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();
228 if (this.fromSeq == seq || this.fromSeq == ds)
230 codon = this.mapping.map.locateInTo(pos, pos);
231 mappedSeq = this.mapping.to;
233 else if (this.mapping.to == seq || this.mapping.to == ds)
235 codon = this.mapping.map.locateInFrom(pos, pos);
236 mappedSeq = this.fromSeq;
241 for (int i = 0; i < codon.length; i += 2)
243 sr.addResult(mappedSeq, codon[i], codon[i + 1]);
249 private List<SequenceToSequenceMapping> mappings;
254 public AlignedCodonFrame()
256 mappings = new ArrayList<>();
260 * Adds a mapping between the dataset sequences for the associated dna and
261 * protein sequence objects
267 public void addMap(SequenceI dnaseq, SequenceI aaseq, MapList map)
269 addMap(dnaseq, aaseq, map, null);
273 * Adds a mapping between the dataset sequences for the associated dna and
274 * protein sequence objects
281 public void addMap(SequenceI dnaseq, SequenceI aaseq, MapList map,
284 // JBPNote DEBUG! THIS !
285 // dnaseq.transferAnnotation(aaseq, mp);
286 // aaseq.transferAnnotation(dnaseq, new Mapping(map.getInverse()));
288 SequenceI fromSeq = (dnaseq.getDatasetSequence() == null) ? dnaseq
289 : dnaseq.getDatasetSequence();
290 SequenceI toSeq = (aaseq.getDatasetSequence() == null) ? aaseq
291 : aaseq.getDatasetSequence();
294 * if we already hold a mapping between these sequences, just add to it
295 * note that 'adding' a duplicate map does nothing; this protects against
296 * creating duplicate mappings in AlignedCodonFrame
298 for (SequenceToSequenceMapping ssm : mappings)
300 if (ssm.fromSeq == fromSeq && ssm.mapping.to == toSeq)
302 ssm.mapping.map.addMapList(map);
308 * otherwise, add a new sequence mapping
310 Mapping mp = new Mapping(toSeq, map);
311 mp.setMappedFromId(mapFromId);
312 mappings.add(new SequenceToSequenceMapping(fromSeq, mp));
315 public SequenceI[] getdnaSeqs()
317 // TODO return a list instead?
319 List<SequenceI> seqs = new ArrayList<>();
320 for (SequenceToSequenceMapping ssm : mappings)
322 seqs.add(ssm.fromSeq);
324 return seqs.toArray(new SequenceI[seqs.size()]);
327 public SequenceI[] getAaSeqs()
329 // TODO not used - remove?
330 List<SequenceI> seqs = new ArrayList<>();
331 for (SequenceToSequenceMapping ssm : mappings)
333 seqs.add(ssm.mapping.to);
335 return seqs.toArray(new SequenceI[seqs.size()]);
338 public MapList[] getdnaToProt()
340 List<MapList> maps = new ArrayList<>();
341 for (SequenceToSequenceMapping ssm : mappings)
343 maps.add(ssm.mapping.map);
345 return maps.toArray(new MapList[maps.size()]);
348 public Mapping[] getProtMappings()
350 List<Mapping> maps = new ArrayList<>();
351 for (SequenceToSequenceMapping ssm : mappings)
353 maps.add(ssm.mapping);
355 return maps.toArray(new Mapping[maps.size()]);
359 * Returns the first mapping found which is to or from the given sequence, or
360 * null if none is found
365 public Mapping getMappingForSequence(SequenceI seq)
367 SequenceI seqDs = seq.getDatasetSequence();
368 seqDs = seqDs != null ? seqDs : seq;
370 for (SequenceToSequenceMapping ssm : mappings)
372 if (ssm.fromSeq == seqDs || ssm.mapping.to == seqDs)
381 * Return the corresponding aligned or dataset aa sequence for given dna
382 * sequence, null if not found.
387 public SequenceI getAaForDnaSeq(SequenceI dnaSeqRef)
389 SequenceI dnads = dnaSeqRef.getDatasetSequence();
390 for (SequenceToSequenceMapping ssm : mappings)
392 if (ssm.fromSeq == dnaSeqRef || ssm.fromSeq == dnads)
394 return ssm.mapping.to;
401 * Return the corresponding aligned or dataset dna sequence for given amino
402 * acid sequence, or null if not found. returns the sequence from the first
403 * mapping found that involves the protein sequence.
408 public SequenceI getDnaForAaSeq(SequenceI aaSeqRef)
410 SequenceI aads = aaSeqRef.getDatasetSequence();
411 for (SequenceToSequenceMapping ssm : mappings)
413 if (ssm.mapping.to == aaSeqRef || ssm.mapping.to == aads)
422 * test to see if codon frame involves seq in any way
425 * a nucleotide or protein sequence
426 * @return true if a mapping exists to or from this sequence to any translated
429 public boolean involvesSequence(SequenceI seq)
431 return getAaForDnaSeq(seq) != null || getDnaForAaSeq(seq) != null;
435 * Add search results for regions in other sequences that translate or are
436 * translated from a particular position in seq
442 * where highlighted regions go
444 public void markMappedRegion(SequenceI seq, int index,
445 SearchResultsI results)
447 SequenceI ds = seq.getDatasetSequence();
448 for (SequenceToSequenceMapping ssm : mappings)
450 ssm.markMappedRegion(ds, index, results);
455 * Returns the DNA codon positions (base 1) for the given position (base 1) in
456 * a mapped protein sequence, or null if no mapping is found.
458 * Intended for use in aligning cDNA to match aligned protein. Only the first
459 * mapping found is returned, so not suitable for use if multiple protein
460 * sequences are mapped to the same cDNA (but aligning cDNA as protein is
461 * ill-defined for this case anyway).
464 * the DNA dataset sequence
466 * residue position (base 1) in a protein sequence
469 public int[] getDnaPosition(SequenceI seq, int aaPos)
472 * Adapted from markMappedRegion().
476 for (SequenceToSequenceMapping ssm : mappings)
478 if (ssm.fromSeq == seq)
480 ml = getdnaToProt()[i];
485 return ml == null ? null : ml.locateInFrom(aaPos, aaPos);
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)
498 public SequenceI findAlignedSequence(SequenceI seq, AlignmentI al)
500 return findAlignedSequence(seq, al, null);
503 * Convenience method to return the first aligned sequence in the given
504 * alignment whose dataset has a mapping with the given (aligned or dataset)
505 * sequence, and optionally the mapping that relates them
509 * @param map - list to add the mapping to
510 * @return sequence from al that maps to seq
512 public SequenceI findAlignedSequence(SequenceI seq, AlignmentI al,List<SequenceToSequenceMapping> map)
515 * Search mapped protein ('to') sequences first.
517 for (SequenceToSequenceMapping ssm : mappings)
519 int mStart=ssm.getMapping().getMap().getFromLowest(),mEnd=ssm.getMapping().map.getFromHighest();
520 if ((ssm.fromSeq == seq || ssm.fromSeq == seq.getDatasetSequence())
521 // here AlignmentUtilsTest. testAlignProteinAsDna_incompleteStartCodon fails because mStart/mEnd is contained by seq
522 // without this filter, we don't get the correct mapping, however
523 )// && seq.getStart()>=mStart && seq.getEnd()<=mEnd)
525 for (SequenceI sourceAligned : al.getSequences())
527 if (ssm.covers(sourceAligned,true,false))
533 return sourceAligned;
540 * Then try mapped dna sequences.
542 for (SequenceToSequenceMapping ssm : mappings)
544 int mStart=ssm.getMapping().getMap().getToLowest(),mEnd=ssm.getMapping().map.getToHighest();
545 if ((ssm.mapping.to == seq
546 || ssm.mapping.to == seq.getDatasetSequence())
547 && seq.getStart()>=mStart && seq.getEnd()<=mEnd)
549 for (SequenceI sourceAligned : al.getSequences())
551 if (ssm.covers(sourceAligned,true,true))
557 return sourceAligned;
567 * Returns the region in the target sequence's dataset that is mapped to the
568 * given position (base 1) in the query sequence's dataset. The region is a
569 * set of start/end position pairs.
576 public int[] getMappedRegion(SequenceI target, SequenceI query,
579 SequenceI targetDs = target.getDatasetSequence() == null ? target
580 : target.getDatasetSequence();
581 SequenceI queryDs = query.getDatasetSequence() == null ? query
582 : query.getDatasetSequence();
583 if (targetDs == null || queryDs == null /*|| dnaToProt == null*/)
587 for (SequenceToSequenceMapping ssm : mappings)
590 * try mapping from target to query
592 if (ssm.fromSeq == targetDs && ssm.mapping.to == queryDs)
594 int[] codon = ssm.mapping.map.locateInFrom(queryPos, queryPos);
601 * else try mapping from query to target
603 else if (ssm.fromSeq == queryDs && ssm.mapping.to == targetDs)
605 int[] codon = ssm.mapping.map.locateInTo(queryPos, queryPos);
616 * Returns the mapped DNA codons for the given position in a protein sequence,
617 * or null if no mapping is found. Returns a list of (e.g.) ['g', 'c', 't']
618 * codons. There may be more than one codon mapped to the protein if (for
619 * example), there are mappings to cDNA variants.
622 * the peptide dataset sequence
624 * residue position (base 1) in the peptide sequence
627 public List<char[]> getMappedCodons(SequenceI protein, int aaPos)
630 SequenceI dnaSeq = null;
631 List<char[]> result = new ArrayList<>();
633 for (SequenceToSequenceMapping ssm : mappings)
635 if (ssm.mapping.to == protein
636 && ssm.mapping.getMap().getFromRatio() == 3)
638 ml = ssm.mapping.map;
639 dnaSeq = ssm.fromSeq;
641 int[] codonPos = ml.locateInFrom(aaPos, aaPos);
642 if (codonPos == null)
648 * Read off the mapped nucleotides (converting to position base 0)
650 codonPos = MappingUtils.flattenRanges(codonPos);
651 int start = dnaSeq.getStart();
652 char c1 = dnaSeq.getCharAt(codonPos[0] - start);
653 char c2 = dnaSeq.getCharAt(codonPos[1] - start);
654 char c3 = dnaSeq.getCharAt(codonPos[2] - start);
655 result.add(new char[] { c1, c2, c3 });
658 return result.isEmpty() ? null : result;
662 * Returns any mappings found which are from the given sequence, and to
663 * distinct sequences.
668 public List<Mapping> getMappingsFromSequence(SequenceI seq)
670 List<Mapping> result = new ArrayList<>();
671 List<SequenceI> related = new ArrayList<>();
672 SequenceI seqDs = seq.getDatasetSequence();
673 seqDs = seqDs != null ? seqDs : seq;
675 for (SequenceToSequenceMapping ssm : mappings)
677 final Mapping mapping = ssm.mapping;
678 if (ssm.fromSeq == seqDs)
680 if (!related.contains(mapping.to))
683 related.add(mapping.to);
691 * Test whether the given sequence is substitutable for one or more dummy
692 * sequences in this mapping
698 public boolean isRealisableWith(SequenceI seq)
700 return realiseWith(seq, false) > 0;
704 * Replace any matchable mapped dummy sequences with the given real one.
705 * Returns the count of sequence mappings instantiated.
710 public int realiseWith(SequenceI seq)
712 return realiseWith(seq, true);
716 * Returns the number of mapped dummy sequences that could be replaced with
717 * the given real sequence.
722 * if true, performs replacements, else only counts
725 protected int realiseWith(SequenceI seq, boolean doUpdate)
727 SequenceI ds = seq.getDatasetSequence() != null
728 ? seq.getDatasetSequence()
733 * check for replaceable DNA ('map from') sequences
735 for (SequenceToSequenceMapping ssm : mappings)
737 SequenceI dna = ssm.fromSeq;
738 if (dna instanceof SequenceDummy
739 && dna.getName().equals(ds.getName()))
741 Mapping mapping = ssm.mapping;
742 int mapStart = mapping.getMap().getFromLowest();
743 int mapEnd = mapping.getMap().getFromHighest();
744 boolean mappable = couldRealiseSequence(dna, ds, mapStart, mapEnd);
750 // TODO: new method ? ds.realise(dna);
751 // might want to copy database refs as well
752 ds.setSequenceFeatures(dna.getSequenceFeatures());
755 System.out.println("Realised mapped sequence " + ds.getName());
761 * check for replaceable protein ('map to') sequences
763 Mapping mapping = ssm.mapping;
764 SequenceI prot = mapping.getTo();
765 int mapStart = mapping.getMap().getToLowest();
766 int mapEnd = mapping.getMap().getToHighest();
767 boolean mappable = couldRealiseSequence(prot, ds, mapStart, mapEnd);
773 // TODO: new method ? ds.realise(dna);
774 // might want to copy database refs as well
775 ds.setSequenceFeatures(dna.getSequenceFeatures());
776 ssm.mapping.setTo(ds);
784 * Helper method to test whether a 'real' sequence could replace a 'dummy'
785 * sequence in the map. The criteria are that they have the same name, and
786 * that the mapped region overlaps the candidate sequence.
794 protected static boolean couldRealiseSequence(SequenceI existing,
795 SequenceI replacement, int mapStart, int mapEnd)
797 if (existing instanceof SequenceDummy
798 && !(replacement instanceof SequenceDummy)
799 && existing.getName().equals(replacement.getName()))
801 int start = replacement.getStart();
802 int end = replacement.getEnd();
803 boolean mappingOverlapsSequence = (mapStart >= start
804 && mapStart <= end) || (mapEnd >= start && mapEnd <= end);
805 if (mappingOverlapsSequence)
814 * Change any mapping to the given sequence to be to its dataset sequence
815 * instead. For use when mappings are created before their referenced
816 * sequences are instantiated, for example when parsing GFF data.
820 public void updateToDataset(SequenceI seq)
822 if (seq == null || seq.getDatasetSequence() == null)
826 SequenceI ds = seq.getDatasetSequence();
828 for (SequenceToSequenceMapping ssm : mappings)
833 if (ssm.fromSeq == seq)
841 if (ssm.mapping.to == seq)
849 * Answers true if this object contains no mappings
853 public boolean isEmpty()
855 return mappings.isEmpty();
859 * Method for debug / inspection purposes only, may change in future
862 public String toString()
864 return mappings == null ? "null" : mappings.toString();
868 * Returns the first mapping found that is between 'fromSeq' and 'toSeq', or
872 * aligned or dataset sequence
874 * aligned or dataset sequence
877 public Mapping getMappingBetween(SequenceI fromSeq, SequenceI toSeq)
879 SequenceI dssFrom = fromSeq.getDatasetSequence() == null ? fromSeq
880 : fromSeq.getDatasetSequence();
881 SequenceI dssTo = toSeq.getDatasetSequence() == null ? toSeq
882 : toSeq.getDatasetSequence();
884 for (SequenceToSequenceMapping mapping : mappings)
886 SequenceI from = mapping.fromSeq;
887 SequenceI to = mapping.mapping.to;
888 if ((from == dssFrom && to == dssTo)
889 || (from == dssTo && to == dssFrom))
891 return mapping.mapping;
898 * Returns a hashcode derived from the list of sequence mappings
900 * @see SequenceToSequenceMapping#hashCode()
901 * @see AbstractList#hashCode()
904 public int hashCode()
906 return this.mappings.hashCode();
910 * Two AlignedCodonFrame objects are equal if they hold the same ordered list
913 * @see SequenceToSequenceMapping#equals
916 public boolean equals(Object obj)
918 if (!(obj instanceof AlignedCodonFrame))
922 return this.mappings.equals(((AlignedCodonFrame) obj).mappings);
925 public List<SequenceToSequenceMapping> getMappings()
931 * Returns the first mapping found which is between the two given sequences,
932 * and covers the full extent of both.
938 public SequenceToSequenceMapping getCoveringMapping(SequenceI seq1,
941 for (SequenceToSequenceMapping mapping : mappings)
943 if (mapping.covers(seq2) && mapping.covers(seq1))
952 * Returns the first mapping found which is between the given sequence and
953 * another, is a triplet mapping (3:1 or 1:3), and covers the full extent of
954 * both sequences involved.
959 public SequenceToSequenceMapping getCoveringCodonMapping(SequenceI seq)
961 for (SequenceToSequenceMapping mapping : mappings)
963 if (mapping.getMapping().getMap().isTripletMap()
964 && mapping.covers(seq))
966 if (mapping.fromSeq == seq
967 && mapping.covers(mapping.getMapping().getTo()))
971 else if (mapping.getMapping().getTo() == seq
972 && mapping.covers(mapping.fromSeq))