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();
178 if (this.fromSeq == seq || this.fromSeq == ds)
180 codon = this.mapping.map.locateInTo(pos, pos);
181 mappedSeq = this.mapping.to;
183 else if (this.mapping.to == seq || this.mapping.to == ds)
185 codon = this.mapping.map.locateInFrom(pos, pos);
186 mappedSeq = this.fromSeq;
191 for (int i = 0; i < codon.length; i += 2)
193 sr.addResult(mappedSeq, codon[i], codon[i + 1]);
199 private List<SequenceToSequenceMapping> mappings;
204 public AlignedCodonFrame()
206 mappings = new ArrayList<>();
210 * Adds a mapping between the dataset sequences for the associated dna and
211 * protein sequence objects
217 public void addMap(SequenceI dnaseq, SequenceI aaseq, MapList map)
219 addMap(dnaseq, aaseq, map, null);
223 * Adds a mapping between the dataset sequences for the associated dna and
224 * protein sequence objects
231 public void addMap(SequenceI dnaseq, SequenceI aaseq, MapList map,
234 // JBPNote DEBUG! THIS !
235 // dnaseq.transferAnnotation(aaseq, mp);
236 // aaseq.transferAnnotation(dnaseq, new Mapping(map.getInverse()));
238 SequenceI fromSeq = (dnaseq.getDatasetSequence() == null) ? dnaseq
239 : dnaseq.getDatasetSequence();
240 SequenceI toSeq = (aaseq.getDatasetSequence() == null) ? aaseq
241 : aaseq.getDatasetSequence();
244 * if we already hold a mapping between these sequences, just add to it
245 * note that 'adding' a duplicate map does nothing; this protects against
246 * creating duplicate mappings in AlignedCodonFrame
248 for (SequenceToSequenceMapping ssm : mappings)
250 if (ssm.fromSeq == fromSeq && ssm.mapping.to == toSeq)
252 ssm.mapping.map.addMapList(map);
258 * otherwise, add a new sequence mapping
260 Mapping mp = new Mapping(toSeq, map);
261 mp.setMappedFromId(mapFromId);
262 mappings.add(new SequenceToSequenceMapping(fromSeq, mp));
265 public SequenceI[] getdnaSeqs()
267 // TODO return a list instead?
269 List<SequenceI> seqs = new ArrayList<>();
270 for (SequenceToSequenceMapping ssm : mappings)
272 seqs.add(ssm.fromSeq);
274 return seqs.toArray(new SequenceI[seqs.size()]);
277 public SequenceI[] getAaSeqs()
279 // TODO not used - remove?
280 List<SequenceI> seqs = new ArrayList<>();
281 for (SequenceToSequenceMapping ssm : mappings)
283 seqs.add(ssm.mapping.to);
285 return seqs.toArray(new SequenceI[seqs.size()]);
288 public MapList[] getdnaToProt()
290 List<MapList> maps = new ArrayList<>();
291 for (SequenceToSequenceMapping ssm : mappings)
293 maps.add(ssm.mapping.map);
295 return maps.toArray(new MapList[maps.size()]);
298 public Mapping[] getProtMappings()
300 List<Mapping> maps = new ArrayList<>();
301 for (SequenceToSequenceMapping ssm : mappings)
303 maps.add(ssm.mapping);
305 return maps.toArray(new Mapping[maps.size()]);
309 * Returns the first mapping found which is to or from the given sequence, or
310 * null if none is found
315 public Mapping getMappingForSequence(SequenceI seq)
317 SequenceI seqDs = seq.getDatasetSequence();
318 seqDs = seqDs != null ? seqDs : seq;
320 for (SequenceToSequenceMapping ssm : mappings)
322 if (ssm.fromSeq == seqDs || ssm.mapping.to == seqDs)
331 * Return the corresponding aligned or dataset aa sequence for given dna
332 * sequence, null if not found.
337 public SequenceI getAaForDnaSeq(SequenceI dnaSeqRef)
339 SequenceI dnads = dnaSeqRef.getDatasetSequence();
340 for (SequenceToSequenceMapping ssm : mappings)
342 if (ssm.fromSeq == dnaSeqRef || ssm.fromSeq == dnads)
344 return ssm.mapping.to;
351 * Return the corresponding aligned or dataset dna sequence for given amino
352 * acid sequence, or null if not found. returns the sequence from the first
353 * mapping found that involves the protein sequence.
358 public SequenceI getDnaForAaSeq(SequenceI aaSeqRef)
360 SequenceI aads = aaSeqRef.getDatasetSequence();
361 for (SequenceToSequenceMapping ssm : mappings)
363 if (ssm.mapping.to == aaSeqRef || ssm.mapping.to == aads)
372 * test to see if codon frame involves seq in any way
375 * a nucleotide or protein sequence
376 * @return true if a mapping exists to or from this sequence to any translated
379 public boolean involvesSequence(SequenceI seq)
381 return getAaForDnaSeq(seq) != null || getDnaForAaSeq(seq) != null;
385 * Add search results for regions in other sequences that translate or are
386 * translated from a particular position in seq
392 * where highlighted regions go
394 public void markMappedRegion(SequenceI seq, int index,
395 SearchResultsI results)
397 SequenceI ds = seq.getDatasetSequence();
398 for (SequenceToSequenceMapping ssm : mappings)
400 ssm.markMappedRegion(ds, index, results);
405 * Returns the DNA codon positions (base 1) for the given position (base 1) in
406 * a mapped protein sequence, or null if no mapping is found.
408 * Intended for use in aligning cDNA to match aligned protein. Only the first
409 * mapping found is returned, so not suitable for use if multiple protein
410 * sequences are mapped to the same cDNA (but aligning cDNA as protein is
411 * ill-defined for this case anyway).
414 * the DNA dataset sequence
416 * residue position (base 1) in a protein sequence
419 public int[] getDnaPosition(SequenceI seq, int aaPos)
422 * Adapted from markMappedRegion().
426 for (SequenceToSequenceMapping ssm : mappings)
428 if (ssm.fromSeq == seq)
430 ml = getdnaToProt()[i];
435 return ml == null ? null : ml.locateInFrom(aaPos, aaPos);
439 * Convenience method to return the first aligned sequence in the given
440 * alignment whose dataset has a mapping with the given (aligned or dataset)
448 public SequenceI findAlignedSequence(SequenceI seq, AlignmentI al)
451 * Search mapped protein ('to') sequences first.
453 for (SequenceToSequenceMapping ssm : mappings)
455 if (ssm.fromSeq == seq || ssm.fromSeq == seq.getDatasetSequence())
457 for (SequenceI sourceAligned : al.getSequences())
459 if (ssm.mapping.to == sourceAligned.getDatasetSequence()
460 || ssm.mapping.to == sourceAligned)
462 return sourceAligned;
469 * Then try mapped dna sequences.
471 for (SequenceToSequenceMapping ssm : mappings)
473 if (ssm.mapping.to == seq
474 || ssm.mapping.to == seq.getDatasetSequence())
476 for (SequenceI sourceAligned : al.getSequences())
478 if (ssm.fromSeq == sourceAligned.getDatasetSequence())
480 return sourceAligned;
490 * Returns the region in the target sequence's dataset that is mapped to the
491 * given position (base 1) in the query sequence's dataset. The region is a
492 * set of start/end position pairs.
499 public int[] getMappedRegion(SequenceI target, SequenceI query,
502 SequenceI targetDs = target.getDatasetSequence() == null ? target
503 : target.getDatasetSequence();
504 SequenceI queryDs = query.getDatasetSequence() == null ? query
505 : query.getDatasetSequence();
506 if (targetDs == null || queryDs == null /*|| dnaToProt == null*/)
510 for (SequenceToSequenceMapping ssm : mappings)
513 * try mapping from target to query
515 if (ssm.fromSeq == targetDs && ssm.mapping.to == queryDs)
517 int[] codon = ssm.mapping.map.locateInFrom(queryPos, queryPos);
524 * else try mapping from query to target
526 else if (ssm.fromSeq == queryDs && ssm.mapping.to == targetDs)
528 int[] codon = ssm.mapping.map.locateInTo(queryPos, queryPos);
539 * Returns the mapped DNA codons for the given position in a protein sequence,
540 * or null if no mapping is found. Returns a list of (e.g.) ['g', 'c', 't']
541 * codons. There may be more than one codon mapped to the protein if (for
542 * example), there are mappings to cDNA variants.
545 * the peptide dataset sequence
547 * residue position (base 1) in the peptide sequence
550 public List<char[]> getMappedCodons(SequenceI protein, int aaPos)
553 SequenceI dnaSeq = null;
554 List<char[]> result = new ArrayList<>();
556 for (SequenceToSequenceMapping ssm : mappings)
558 if (ssm.mapping.to == protein
559 && ssm.mapping.getMap().getFromRatio() == 3)
561 ml = ssm.mapping.map;
562 dnaSeq = ssm.fromSeq;
564 int[] codonPos = ml.locateInFrom(aaPos, aaPos);
565 if (codonPos == null)
571 * Read off the mapped nucleotides (converting to position base 0)
573 codonPos = MappingUtils.flattenRanges(codonPos);
574 int start = dnaSeq.getStart();
575 char c1 = dnaSeq.getCharAt(codonPos[0] - start);
576 char c2 = dnaSeq.getCharAt(codonPos[1] - start);
577 char c3 = dnaSeq.getCharAt(codonPos[2] - start);
578 result.add(new char[] { c1, c2, c3 });
581 return result.isEmpty() ? null : result;
585 * Returns any mappings found which are from the given sequence, and to
586 * distinct sequences.
591 public List<Mapping> getMappingsFromSequence(SequenceI seq)
593 List<Mapping> result = new ArrayList<>();
594 List<SequenceI> related = new ArrayList<>();
595 SequenceI seqDs = seq.getDatasetSequence();
596 seqDs = seqDs != null ? seqDs : seq;
598 for (SequenceToSequenceMapping ssm : mappings)
600 final Mapping mapping = ssm.mapping;
601 if (ssm.fromSeq == seqDs)
603 if (!related.contains(mapping.to))
606 related.add(mapping.to);
614 * Test whether the given sequence is substitutable for one or more dummy
615 * sequences in this mapping
621 public boolean isRealisableWith(SequenceI seq)
623 return realiseWith(seq, false) > 0;
627 * Replace any matchable mapped dummy sequences with the given real one.
628 * Returns the count of sequence mappings instantiated.
633 public int realiseWith(SequenceI seq)
635 return realiseWith(seq, true);
639 * Returns the number of mapped dummy sequences that could be replaced with
640 * the given real sequence.
645 * if true, performs replacements, else only counts
648 protected int realiseWith(SequenceI seq, boolean doUpdate)
650 SequenceI ds = seq.getDatasetSequence() != null
651 ? seq.getDatasetSequence()
656 * check for replaceable DNA ('map from') sequences
658 for (SequenceToSequenceMapping ssm : mappings)
660 SequenceI dna = ssm.fromSeq;
661 if (dna instanceof SequenceDummy
662 && dna.getName().equals(ds.getName()))
664 Mapping mapping = ssm.mapping;
665 int mapStart = mapping.getMap().getFromLowest();
666 int mapEnd = mapping.getMap().getFromHighest();
667 boolean mappable = couldRealiseSequence(dna, ds, mapStart, mapEnd);
673 // TODO: new method ? ds.realise(dna);
674 // might want to copy database refs as well
675 ds.setSequenceFeatures(dna.getSequenceFeatures());
678 System.out.println("Realised mapped sequence " + ds.getName());
684 * check for replaceable protein ('map to') sequences
686 Mapping mapping = ssm.mapping;
687 SequenceI prot = mapping.getTo();
688 int mapStart = mapping.getMap().getToLowest();
689 int mapEnd = mapping.getMap().getToHighest();
690 boolean mappable = couldRealiseSequence(prot, ds, mapStart, mapEnd);
696 // TODO: new method ? ds.realise(dna);
697 // might want to copy database refs as well
698 ds.setSequenceFeatures(dna.getSequenceFeatures());
699 ssm.mapping.setTo(ds);
707 * Helper method to test whether a 'real' sequence could replace a 'dummy'
708 * sequence in the map. The criteria are that they have the same name, and
709 * that the mapped region overlaps the candidate sequence.
717 protected static boolean couldRealiseSequence(SequenceI existing,
718 SequenceI replacement, int mapStart, int mapEnd)
720 if (existing instanceof SequenceDummy
721 && !(replacement instanceof SequenceDummy)
722 && existing.getName().equals(replacement.getName()))
724 int start = replacement.getStart();
725 int end = replacement.getEnd();
726 boolean mappingOverlapsSequence = (mapStart >= start
727 && mapStart <= end) || (mapEnd >= start && mapEnd <= end);
728 if (mappingOverlapsSequence)
737 * Change any mapping to the given sequence to be to its dataset sequence
738 * instead. For use when mappings are created before their referenced
739 * sequences are instantiated, for example when parsing GFF data.
743 public void updateToDataset(SequenceI seq)
745 if (seq == null || seq.getDatasetSequence() == null)
749 SequenceI ds = seq.getDatasetSequence();
751 for (SequenceToSequenceMapping ssm : mappings)
756 if (ssm.fromSeq == seq)
764 if (ssm.mapping.to == seq)
772 * Answers true if this object contains no mappings
776 public boolean isEmpty()
778 return mappings.isEmpty();
782 * Method for debug / inspection purposes only, may change in future
785 public String toString()
787 return mappings == null ? "null" : mappings.toString();
791 * Returns the first mapping found that is between 'fromSeq' and 'toSeq', or
795 * aligned or dataset sequence
797 * aligned or dataset sequence
800 public Mapping getMappingBetween(SequenceI fromSeq, SequenceI toSeq)
802 SequenceI dssFrom = fromSeq.getDatasetSequence() == null ? fromSeq
803 : fromSeq.getDatasetSequence();
804 SequenceI dssTo = toSeq.getDatasetSequence() == null ? toSeq
805 : toSeq.getDatasetSequence();
807 for (SequenceToSequenceMapping mapping : mappings)
809 SequenceI from = mapping.fromSeq;
810 SequenceI to = mapping.mapping.to;
811 if ((from == dssFrom && to == dssTo)
812 || (from == dssTo && to == dssFrom))
814 return mapping.mapping;
821 * Returns a hashcode derived from the list of sequence mappings
823 * @see SequenceToSequenceMapping#hashCode()
824 * @see AbstractList#hashCode()
827 public int hashCode()
829 return this.mappings.hashCode();
833 * Two AlignedCodonFrame objects are equal if they hold the same ordered list
836 * @see SequenceToSequenceMapping#equals
839 public boolean equals(Object obj)
841 if (!(obj instanceof AlignedCodonFrame))
845 return this.mappings.equals(((AlignedCodonFrame) obj).mappings);
848 public List<SequenceToSequenceMapping> getMappings()
854 * Returns the first mapping found which is between the two given sequences,
855 * and covers the full extent of both.
861 public SequenceToSequenceMapping getCoveringMapping(SequenceI seq1,
864 for (SequenceToSequenceMapping mapping : mappings)
866 if (mapping.covers(seq2) && mapping.covers(seq1))
875 * Returns the first mapping found which is between the given sequence and
876 * another, is a triplet mapping (3:1 or 1:3), and covers the full extent of
877 * both sequences involved.
882 public SequenceToSequenceMapping getCoveringCodonMapping(SequenceI seq)
884 for (SequenceToSequenceMapping mapping : mappings)
886 if (mapping.getMapping().getMap().isTripletMap()
887 && mapping.covers(seq))
889 if (mapping.fromSeq == seq
890 && mapping.covers(mapping.getMapping().getTo()))
894 else if (mapping.getMapping().getTo() == seq
895 && mapping.covers(mapping.fromSeq))