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 List<int[]> mappedRanges = null;
122 MapList mapList = mapping.getMap();
123 if (fromSeq == seq || fromSeq == seq.getDatasetSequence())
125 mappedRanges = mapList.getFromRanges();
127 else if (mapping.to == seq || mapping.to == seq.getDatasetSequence())
129 mappedRanges = mapList.getToRanges();
137 * check that each mapped range lies within the sequence range
138 * (necessary for circular CDS - example EMBL:J03321:AAA91567)
139 * and mapped length covers (at least) sequence length
142 for (int[] range : mappedRanges)
144 int from = Math.min(range[0], range[1]);
145 int to = Math.max(range[0], range[1]);
146 if (from < seq.getStart() || to > seq.getEnd())
150 length += (to - from + 1);
152 // add 1 to mapped length to allow for a mapped stop codon
153 if (length + 1 < (seq.getEnd() - seq.getStart() + 1))
161 * Adds any regions mapped to or from position {@code pos} in sequence
162 * {@code seq} to the given search results
168 public void markMappedRegion(SequenceI seq, int pos, SearchResultsI sr)
171 SequenceI mappedSeq = null;
172 SequenceI ds = seq.getDatasetSequence();
174 if (this.fromSeq == seq || this.fromSeq == ds)
176 codon = this.mapping.map.locateInTo(pos, pos);
177 mappedSeq = this.mapping.to;
179 else if (this.mapping.to == seq || this.mapping.to == ds)
181 codon = this.mapping.map.locateInFrom(pos, pos);
182 mappedSeq = this.fromSeq;
187 for (int i = 0; i < codon.length; i += 2)
189 sr.addResult(mappedSeq, codon[i], codon[i + 1]);
195 private List<SequenceToSequenceMapping> mappings;
200 public AlignedCodonFrame()
202 mappings = new ArrayList<>();
206 * Adds a mapping between the dataset sequences for the associated dna and
207 * protein sequence objects
213 public void addMap(SequenceI dnaseq, SequenceI aaseq, MapList map)
215 addMap(dnaseq, aaseq, map, null);
219 * Adds a mapping between the dataset sequences for the associated dna and
220 * protein sequence objects
227 public void addMap(SequenceI dnaseq, SequenceI aaseq, MapList map,
230 // JBPNote DEBUG! THIS !
231 // dnaseq.transferAnnotation(aaseq, mp);
232 // aaseq.transferAnnotation(dnaseq, new Mapping(map.getInverse()));
234 SequenceI fromSeq = (dnaseq.getDatasetSequence() == null) ? dnaseq
235 : dnaseq.getDatasetSequence();
236 SequenceI toSeq = (aaseq.getDatasetSequence() == null) ? aaseq
237 : aaseq.getDatasetSequence();
240 * if we already hold a mapping between these sequences, just add to it
241 * note that 'adding' a duplicate map does nothing; this protects against
242 * creating duplicate mappings in AlignedCodonFrame
244 for (SequenceToSequenceMapping ssm : mappings)
246 if (ssm.fromSeq == fromSeq && ssm.mapping.to == toSeq)
248 ssm.mapping.map.addMapList(map);
254 * otherwise, add a new sequence mapping
256 Mapping mp = new Mapping(toSeq, map);
257 mp.setMappedFromId(mapFromId);
258 mappings.add(new SequenceToSequenceMapping(fromSeq, mp));
261 public SequenceI[] getdnaSeqs()
263 // TODO return a list instead?
265 List<SequenceI> seqs = new ArrayList<>();
266 for (SequenceToSequenceMapping ssm : mappings)
268 seqs.add(ssm.fromSeq);
270 return seqs.toArray(new SequenceI[seqs.size()]);
273 public SequenceI[] getAaSeqs()
275 // TODO not used - remove?
276 List<SequenceI> seqs = new ArrayList<>();
277 for (SequenceToSequenceMapping ssm : mappings)
279 seqs.add(ssm.mapping.to);
281 return seqs.toArray(new SequenceI[seqs.size()]);
284 public MapList[] getdnaToProt()
286 List<MapList> maps = new ArrayList<>();
287 for (SequenceToSequenceMapping ssm : mappings)
289 maps.add(ssm.mapping.map);
291 return maps.toArray(new MapList[maps.size()]);
294 public Mapping[] getProtMappings()
296 List<Mapping> maps = new ArrayList<>();
297 for (SequenceToSequenceMapping ssm : mappings)
299 maps.add(ssm.mapping);
301 return maps.toArray(new Mapping[maps.size()]);
305 * Returns the first mapping found which is to or from the given sequence, or
306 * null if none is found
311 public Mapping getMappingForSequence(SequenceI seq)
313 SequenceI seqDs = seq.getDatasetSequence();
314 seqDs = seqDs != null ? seqDs : seq;
316 for (SequenceToSequenceMapping ssm : mappings)
318 if (ssm.fromSeq == seqDs || ssm.mapping.to == seqDs)
327 * Return the corresponding aligned or dataset aa sequence for given dna
328 * sequence, null if not found.
333 public SequenceI getAaForDnaSeq(SequenceI dnaSeqRef)
335 SequenceI dnads = dnaSeqRef.getDatasetSequence();
336 for (SequenceToSequenceMapping ssm : mappings)
338 if (ssm.fromSeq == dnaSeqRef || ssm.fromSeq == dnads)
340 return ssm.mapping.to;
347 * Return the corresponding aligned or dataset dna sequence for given amino
348 * acid sequence, or null if not found. returns the sequence from the first
349 * mapping found that involves the protein sequence.
354 public SequenceI getDnaForAaSeq(SequenceI aaSeqRef)
356 SequenceI aads = aaSeqRef.getDatasetSequence();
357 for (SequenceToSequenceMapping ssm : mappings)
359 if (ssm.mapping.to == aaSeqRef || ssm.mapping.to == aads)
368 * test to see if codon frame involves seq in any way
371 * a nucleotide or protein sequence
372 * @return true if a mapping exists to or from this sequence to any translated
375 public boolean involvesSequence(SequenceI seq)
377 return getAaForDnaSeq(seq) != null || getDnaForAaSeq(seq) != null;
381 * Add search results for regions in other sequences that translate or are
382 * translated from a particular position in seq
388 * where highlighted regions go
390 public void markMappedRegion(SequenceI seq, int index,
391 SearchResultsI results)
393 SequenceI ds = seq.getDatasetSequence();
394 for (SequenceToSequenceMapping ssm : mappings)
396 ssm.markMappedRegion(ds, index, results);
401 * Returns the DNA codon positions (base 1) for the given position (base 1) in
402 * a mapped protein sequence, or null if no mapping is found.
404 * Intended for use in aligning cDNA to match aligned protein. Only the first
405 * mapping found is returned, so not suitable for use if multiple protein
406 * sequences are mapped to the same cDNA (but aligning cDNA as protein is
407 * ill-defined for this case anyway).
410 * the DNA dataset sequence
412 * residue position (base 1) in a protein sequence
415 public int[] getDnaPosition(SequenceI seq, int aaPos)
418 * Adapted from markMappedRegion().
422 for (SequenceToSequenceMapping ssm : mappings)
424 if (ssm.fromSeq == seq)
426 ml = getdnaToProt()[i];
431 return ml == null ? null : ml.locateInFrom(aaPos, aaPos);
435 * Convenience method to return the first aligned sequence in the given
436 * alignment whose dataset has a mapping with the given (aligned or dataset)
444 public SequenceI findAlignedSequence(SequenceI seq, AlignmentI al)
447 * Search mapped protein ('to') sequences first.
449 for (SequenceToSequenceMapping ssm : mappings)
451 if (ssm.fromSeq == seq || ssm.fromSeq == seq.getDatasetSequence())
453 for (SequenceI sourceAligned : al.getSequences())
455 if (ssm.mapping.to == sourceAligned.getDatasetSequence()
456 || ssm.mapping.to == sourceAligned)
458 return sourceAligned;
465 * Then try mapped dna sequences.
467 for (SequenceToSequenceMapping ssm : mappings)
469 if (ssm.mapping.to == seq
470 || ssm.mapping.to == seq.getDatasetSequence())
472 for (SequenceI sourceAligned : al.getSequences())
474 if (ssm.fromSeq == sourceAligned.getDatasetSequence())
476 return sourceAligned;
486 * Returns the region in the target sequence's dataset that is mapped to the
487 * given position (base 1) in the query sequence's dataset. The region is a
488 * set of start/end position pairs.
495 public int[] getMappedRegion(SequenceI target, SequenceI query,
498 SequenceI targetDs = target.getDatasetSequence() == null ? target
499 : target.getDatasetSequence();
500 SequenceI queryDs = query.getDatasetSequence() == null ? query
501 : query.getDatasetSequence();
502 if (targetDs == null || queryDs == null /*|| dnaToProt == null*/)
506 for (SequenceToSequenceMapping ssm : mappings)
509 * try mapping from target to query
511 if (ssm.fromSeq == targetDs && ssm.mapping.to == queryDs)
513 int[] codon = ssm.mapping.map.locateInFrom(queryPos, queryPos);
520 * else try mapping from query to target
522 else if (ssm.fromSeq == queryDs && ssm.mapping.to == targetDs)
524 int[] codon = ssm.mapping.map.locateInTo(queryPos, queryPos);
535 * Returns the mapped DNA codons for the given position in a protein sequence,
536 * or null if no mapping is found. Returns a list of (e.g.) ['g', 'c', 't']
537 * codons. There may be more than one codon mapped to the protein if (for
538 * example), there are mappings to cDNA variants.
541 * the peptide dataset sequence
543 * residue position (base 1) in the peptide sequence
546 public List<char[]> getMappedCodons(SequenceI protein, int aaPos)
549 SequenceI dnaSeq = null;
550 List<char[]> result = new ArrayList<>();
552 for (SequenceToSequenceMapping ssm : mappings)
554 if (ssm.mapping.to == protein
555 && ssm.mapping.getMap().getFromRatio() == 3)
557 ml = ssm.mapping.map;
558 dnaSeq = ssm.fromSeq;
560 int[] codonPos = ml.locateInFrom(aaPos, aaPos);
561 if (codonPos == null)
567 * Read off the mapped nucleotides (converting to position base 0)
569 codonPos = MappingUtils.flattenRanges(codonPos);
570 int start = dnaSeq.getStart();
571 char c1 = dnaSeq.getCharAt(codonPos[0] - start);
572 char c2 = dnaSeq.getCharAt(codonPos[1] - start);
573 char c3 = dnaSeq.getCharAt(codonPos[2] - start);
574 result.add(new char[] { c1, c2, c3 });
577 return result.isEmpty() ? null : result;
581 * Returns any mappings found which are from the given sequence, and to
582 * distinct sequences.
587 public List<Mapping> getMappingsFromSequence(SequenceI seq)
589 List<Mapping> result = new ArrayList<>();
590 List<SequenceI> related = new ArrayList<>();
591 SequenceI seqDs = seq.getDatasetSequence();
592 seqDs = seqDs != null ? seqDs : seq;
594 for (SequenceToSequenceMapping ssm : mappings)
596 final Mapping mapping = ssm.mapping;
597 if (ssm.fromSeq == seqDs)
599 if (!related.contains(mapping.to))
602 related.add(mapping.to);
610 * Test whether the given sequence is substitutable for one or more dummy
611 * sequences in this mapping
617 public boolean isRealisableWith(SequenceI seq)
619 return realiseWith(seq, false) > 0;
623 * Replace any matchable mapped dummy sequences with the given real one.
624 * Returns the count of sequence mappings instantiated.
629 public int realiseWith(SequenceI seq)
631 return realiseWith(seq, true);
635 * Returns the number of mapped dummy sequences that could be replaced with
636 * the given real sequence.
641 * if true, performs replacements, else only counts
644 protected int realiseWith(SequenceI seq, boolean doUpdate)
646 SequenceI ds = seq.getDatasetSequence() != null
647 ? seq.getDatasetSequence()
652 * check for replaceable DNA ('map from') sequences
654 for (SequenceToSequenceMapping ssm : mappings)
656 SequenceI dna = ssm.fromSeq;
657 if (dna instanceof SequenceDummy
658 && dna.getName().equals(ds.getName()))
660 Mapping mapping = ssm.mapping;
661 int mapStart = mapping.getMap().getFromLowest();
662 int mapEnd = mapping.getMap().getFromHighest();
663 boolean mappable = couldRealiseSequence(dna, ds, mapStart, mapEnd);
669 // TODO: new method ? ds.realise(dna);
670 // might want to copy database refs as well
671 ds.setSequenceFeatures(dna.getSequenceFeatures());
674 System.out.println("Realised mapped sequence " + ds.getName());
680 * check for replaceable protein ('map to') sequences
682 Mapping mapping = ssm.mapping;
683 SequenceI prot = mapping.getTo();
684 int mapStart = mapping.getMap().getToLowest();
685 int mapEnd = mapping.getMap().getToHighest();
686 boolean mappable = couldRealiseSequence(prot, ds, mapStart, mapEnd);
692 // TODO: new method ? ds.realise(dna);
693 // might want to copy database refs as well
694 ds.setSequenceFeatures(dna.getSequenceFeatures());
695 ssm.mapping.setTo(ds);
703 * Helper method to test whether a 'real' sequence could replace a 'dummy'
704 * sequence in the map. The criteria are that they have the same name, and
705 * that the mapped region overlaps the candidate sequence.
713 protected static boolean couldRealiseSequence(SequenceI existing,
714 SequenceI replacement, int mapStart, int mapEnd)
716 if (existing instanceof SequenceDummy
717 && !(replacement instanceof SequenceDummy)
718 && existing.getName().equals(replacement.getName()))
720 int start = replacement.getStart();
721 int end = replacement.getEnd();
722 boolean mappingOverlapsSequence = (mapStart >= start
723 && mapStart <= end) || (mapEnd >= start && mapEnd <= end);
724 if (mappingOverlapsSequence)
733 * Change any mapping to the given sequence to be to its dataset sequence
734 * instead. For use when mappings are created before their referenced
735 * sequences are instantiated, for example when parsing GFF data.
739 public void updateToDataset(SequenceI seq)
741 if (seq == null || seq.getDatasetSequence() == null)
745 SequenceI ds = seq.getDatasetSequence();
747 for (SequenceToSequenceMapping ssm : mappings)
752 if (ssm.fromSeq == seq)
760 if (ssm.mapping.to == seq)
768 * Answers true if this object contains no mappings
772 public boolean isEmpty()
774 return mappings.isEmpty();
778 * Method for debug / inspection purposes only, may change in future
781 public String toString()
783 return mappings == null ? "null" : mappings.toString();
787 * Returns the first mapping found that is between 'fromSeq' and 'toSeq', or
791 * aligned or dataset sequence
793 * aligned or dataset sequence
796 public Mapping getMappingBetween(SequenceI fromSeq, SequenceI toSeq)
798 SequenceI dssFrom = fromSeq.getDatasetSequence() == null ? fromSeq
799 : fromSeq.getDatasetSequence();
800 SequenceI dssTo = toSeq.getDatasetSequence() == null ? toSeq
801 : toSeq.getDatasetSequence();
803 for (SequenceToSequenceMapping mapping : mappings)
805 SequenceI from = mapping.fromSeq;
806 SequenceI to = mapping.mapping.to;
807 if ((from == dssFrom && to == dssTo)
808 || (from == dssTo && to == dssFrom))
810 return mapping.mapping;
817 * Returns a hashcode derived from the list of sequence mappings
819 * @see SequenceToSequenceMapping#hashCode()
820 * @see AbstractList#hashCode()
823 public int hashCode()
825 return this.mappings.hashCode();
829 * Two AlignedCodonFrame objects are equal if they hold the same ordered list
832 * @see SequenceToSequenceMapping#equals
835 public boolean equals(Object obj)
837 if (!(obj instanceof AlignedCodonFrame))
841 return this.mappings.equals(((AlignedCodonFrame) obj).mappings);
844 public List<SequenceToSequenceMapping> getMappings()
850 * Returns the first mapping found which is between the two given sequences,
851 * and covers the full extent of both.
857 public SequenceToSequenceMapping getCoveringMapping(SequenceI seq1,
860 for (SequenceToSequenceMapping mapping : mappings)
862 if (mapping.covers(seq2) && mapping.covers(seq1))
871 * Returns the first mapping found which is between the given sequence and
872 * another, is a triplet mapping (3:1 or 1:3), and covers the full extent of
873 * both sequences involved.
878 public SequenceToSequenceMapping getCoveringCodonMapping(SequenceI seq)
880 for (SequenceToSequenceMapping mapping : mappings)
882 if (mapping.getMapping().getMap().isTripletMap()
883 && mapping.covers(seq))
885 if (mapping.fromSeq == seq
886 && mapping.covers(mapping.getMapping().getTo()))
890 else if (mapping.getMapping().getTo() == seq
891 && mapping.covers(mapping.fromSeq))