping = ssm.mapping;
738 int mapStart = mapping.getMap().getFromLowest();
739 int mapEnd = mapping.getMap().getFromHighest();
740 boolean mappable = couldRealiseSequence(dna, ds, mapStart, mapEnd);
746 // TODO: new method ? ds.realise(dna);
747 // might want to copy database refs as well
748 ds.setSequenceFeatures(dna.getSequenceFeatures());
751 System.out.println("Realised mapped sequence " + ds.getName());
757 * check for replaceable protein ('map to') sequences
759 Mapping mapping = ssm.mapping;
760 SequenceI prot = mapping.getTo();
761 int mapStart = mapping.getMap().getToLowest();
762 int mapEnd = mapping.getMap().getToHighest();
763 boolean mappable = couldRealiseSequence(prot, ds, mapStart, mapEnd);
769 // TODO: new method ? ds.realise(dna);
770 // might want to copy database refs as well
771 ds.setSequenceFeatures(dna.getSequenceFeatures());
772 ssm.mapping.setTo(ds);
780 * Helper method to test whether a 'real' sequence could replace a 'dummy'
781 * sequence in the map. The criteria are that they have the same name, and
782 * that the mapped region overlaps the candidate sequence.
790 protected static boolean couldRealiseSequence(SequenceI existing,
791 SequenceI replacement, int mapStart, int mapEnd)
793 if (existing instanceof SequenceDummy
794 && !(replacement instanceof SequenceDummy)
795 && existing.getName().equals(replacement.getName()))
797 int start = replacement.getStart();
798 int end = replacement.getEnd();
799 boolean mappingOverlapsSequence = (mapStart >= start
800 && mapStart <= end) || (mapEnd >= start && mapEnd <= end);
801 if (mappingOverlapsSequence)
810 * Change any mapping to the given sequence to be to its dataset sequence
811 * instead. For use when mappings are created before their referenced
812 * sequences are instantiated, for example when parsing GFF data.
816 public void updateToDataset(SequenceI seq)
818 if (seq == null || seq.getDatasetSequence() == null)
822 SequenceI ds = seq.getDatasetSequence();
824 for (SequenceToSequenceMapping ssm : mappings)
829 if (ssm.fromSeq == seq)
837 if (ssm.mapping.to == seq)
845 * Answers true if this object contains no mappings
849 public boolean isEmpty()
851 return mappings.isEmpty();
855 * Method for debug / inspection purposes only, may change in future
858 public String toString()
860 return mappings == null ? "null" : mappings.toString();
864 * Returns the first mapping found that is between 'fromSeq' and 'toSeq', or
868 * aligned or dataset sequence
870 * aligned or dataset sequence
873 public Mapping getMappingBetween(SequenceI fromSeq, SequenceI toSeq)
875 SequenceI dssFrom = fromSeq.getDatasetSequence() == null ? fromSeq
876 : fromSeq.getDatasetSequence();
877 SequenceI dssTo = toSeq.getDatasetSequence() == null ? toSeq
878 : toSeq.getDatasetSequence();
880 for (SequenceToSequenceMapping mapping : mappings)
882 SequenceI from = mapping.fromSeq;
883 SequenceI to = mapping.mapping.to;
884 if ((from == dssFrom && to == dssTo)
885 || (from == dssTo && to == dssFrom))
887 return mapping.mapping;
894 * Returns a hashcode derived from the list of sequence mappings
896 * @see SequenceToSequenceMapping#hashCode()
897 * @see AbstractList#hashCode()
900 public int hashCode()
902 return this.mappings.hashCode();
906 * Two AlignedCodonFrame objects are equal if they hold the same ordered list
909 * @see SequenceToSequenceMapping#equals
912 public boolean equals(Object obj)
914 if (!(obj instanceof AlignedCodonFrame))
918 return this.mappings.equals(((AlignedCodonFrame) obj).mappings);
921 public List<SequenceToSequenceMapping> getMappings()
927 * Returns the first mapping found which is between the two given sequences,
928 * and covers the full extent of both.
934 public SequenceToSequenceMapping getCoveringMapping(SequenceI seq1,
937 for (SequenceToSequenceMapping mapping : mappings)
939 if (mapping.covers(seq2) && mapping.covers(seq1))
948 * Returns the first mapping found which is between the given sequence and
949 * another, is a triplet mapping (3:1 or 1:3), and covers the full extent of
950 * both sequences involved.
955 public SequenceToSequenceMapping getCoveringCodonMapping(SequenceI seq)
957 for (SequenceToSequenceMapping mapping : mappings)
959 if (mapping.getMapping().getMap().isTripletMap()
960 && mapping.covers(seq))
962 if (mapping.fromSeq == seq
963 && mapping.covers(mapping.getMapping().getTo()))
967 else if (mapping.getMapping().getTo() == seq
968 && mapping.covers(mapping.fromSeq))