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.analysis;
23 import java.util.ArrayList;
24 import java.util.Arrays;
25 import java.util.Collection;
26 import java.util.Collections;
27 import java.util.HashMap;
28 import java.util.HashSet;
29 import java.util.Iterator;
30 import java.util.LinkedHashMap;
31 import java.util.List;
32 import java.util.Locale;
34 import java.util.Map.Entry;
35 import java.util.NoSuchElementException;
37 import java.util.SortedMap;
38 import java.util.TreeMap;
40 import jalview.api.AlignCalcWorkerI;
41 import jalview.bin.Console;
42 import jalview.commands.RemoveGapColCommand;
43 import jalview.datamodel.AlignedCodon;
44 import jalview.datamodel.AlignedCodonFrame;
45 import jalview.datamodel.AlignedCodonFrame.SequenceToSequenceMapping;
46 import jalview.datamodel.Alignment;
47 import jalview.datamodel.AlignmentAnnotation;
48 import jalview.datamodel.AlignmentI;
49 import jalview.datamodel.ContactMatrixI;
50 import jalview.datamodel.DBRefEntry;
51 import jalview.datamodel.GeneLociI;
52 import jalview.datamodel.IncompleteCodonException;
53 import jalview.datamodel.Mapping;
54 import jalview.datamodel.Sequence;
55 import jalview.datamodel.SequenceFeature;
56 import jalview.datamodel.SequenceGroup;
57 import jalview.datamodel.SequenceI;
58 import jalview.datamodel.features.SequenceFeatures;
59 import jalview.gui.AlignmentPanel;
60 import jalview.io.gff.SequenceOntologyI;
61 import jalview.schemes.ResidueProperties;
62 import jalview.util.Comparison;
63 import jalview.util.DBRefUtils;
64 import jalview.util.IntRangeComparator;
65 import jalview.util.MapList;
66 import jalview.util.MappingUtils;
67 import jalview.workers.SecondaryStructureConsensusThread;
70 * grab bag of useful alignment manipulation operations Expect these to be
71 * refactored elsewhere at some point.
76 public class AlignmentUtils
78 private static final int CODON_LENGTH = 3;
80 private static final String SEQUENCE_VARIANT = "sequence_variant:";
82 private static final String SS_ANNOTATION_LABEL = "Secondary Structure";
85 * the 'id' attribute is provided for variant features fetched from
86 * Ensembl using its REST service with JSON format
88 public static final String VARIANT_ID = "id";
91 * A data model to hold the 'normal' base value at a position, and an optional
92 * sequence variant feature
94 static final class DnaVariant
98 SequenceFeature variant;
100 DnaVariant(String nuc)
106 DnaVariant(String nuc, SequenceFeature var)
112 public String getSource()
114 return variant == null ? null : variant.getFeatureGroup();
118 * toString for aid in the debugger only
121 public String toString()
123 return base + ":" + (variant == null ? "" : variant.getDescription());
128 * given an existing alignment, create a new alignment including all, or up to
129 * flankSize additional symbols from each sequence's dataset sequence
135 public static AlignmentI expandContext(AlignmentI core, int flankSize)
137 List<SequenceI> sq = new ArrayList<>();
139 for (SequenceI s : core.getSequences())
141 SequenceI newSeq = s.deriveSequence();
142 final int newSeqStart = newSeq.getStart() - 1;
143 if (newSeqStart > maxoffset
144 && newSeq.getDatasetSequence().getStart() < s.getStart())
146 maxoffset = newSeqStart;
152 maxoffset = Math.min(maxoffset, flankSize);
156 * now add offset left and right to create an expanded alignment
158 for (SequenceI s : sq)
161 while (ds.getDatasetSequence() != null)
163 ds = ds.getDatasetSequence();
165 int s_end = s.findPosition(s.getStart() + s.getLength());
166 // find available flanking residues for sequence
167 int ustream_ds = s.getStart() - ds.getStart();
168 int dstream_ds = ds.getEnd() - s_end;
170 // build new flanked sequence
172 // compute gap padding to start of flanking sequence
173 int offset = maxoffset - ustream_ds;
175 // padding is gapChar x ( maxoffset - min(ustream_ds, flank)
178 if (flankSize < ustream_ds)
180 // take up to flankSize residues
181 offset = maxoffset - flankSize;
182 ustream_ds = flankSize;
184 if (flankSize <= dstream_ds)
186 dstream_ds = flankSize - 1;
189 // TODO use Character.toLowerCase to avoid creating String objects?
190 char[] upstream = new String(ds
191 .getSequence(s.getStart() - 1 - ustream_ds, s.getStart() - 1))
192 .toLowerCase(Locale.ROOT).toCharArray();
193 char[] downstream = new String(
194 ds.getSequence(s_end - 1, s_end + dstream_ds))
195 .toLowerCase(Locale.ROOT).toCharArray();
196 char[] coreseq = s.getSequence();
197 char[] nseq = new char[offset + upstream.length + downstream.length
199 char c = core.getGapCharacter();
202 for (; p < offset; p++)
207 System.arraycopy(upstream, 0, nseq, p, upstream.length);
208 System.arraycopy(coreseq, 0, nseq, p + upstream.length,
210 System.arraycopy(downstream, 0, nseq,
211 p + coreseq.length + upstream.length, downstream.length);
212 s.setSequence(new String(nseq));
213 s.setStart(s.getStart() - ustream_ds);
214 s.setEnd(s_end + downstream.length);
216 AlignmentI newAl = new jalview.datamodel.Alignment(
217 sq.toArray(new SequenceI[0]));
218 for (SequenceI s : sq)
220 if (s.getAnnotation() != null)
222 for (AlignmentAnnotation aa : s.getAnnotation())
224 aa.adjustForAlignment(); // JAL-1712 fix
225 newAl.addAnnotation(aa);
229 newAl.setDataset(core.getDataset());
234 * Returns the index (zero-based position) of a sequence in an alignment, or
241 public static int getSequenceIndex(AlignmentI al, SequenceI seq)
245 for (SequenceI alSeq : al.getSequences())
258 * Returns a map of lists of sequences in the alignment, keyed by sequence
259 * name. For use in mapping between different alignment views of the same
262 * @see jalview.datamodel.AlignmentI#getSequencesByName()
264 public static Map<String, List<SequenceI>> getSequencesByName(
267 Map<String, List<SequenceI>> theMap = new LinkedHashMap<>();
268 for (SequenceI seq : al.getSequences())
270 String name = seq.getName();
273 List<SequenceI> seqs = theMap.get(name);
276 seqs = new ArrayList<>();
277 theMap.put(name, seqs);
286 * Build mapping of protein to cDNA alignment. Mappings are made between
287 * sequences where the cDNA translates to the protein sequence. Any new
288 * mappings are added to the protein alignment. Returns true if any mappings
289 * either already exist or were added, else false.
291 * @param proteinAlignment
292 * @param cdnaAlignment
295 public static boolean mapProteinAlignmentToCdna(
296 final AlignmentI proteinAlignment, final AlignmentI cdnaAlignment)
298 if (proteinAlignment == null || cdnaAlignment == null)
303 Set<SequenceI> mappedDna = new HashSet<>();
304 Set<SequenceI> mappedProtein = new HashSet<>();
307 * First pass - map sequences where cross-references exist. This include
308 * 1-to-many mappings to support, for example, variant cDNA.
310 boolean mappingPerformed = mapProteinToCdna(proteinAlignment,
311 cdnaAlignment, mappedDna, mappedProtein, true);
314 * Second pass - map sequences where no cross-references exist. This only
315 * does 1-to-1 mappings and assumes corresponding sequences are in the same
316 * order in the alignments.
318 mappingPerformed |= mapProteinToCdna(proteinAlignment, cdnaAlignment,
319 mappedDna, mappedProtein, false);
320 return mappingPerformed;
324 * Make mappings between compatible sequences (where the cDNA translation
325 * matches the protein).
327 * @param proteinAlignment
328 * @param cdnaAlignment
330 * a set of mapped DNA sequences (to add to)
331 * @param mappedProtein
332 * a set of mapped Protein sequences (to add to)
334 * if true, only map sequences where xrefs exist
337 protected static boolean mapProteinToCdna(
338 final AlignmentI proteinAlignment, final AlignmentI cdnaAlignment,
339 Set<SequenceI> mappedDna, Set<SequenceI> mappedProtein,
342 boolean mappingExistsOrAdded = false;
343 List<SequenceI> thisSeqs = proteinAlignment.getSequences();
344 for (SequenceI aaSeq : thisSeqs)
346 boolean proteinMapped = false;
347 AlignedCodonFrame acf = new AlignedCodonFrame();
349 for (SequenceI cdnaSeq : cdnaAlignment.getSequences())
352 * Always try to map if sequences have xref to each other; this supports
353 * variant cDNA or alternative splicing for a protein sequence.
355 * If no xrefs, try to map progressively, assuming that alignments have
356 * mappable sequences in corresponding order. These are not
357 * many-to-many, as that would risk mixing species with similar cDNA
360 if (xrefsOnly && !AlignmentUtils.haveCrossRef(aaSeq, cdnaSeq))
366 * Don't map non-xrefd sequences more than once each. This heuristic
367 * allows us to pair up similar sequences in ordered alignments.
369 if (!xrefsOnly && (mappedProtein.contains(aaSeq)
370 || mappedDna.contains(cdnaSeq)))
374 if (mappingExists(proteinAlignment.getCodonFrames(),
375 aaSeq.getDatasetSequence(), cdnaSeq.getDatasetSequence()))
377 mappingExistsOrAdded = true;
381 MapList map = mapCdnaToProtein(aaSeq, cdnaSeq);
384 acf.addMap(cdnaSeq, aaSeq, map);
385 mappingExistsOrAdded = true;
386 proteinMapped = true;
387 mappedDna.add(cdnaSeq);
388 mappedProtein.add(aaSeq);
394 proteinAlignment.addCodonFrame(acf);
397 return mappingExistsOrAdded;
401 * Answers true if the mappings include one between the given (dataset)
404 protected static boolean mappingExists(List<AlignedCodonFrame> mappings,
405 SequenceI aaSeq, SequenceI cdnaSeq)
407 if (mappings != null)
409 for (AlignedCodonFrame acf : mappings)
411 if (cdnaSeq == acf.getDnaForAaSeq(aaSeq))
421 * Builds a mapping (if possible) of a cDNA to a protein sequence.
423 * <li>first checks if the cdna translates exactly to the protein
425 * <li>else checks for translation after removing a STOP codon</li>
426 * <li>else checks for translation after removing a START codon</li>
427 * <li>if that fails, inspect CDS features on the cDNA sequence</li>
429 * Returns null if no mapping is determined.
432 * the aligned protein sequence
434 * the aligned cdna sequence
437 public static MapList mapCdnaToProtein(SequenceI proteinSeq,
441 * Here we handle either dataset sequence set (desktop) or absent (applet).
442 * Use only the char[] form of the sequence to avoid creating possibly large
445 final SequenceI proteinDataset = proteinSeq.getDatasetSequence();
446 char[] aaSeqChars = proteinDataset != null
447 ? proteinDataset.getSequence()
448 : proteinSeq.getSequence();
449 final SequenceI cdnaDataset = cdnaSeq.getDatasetSequence();
450 char[] cdnaSeqChars = cdnaDataset != null ? cdnaDataset.getSequence()
451 : cdnaSeq.getSequence();
452 if (aaSeqChars == null || cdnaSeqChars == null)
458 * cdnaStart/End, proteinStartEnd are base 1 (for dataset sequence mapping)
460 final int mappedLength = CODON_LENGTH * aaSeqChars.length;
461 int cdnaLength = cdnaSeqChars.length;
462 int cdnaStart = cdnaSeq.getStart();
463 int cdnaEnd = cdnaSeq.getEnd();
464 final int proteinStart = proteinSeq.getStart();
465 final int proteinEnd = proteinSeq.getEnd();
468 * If lengths don't match, try ignoring stop codon (if present)
470 if (cdnaLength != mappedLength && cdnaLength > 2)
472 String lastCodon = String.valueOf(cdnaSeqChars,
473 cdnaLength - CODON_LENGTH, CODON_LENGTH)
474 .toUpperCase(Locale.ROOT);
475 for (String stop : ResidueProperties.STOP_CODONS)
477 if (lastCodon.equals(stop))
479 cdnaEnd -= CODON_LENGTH;
480 cdnaLength -= CODON_LENGTH;
487 * If lengths still don't match, try ignoring start codon.
490 if (cdnaLength != mappedLength && cdnaLength > 2
491 && String.valueOf(cdnaSeqChars, 0, CODON_LENGTH)
492 .toUpperCase(Locale.ROOT)
493 .equals(ResidueProperties.START))
495 startOffset += CODON_LENGTH;
496 cdnaStart += CODON_LENGTH;
497 cdnaLength -= CODON_LENGTH;
500 if (translatesAs(cdnaSeqChars, startOffset, aaSeqChars))
503 * protein is translation of dna (+/- start/stop codons)
505 MapList map = new MapList(new int[] { cdnaStart, cdnaEnd },
507 { proteinStart, proteinEnd }, CODON_LENGTH, 1);
512 * translation failed - try mapping CDS annotated regions of dna
514 return mapCdsToProtein(cdnaSeq, proteinSeq);
518 * Test whether the given cdna sequence, starting at the given offset,
519 * translates to the given amino acid sequence, using the standard translation
520 * table. Designed to fail fast i.e. as soon as a mismatch position is found.
522 * @param cdnaSeqChars
527 protected static boolean translatesAs(char[] cdnaSeqChars, int cdnaStart,
530 if (cdnaSeqChars == null || aaSeqChars == null)
536 int dnaPos = cdnaStart;
537 for (; dnaPos < cdnaSeqChars.length - 2
538 && aaPos < aaSeqChars.length; dnaPos += CODON_LENGTH, aaPos++)
540 String codon = String.valueOf(cdnaSeqChars, dnaPos, CODON_LENGTH);
541 final String translated = ResidueProperties.codonTranslate(codon);
544 * allow * in protein to match untranslatable in dna
546 final char aaRes = aaSeqChars[aaPos];
547 if ((translated == null || ResidueProperties.STOP.equals(translated))
552 if (translated == null || !(aaRes == translated.charAt(0)))
555 // jalview.bin.Console.outPrintln(("Mismatch at " + i + "/" + aaResidue
557 // + codon + "(" + translated + ") != " + aaRes));
563 * check we matched all of the protein sequence
565 if (aaPos != aaSeqChars.length)
571 * check we matched all of the dna except
572 * for optional trailing STOP codon
574 if (dnaPos == cdnaSeqChars.length)
578 if (dnaPos == cdnaSeqChars.length - CODON_LENGTH)
580 String codon = String.valueOf(cdnaSeqChars, dnaPos, CODON_LENGTH);
581 if (ResidueProperties.STOP
582 .equals(ResidueProperties.codonTranslate(codon)))
591 * Align sequence 'seq' to match the alignment of a mapped sequence. Note this
592 * currently assumes that we are aligning cDNA to match protein.
595 * the sequence to be realigned
597 * the alignment whose sequence alignment is to be 'copied'
599 * character string represent a gap in the realigned sequence
600 * @param preserveUnmappedGaps
601 * @param preserveMappedGaps
602 * @return true if the sequence was realigned, false if it could not be
604 public static boolean alignSequenceAs(SequenceI seq, AlignmentI al,
605 String gap, boolean preserveMappedGaps,
606 boolean preserveUnmappedGaps)
609 * Get any mappings from the source alignment to the target (dataset)
612 // TODO there may be one AlignedCodonFrame per dataset sequence, or one with
613 // all mappings. Would it help to constrain this?
614 List<AlignedCodonFrame> mappings = al.getCodonFrame(seq);
615 if (mappings == null || mappings.isEmpty())
621 * Locate the aligned source sequence whose dataset sequence is mapped. We
622 * just take the first match here (as we can't align like more than one
625 SequenceI alignFrom = null;
626 AlignedCodonFrame mapping = null;
627 for (AlignedCodonFrame mp : mappings)
629 alignFrom = mp.findAlignedSequence(seq, al);
630 if (alignFrom != null)
637 if (alignFrom == null)
641 alignSequenceAs(seq, alignFrom, mapping, gap, al.getGapCharacter(),
642 preserveMappedGaps, preserveUnmappedGaps);
647 * Align sequence 'alignTo' the same way as 'alignFrom', using the mapping to
648 * match residues and codons. Flags control whether existing gaps in unmapped
649 * (intron) and mapped (exon) regions are preserved or not. Gaps between
650 * intron and exon are only retained if both flags are set.
657 * @param preserveUnmappedGaps
658 * @param preserveMappedGaps
660 public static void alignSequenceAs(SequenceI alignTo, SequenceI alignFrom,
661 AlignedCodonFrame mapping, String myGap, char sourceGap,
662 boolean preserveMappedGaps, boolean preserveUnmappedGaps)
664 // TODO generalise to work for Protein-Protein, dna-dna, dna-protein
666 // aligned and dataset sequence positions, all base zero
670 int basesWritten = 0;
671 char myGapChar = myGap.charAt(0);
672 int ratio = myGap.length();
674 int fromOffset = alignFrom.getStart() - 1;
675 int toOffset = alignTo.getStart() - 1;
676 int sourceGapMappedLength = 0;
677 boolean inExon = false;
678 final int toLength = alignTo.getLength();
679 final int fromLength = alignFrom.getLength();
680 StringBuilder thisAligned = new StringBuilder(2 * toLength);
683 * Traverse the 'model' aligned sequence
685 for (int i = 0; i < fromLength; i++)
687 char sourceChar = alignFrom.getCharAt(i);
688 if (sourceChar == sourceGap)
690 sourceGapMappedLength += ratio;
695 * Found a non-gap character. Locate its mapped region if any.
698 // Note mapping positions are base 1, our sequence positions base 0
699 int[] mappedPos = mapping.getMappedRegion(alignTo, alignFrom,
700 sourceDsPos + fromOffset);
701 if (mappedPos == null)
704 * unmapped position; treat like a gap
706 sourceGapMappedLength += ratio;
707 // jalview.bin.Console.errPrintln("Can't align: no codon mapping to
709 // + sourceDsPos + "(" + sourceChar + ")");
714 int mappedCodonStart = mappedPos[0]; // position (1...) of codon start
715 int mappedCodonEnd = mappedPos[mappedPos.length - 1]; // codon end pos
716 StringBuilder trailingCopiedGap = new StringBuilder();
719 * Copy dna sequence up to and including this codon. Optionally, include
720 * gaps before the codon starts (in introns) and/or after the codon starts
723 * Note this only works for 'linear' splicing, not reverse or interleaved.
724 * But then 'align dna as protein' doesn't make much sense otherwise.
726 int intronLength = 0;
727 while (basesWritten + toOffset < mappedCodonEnd
728 && thisSeqPos < toLength)
730 final char c = alignTo.getCharAt(thisSeqPos++);
734 int sourcePosition = basesWritten + toOffset;
735 if (sourcePosition < mappedCodonStart)
738 * Found an unmapped (intron) base. First add in any preceding gaps
741 if (preserveUnmappedGaps && trailingCopiedGap.length() > 0)
743 thisAligned.append(trailingCopiedGap.toString());
744 intronLength += trailingCopiedGap.length();
745 trailingCopiedGap = new StringBuilder();
752 final boolean startOfCodon = sourcePosition == mappedCodonStart;
753 int gapsToAdd = calculateGapsToInsert(preserveMappedGaps,
754 preserveUnmappedGaps, sourceGapMappedLength, inExon,
755 trailingCopiedGap.length(), intronLength, startOfCodon);
756 for (int k = 0; k < gapsToAdd; k++)
758 thisAligned.append(myGapChar);
760 sourceGapMappedLength = 0;
763 thisAligned.append(c);
764 trailingCopiedGap = new StringBuilder();
768 if (inExon && preserveMappedGaps)
770 trailingCopiedGap.append(myGapChar);
772 else if (!inExon && preserveUnmappedGaps)
774 trailingCopiedGap.append(myGapChar);
781 * At end of model aligned sequence. Copy any remaining target sequence, optionally
782 * including (intron) gaps.
784 while (thisSeqPos < toLength)
786 final char c = alignTo.getCharAt(thisSeqPos++);
787 if (c != myGapChar || preserveUnmappedGaps)
789 thisAligned.append(c);
791 sourceGapMappedLength--;
795 * finally add gaps to pad for any trailing source gaps or
796 * unmapped characters
798 if (preserveUnmappedGaps)
800 while (sourceGapMappedLength > 0)
802 thisAligned.append(myGapChar);
803 sourceGapMappedLength--;
808 * All done aligning, set the aligned sequence.
810 alignTo.setSequence(new String(thisAligned));
814 * Helper method to work out how many gaps to insert when realigning.
816 * @param preserveMappedGaps
817 * @param preserveUnmappedGaps
818 * @param sourceGapMappedLength
820 * @param trailingCopiedGap
821 * @param intronLength
822 * @param startOfCodon
825 protected static int calculateGapsToInsert(boolean preserveMappedGaps,
826 boolean preserveUnmappedGaps, int sourceGapMappedLength,
827 boolean inExon, int trailingGapLength, int intronLength,
828 final boolean startOfCodon)
834 * Reached start of codon. Ignore trailing gaps in intron unless we are
835 * preserving gaps in both exon and intron. Ignore them anyway if the
836 * protein alignment introduces a gap at least as large as the intronic
839 if (inExon && !preserveMappedGaps)
841 trailingGapLength = 0;
843 if (!inExon && !(preserveMappedGaps && preserveUnmappedGaps))
845 trailingGapLength = 0;
849 gapsToAdd = Math.max(sourceGapMappedLength, trailingGapLength);
853 if (intronLength + trailingGapLength <= sourceGapMappedLength)
855 gapsToAdd = sourceGapMappedLength - intronLength;
859 gapsToAdd = Math.min(
860 intronLength + trailingGapLength - sourceGapMappedLength,
868 * second or third base of codon; check for any gaps in dna
870 if (!preserveMappedGaps)
872 trailingGapLength = 0;
874 gapsToAdd = Math.max(sourceGapMappedLength, trailingGapLength);
880 * Realigns the given protein to match the alignment of the dna, using codon
881 * mappings to translate aligned codon positions to protein residues.
884 * the alignment whose sequences are realigned by this method
886 * the dna alignment whose alignment we are 'copying'
887 * @return the number of sequences that were realigned
889 public static int alignProteinAsDna(AlignmentI protein, AlignmentI dna)
891 if (protein.isNucleotide() || !dna.isNucleotide())
894 .errPrintln("Wrong alignment type in alignProteinAsDna");
897 List<SequenceI> unmappedProtein = new ArrayList<>();
898 Map<AlignedCodon, Map<SequenceI, AlignedCodon>> alignedCodons = buildCodonColumnsMap(
899 protein, dna, unmappedProtein);
900 return alignProteinAs(protein, alignedCodons, unmappedProtein);
904 * Realigns the given dna to match the alignment of the protein, using codon
905 * mappings to translate aligned peptide positions to codons.
907 * Always produces a padded CDS alignment.
910 * the alignment whose sequences are realigned by this method
912 * the protein alignment whose alignment we are 'copying'
913 * @return the number of sequences that were realigned
915 public static int alignCdsAsProtein(AlignmentI dna, AlignmentI protein)
917 if (protein.isNucleotide() || !dna.isNucleotide())
920 .errPrintln("Wrong alignment type in alignProteinAsDna");
923 // todo: implement this
924 List<AlignedCodonFrame> mappings = protein.getCodonFrames();
925 int alignedCount = 0;
926 int width = 0; // alignment width for padding CDS
927 for (SequenceI dnaSeq : dna.getSequences())
929 if (alignCdsSequenceAsProtein(dnaSeq, protein, mappings,
930 dna.getGapCharacter()))
934 width = Math.max(dnaSeq.getLength(), width);
938 for (SequenceI dnaSeq : dna.getSequences())
940 oldwidth = dnaSeq.getLength();
941 diff = width - oldwidth;
944 dnaSeq.insertCharAt(oldwidth, diff, dna.getGapCharacter());
951 * Helper method to align (if possible) the dna sequence to match the
952 * alignment of a mapped protein sequence. This is currently limited to
953 * handling coding sequence only.
961 static boolean alignCdsSequenceAsProtein(SequenceI cdsSeq,
962 AlignmentI protein, List<AlignedCodonFrame> mappings,
965 SequenceI cdsDss = cdsSeq.getDatasetSequence();
969 .println("alignCdsSequenceAsProtein needs aligned sequence!");
973 List<AlignedCodonFrame> dnaMappings = MappingUtils
974 .findMappingsForSequence(cdsSeq, mappings);
975 for (AlignedCodonFrame mapping : dnaMappings)
977 SequenceI peptide = mapping.findAlignedSequence(cdsSeq, protein);
980 final int peptideLength = peptide.getLength();
981 Mapping map = mapping.getMappingBetween(cdsSeq, peptide);
984 MapList mapList = map.getMap();
985 if (map.getTo() == peptide.getDatasetSequence())
987 mapList = mapList.getInverse();
989 final int cdsLength = cdsDss.getLength();
990 int mappedFromLength = MappingUtils
991 .getLength(mapList.getFromRanges());
992 int mappedToLength = MappingUtils
993 .getLength(mapList.getToRanges());
994 boolean addStopCodon = (cdsLength == mappedFromLength
995 * CODON_LENGTH + CODON_LENGTH)
996 || (peptide.getDatasetSequence()
997 .getLength() == mappedFromLength - 1);
998 if (cdsLength != mappedToLength && !addStopCodon)
1000 jalview.bin.Console.errPrintln(String.format(
1001 "Can't align cds as protein (length mismatch %d/%d): %s",
1002 cdsLength, mappedToLength, cdsSeq.getName()));
1006 * pre-fill the aligned cds sequence with gaps
1008 char[] alignedCds = new char[peptideLength * CODON_LENGTH
1009 + (addStopCodon ? CODON_LENGTH : 0)];
1010 Arrays.fill(alignedCds, gapChar);
1013 * walk over the aligned peptide sequence and insert mapped
1014 * codons for residues in the aligned cds sequence
1016 int copiedBases = 0;
1017 int cdsStart = cdsDss.getStart();
1018 int proteinPos = peptide.getStart() - 1;
1021 for (int col = 0; col < peptideLength; col++)
1023 char residue = peptide.getCharAt(col);
1025 if (Comparison.isGap(residue))
1027 cdsCol += CODON_LENGTH;
1032 int[] codon = mapList.locateInTo(proteinPos, proteinPos);
1035 // e.g. incomplete start codon, X in peptide
1036 cdsCol += CODON_LENGTH;
1040 for (int j = codon[0]; j <= codon[1]; j++)
1042 char mappedBase = cdsDss.getCharAt(j - cdsStart);
1043 alignedCds[cdsCol++] = mappedBase;
1051 * append stop codon if not mapped from protein,
1052 * closing it up to the end of the mapped sequence
1054 if (copiedBases == cdsLength - CODON_LENGTH)
1056 for (int i = alignedCds.length - 1; i >= 0; i--)
1058 if (!Comparison.isGap(alignedCds[i]))
1060 cdsCol = i + 1; // gap just after end of sequence
1064 for (int i = cdsLength - CODON_LENGTH; i < cdsLength; i++)
1066 alignedCds[cdsCol++] = cdsDss.getCharAt(i);
1069 cdsSeq.setSequence(new String(alignedCds));
1078 * Builds a map whose key is an aligned codon position (3 alignment column
1079 * numbers base 0), and whose value is a map from protein sequence to each
1080 * protein's peptide residue for that codon. The map generates an ordering of
1081 * the codons, and allows us to read off the peptides at each position in
1082 * order to assemble 'aligned' protein sequences.
1085 * the protein alignment
1087 * the coding dna alignment
1088 * @param unmappedProtein
1089 * any unmapped proteins are added to this list
1092 protected static Map<AlignedCodon, Map<SequenceI, AlignedCodon>> buildCodonColumnsMap(
1093 AlignmentI protein, AlignmentI dna,
1094 List<SequenceI> unmappedProtein)
1097 * maintain a list of any proteins with no mappings - these will be
1098 * rendered 'as is' in the protein alignment as we can't align them
1100 unmappedProtein.addAll(protein.getSequences());
1102 List<AlignedCodonFrame> mappings = protein.getCodonFrames();
1105 * Map will hold, for each aligned codon position e.g. [3, 5, 6], a map of
1106 * {dnaSequence, {proteinSequence, codonProduct}} at that position. The
1107 * comparator keeps the codon positions ordered.
1109 Map<AlignedCodon, Map<SequenceI, AlignedCodon>> alignedCodons = new TreeMap<>(
1110 new CodonComparator());
1112 for (SequenceI dnaSeq : dna.getSequences())
1114 for (AlignedCodonFrame mapping : mappings)
1116 SequenceI prot = mapping.findAlignedSequence(dnaSeq, protein);
1119 Mapping seqMap = mapping.getMappingForSequence(dnaSeq);
1120 addCodonPositions(dnaSeq, prot, protein.getGapCharacter(), seqMap,
1122 unmappedProtein.remove(prot);
1128 * Finally add any unmapped peptide start residues (e.g. for incomplete
1129 * codons) as if at the codon position before the second residue
1131 // TODO resolve JAL-2022 so this fudge can be removed
1132 int mappedSequenceCount = protein.getHeight() - unmappedProtein.size();
1133 addUnmappedPeptideStarts(alignedCodons, mappedSequenceCount);
1135 return alignedCodons;
1139 * Scans for any protein mapped from position 2 (meaning unmapped start
1140 * position e.g. an incomplete codon), and synthesizes a 'codon' for it at the
1141 * preceding position in the alignment
1143 * @param alignedCodons
1144 * the codon-to-peptide map
1145 * @param mappedSequenceCount
1146 * the number of distinct sequences in the map
1148 protected static void addUnmappedPeptideStarts(
1149 Map<AlignedCodon, Map<SequenceI, AlignedCodon>> alignedCodons,
1150 int mappedSequenceCount)
1152 // TODO delete this ugly hack once JAL-2022 is resolved
1153 // i.e. we can model startPhase > 0 (incomplete start codon)
1155 List<SequenceI> sequencesChecked = new ArrayList<>();
1156 AlignedCodon lastCodon = null;
1157 Map<SequenceI, AlignedCodon> toAdd = new HashMap<>();
1159 for (Entry<AlignedCodon, Map<SequenceI, AlignedCodon>> entry : alignedCodons
1162 for (Entry<SequenceI, AlignedCodon> sequenceCodon : entry.getValue()
1165 SequenceI seq = sequenceCodon.getKey();
1166 if (sequencesChecked.contains(seq))
1170 sequencesChecked.add(seq);
1171 AlignedCodon codon = sequenceCodon.getValue();
1172 if (codon.peptideCol > 1)
1174 jalview.bin.Console.errPrintln(
1175 "Problem mapping protein with >1 unmapped start positions: "
1178 else if (codon.peptideCol == 1)
1181 * first position (peptideCol == 0) was unmapped - add it
1183 if (lastCodon != null)
1185 AlignedCodon firstPeptide = new AlignedCodon(lastCodon.pos1,
1186 lastCodon.pos2, lastCodon.pos3,
1187 String.valueOf(seq.getCharAt(0)), 0);
1188 toAdd.put(seq, firstPeptide);
1193 * unmapped residue at start of alignment (no prior column) -
1194 * 'insert' at nominal codon [0, 0, 0]
1196 AlignedCodon firstPeptide = new AlignedCodon(0, 0, 0,
1197 String.valueOf(seq.getCharAt(0)), 0);
1198 toAdd.put(seq, firstPeptide);
1201 if (sequencesChecked.size() == mappedSequenceCount)
1203 // no need to check past first mapped position in all sequences
1207 lastCodon = entry.getKey();
1211 * add any new codons safely after iterating over the map
1213 for (Entry<SequenceI, AlignedCodon> startCodon : toAdd.entrySet())
1215 addCodonToMap(alignedCodons, startCodon.getValue(),
1216 startCodon.getKey());
1221 * Update the aligned protein sequences to match the codon alignments given in
1225 * @param alignedCodons
1226 * an ordered map of codon positions (columns), with sequence/peptide
1227 * values present in each column
1228 * @param unmappedProtein
1231 protected static int alignProteinAs(AlignmentI protein,
1232 Map<AlignedCodon, Map<SequenceI, AlignedCodon>> alignedCodons,
1233 List<SequenceI> unmappedProtein)
1236 * prefill peptide sequences with gaps
1238 int alignedWidth = alignedCodons.size();
1239 char[] gaps = new char[alignedWidth];
1240 Arrays.fill(gaps, protein.getGapCharacter());
1241 Map<SequenceI, char[]> peptides = new HashMap<>();
1242 for (SequenceI seq : protein.getSequences())
1244 if (!unmappedProtein.contains(seq))
1246 peptides.put(seq, Arrays.copyOf(gaps, gaps.length));
1251 * Traverse the codons left to right (as defined by CodonComparator)
1252 * and insert peptides in each column where the sequence is mapped.
1253 * This gives a peptide 'alignment' where residues are aligned if their
1254 * corresponding codons occupy the same columns in the cdna alignment.
1257 for (AlignedCodon codon : alignedCodons.keySet())
1259 final Map<SequenceI, AlignedCodon> columnResidues = alignedCodons
1261 for (Entry<SequenceI, AlignedCodon> entry : columnResidues.entrySet())
1263 char residue = entry.getValue().product.charAt(0);
1264 peptides.get(entry.getKey())[column] = residue;
1270 * and finally set the constructed sequences
1272 for (Entry<SequenceI, char[]> entry : peptides.entrySet())
1274 entry.getKey().setSequence(new String(entry.getValue()));
1281 * Populate the map of aligned codons by traversing the given sequence
1282 * mapping, locating the aligned positions of mapped codons, and adding those
1283 * positions and their translation products to the map.
1286 * the aligned sequence we are mapping from
1288 * the sequence to be aligned to the codons
1290 * the gap character in the dna sequence
1292 * a mapping to a sequence translation
1293 * @param alignedCodons
1294 * the map we are building up
1296 static void addCodonPositions(SequenceI dna, SequenceI protein,
1297 char gapChar, Mapping seqMap,
1298 Map<AlignedCodon, Map<SequenceI, AlignedCodon>> alignedCodons)
1300 Iterator<AlignedCodon> codons = seqMap.getCodonIterator(dna, gapChar);
1303 * add codon positions, and their peptide translations, to the alignment
1304 * map, while remembering the first codon mapped
1306 while (codons.hasNext())
1310 AlignedCodon codon = codons.next();
1311 addCodonToMap(alignedCodons, codon, protein);
1312 } catch (IncompleteCodonException e)
1314 // possible incomplete trailing codon - ignore
1315 } catch (NoSuchElementException e)
1317 // possibly peptide lacking STOP
1323 * Helper method to add a codon-to-peptide entry to the aligned codons map
1325 * @param alignedCodons
1329 protected static void addCodonToMap(
1330 Map<AlignedCodon, Map<SequenceI, AlignedCodon>> alignedCodons,
1331 AlignedCodon codon, SequenceI protein)
1333 Map<SequenceI, AlignedCodon> seqProduct = alignedCodons.get(codon);
1334 if (seqProduct == null)
1336 seqProduct = new HashMap<>();
1337 alignedCodons.put(codon, seqProduct);
1339 seqProduct.put(protein, codon);
1343 * Returns true if a cDNA/Protein mapping either exists, or could be made,
1344 * between at least one pair of sequences in the two alignments. Currently,
1347 * <li>One alignment must be nucleotide, and the other protein</li>
1348 * <li>At least one pair of sequences must be already mapped, or mappable</li>
1349 * <li>Mappable means the nucleotide translation matches the protein
1351 * <li>The translation may ignore start and stop codons if present in the
1359 public static boolean isMappable(AlignmentI al1, AlignmentI al2)
1361 if (al1 == null || al2 == null)
1367 * Require one nucleotide and one protein
1369 if (al1.isNucleotide() == al2.isNucleotide())
1373 AlignmentI dna = al1.isNucleotide() ? al1 : al2;
1374 AlignmentI protein = dna == al1 ? al2 : al1;
1375 List<AlignedCodonFrame> mappings = protein.getCodonFrames();
1376 for (SequenceI dnaSeq : dna.getSequences())
1378 for (SequenceI proteinSeq : protein.getSequences())
1380 if (isMappable(dnaSeq, proteinSeq, mappings))
1390 * Returns true if the dna sequence is mapped, or could be mapped, to the
1398 protected static boolean isMappable(SequenceI dnaSeq,
1399 SequenceI proteinSeq, List<AlignedCodonFrame> mappings)
1401 if (dnaSeq == null || proteinSeq == null)
1406 SequenceI dnaDs = dnaSeq.getDatasetSequence() == null ? dnaSeq
1407 : dnaSeq.getDatasetSequence();
1408 SequenceI proteinDs = proteinSeq.getDatasetSequence() == null
1410 : proteinSeq.getDatasetSequence();
1412 for (AlignedCodonFrame mapping : mappings)
1414 if (proteinDs == mapping.getAaForDnaSeq(dnaDs))
1424 * Just try to make a mapping (it is not yet stored), test whether
1427 return mapCdnaToProtein(proteinDs, dnaDs) != null;
1431 * Finds any reference annotations associated with the sequences in
1432 * sequenceScope, that are not already added to the alignment, and adds them
1433 * to the 'candidates' map. Also populates a lookup table of annotation
1434 * labels, keyed by calcId, for use in constructing tooltips or the like.
1436 * @param sequenceScope
1437 * the sequences to scan for reference annotations
1438 * @param labelForCalcId
1439 * (optional) map to populate with label for calcId
1441 * map to populate with annotations for sequence
1443 * the alignment to check for presence of annotations
1445 public static void findAddableReferenceAnnotations(
1446 List<SequenceI> sequenceScope, Map<String, String> labelForCalcId,
1447 final Map<SequenceI, List<AlignmentAnnotation>> candidates,
1450 if (sequenceScope == null)
1456 * For each sequence in scope, make a list of any annotations on the
1457 * underlying dataset sequence which are not already on the alignment.
1459 * Add to a map of { alignmentSequence, <List of annotations to add> }
1461 for (SequenceI seq : sequenceScope)
1463 SequenceI dataset = seq.getDatasetSequence();
1464 if (dataset == null)
1468 AlignmentAnnotation[] datasetAnnotations = dataset.getAnnotation();
1469 if (datasetAnnotations == null)
1473 final List<AlignmentAnnotation> result = new ArrayList<>();
1474 for (AlignmentAnnotation dsann : datasetAnnotations)
1477 * Find matching annotations on the alignment. If none is found, then
1478 * add this annotation to the list of 'addable' annotations for this
1481 final Iterable<AlignmentAnnotation> matchedAlignmentAnnotations = al
1482 .findAnnotations(seq, dsann.getCalcId(), dsann.label);
1483 boolean found = false;
1484 if (matchedAlignmentAnnotations != null)
1486 for (AlignmentAnnotation matched : matchedAlignmentAnnotations)
1488 if (dsann.description.equals(matched.description))
1498 if (labelForCalcId != null)
1500 labelForCalcId.put(dsann.getCalcId(), dsann.label);
1505 * Save any addable annotations for this sequence
1507 if (!result.isEmpty())
1509 candidates.put(seq, result);
1515 * Adds annotations to the top of the alignment annotations, in the same order
1516 * as their related sequences. If you already have an annotation and want to
1517 * add it to a sequence in an alignment use {@code addReferenceAnnotationTo}
1519 * @param annotations
1520 * the annotations to add
1522 * the alignment to add them to
1523 * @param selectionGroup
1524 * current selection group - may be null, if provided then any added
1525 * annotation will be trimmed to just those columns in the selection
1528 public static void addReferenceAnnotations(
1529 Map<SequenceI, List<AlignmentAnnotation>> annotations,
1530 final AlignmentI alignment, final SequenceGroup selectionGroup)
1532 for (SequenceI seq : annotations.keySet())
1534 for (AlignmentAnnotation ann : annotations.get(seq))
1536 addReferenceAnnotationTo(alignment, seq, ann, selectionGroup);
1542 public static boolean isSSAnnotationPresent( Map<SequenceI, List<AlignmentAnnotation>> annotations) {
1544 for (SequenceI seq : annotations.keySet())
1546 for (AlignmentAnnotation ann : annotations.get(seq))
1548 if(ann.getDescription(false).startsWith(SS_ANNOTATION_LABEL)) {
1557 * Make a copy of a reference annotation {@code ann} and add it to an
1558 * alignment sequence {@code seq} in {@code alignment}, optionally limited to
1559 * the extent of {@code selectionGroup}
1564 * @param selectionGroup
1565 * current selection group - may be null, if provided then any added
1566 * annotation will be trimmed to just those columns in the selection
1568 * @return annotation added to {@code seq and {@code alignment}
1570 public static AlignmentAnnotation addReferenceAnnotationTo(
1571 final AlignmentI alignment, final SequenceI seq,
1572 final AlignmentAnnotation ann, final SequenceGroup selectionGroup)
1574 AlignmentAnnotation copyAnn = new AlignmentAnnotation(ann);
1576 int endRes = ann.annotations.length;
1577 if (selectionGroup != null)
1579 startRes = -1 + Math.min(seq.getEnd(), Math.max(seq.getStart(),
1580 seq.findPosition(selectionGroup.getStartRes())));
1581 endRes = -1 + Math.min(seq.getEnd(),
1582 seq.findPosition(selectionGroup.getEndRes()));
1585 copyAnn.restrict(startRes, endRes + 0);
1588 * Add to the sequence (sets copyAnn.datasetSequence), unless the
1589 * original annotation is already on the sequence.
1591 if (!seq.hasAnnotation(ann))
1593 ContactMatrixI cm = seq.getDatasetSequence().getContactMatrixFor(ann);
1596 seq.addContactListFor(copyAnn, cm);
1598 seq.addAlignmentAnnotation(copyAnn);
1601 copyAnn.adjustForAlignment();
1602 // add to the alignment and set visible
1603 alignment.addAnnotation(copyAnn);
1604 copyAnn.visible = true;
1610 * Set visibility of alignment annotations of specified types (labels), for
1611 * specified sequences. This supports controls like "Show all secondary
1612 * structure", "Hide all Temp factor", etc.
1614 * @al the alignment to scan for annotations
1616 * the types (labels) of annotations to be updated
1617 * @param forSequences
1618 * if not null, only annotations linked to one of these sequences are
1619 * in scope for update; if null, acts on all sequence annotations
1621 * if this flag is true, 'types' is ignored (label not checked)
1623 * if true, set visibility on, else set off
1625 public static void showOrHideSequenceAnnotations(AlignmentI al,
1626 Collection<String> types, List<SequenceI> forSequences,
1627 boolean anyType, boolean doShow)
1629 AlignmentAnnotation[] anns = al.getAlignmentAnnotation();
1632 for (AlignmentAnnotation aa : anns)
1634 if (anyType || types.contains(aa.label))
1636 if ((aa.sequenceRef != null) && (forSequences == null
1637 || forSequences.contains(aa.sequenceRef)))
1639 aa.visible = doShow;
1646 public static AlignmentAnnotation getFirstSequenceAnnotationOfType(
1647 AlignmentI al, int graphType)
1649 AlignmentAnnotation[] anns = al.getAlignmentAnnotation();
1652 for (AlignmentAnnotation aa : anns)
1654 if (aa.sequenceRef != null && aa.graph == graphType)
1662 * Returns true if either sequence has a cross-reference to the other
1668 public static boolean haveCrossRef(SequenceI seq1, SequenceI seq2)
1670 // Note: moved here from class CrossRef as the latter class has dependencies
1671 // not availability to the applet's classpath
1672 return hasCrossRef(seq1, seq2) || hasCrossRef(seq2, seq1);
1676 * Returns true if seq1 has a cross-reference to seq2. Currently this assumes
1677 * that sequence name is structured as Source|AccessionId.
1683 public static boolean hasCrossRef(SequenceI seq1, SequenceI seq2)
1685 if (seq1 == null || seq2 == null)
1689 String name = seq2.getName();
1690 final List<DBRefEntry> xrefs = seq1.getDBRefs();
1693 for (int ix = 0, nx = xrefs.size(); ix < nx; ix++)
1695 DBRefEntry xref = xrefs.get(ix);
1696 String xrefName = xref.getSource() + "|" + xref.getAccessionId();
1697 // case-insensitive test, consistent with DBRefEntry.equalRef()
1698 if (xrefName.equalsIgnoreCase(name))
1708 * Constructs an alignment consisting of the mapped (CDS) regions in the given
1709 * nucleotide sequences, and updates mappings to match. The CDS sequences are
1710 * added to the original alignment's dataset, which is shared by the new
1711 * alignment. Mappings from nucleotide to CDS, and from CDS to protein, are
1712 * added to the alignment dataset.
1715 * aligned nucleotide (dna or cds) sequences
1717 * the alignment dataset the sequences belong to
1719 * (optional) to restrict results to CDS that map to specified
1721 * @return an alignment whose sequences are the cds-only parts of the dna
1722 * sequences (or null if no mappings are found)
1724 public static AlignmentI makeCdsAlignment(SequenceI[] dna,
1725 AlignmentI dataset, SequenceI[] products)
1727 if (dataset == null || dataset.getDataset() != null)
1729 throw new IllegalArgumentException(
1730 "IMPLEMENTATION ERROR: dataset.getDataset() must be null!");
1732 List<SequenceI> foundSeqs = new ArrayList<>();
1733 List<SequenceI> cdsSeqs = new ArrayList<>();
1734 List<AlignedCodonFrame> mappings = dataset.getCodonFrames();
1735 HashSet<SequenceI> productSeqs = null;
1736 if (products != null)
1738 productSeqs = new HashSet<>();
1739 for (SequenceI seq : products)
1741 productSeqs.add(seq.getDatasetSequence() == null ? seq
1742 : seq.getDatasetSequence());
1747 * Construct CDS sequences from mappings on the alignment dataset.
1749 * - find the protein product(s) mapped to from each dna sequence
1750 * - if the mapping covers the whole dna sequence (give or take start/stop
1751 * codon), take the dna as the CDS sequence
1752 * - else search dataset mappings for a suitable dna sequence, i.e. one
1753 * whose whole sequence is mapped to the protein
1754 * - if no sequence found, construct one from the dna sequence and mapping
1755 * (and add it to dataset so it is found if this is repeated)
1757 for (SequenceI dnaSeq : dna)
1759 SequenceI dnaDss = dnaSeq.getDatasetSequence() == null ? dnaSeq
1760 : dnaSeq.getDatasetSequence();
1762 List<AlignedCodonFrame> seqMappings = MappingUtils
1763 .findMappingsForSequence(dnaSeq, mappings);
1764 for (AlignedCodonFrame mapping : seqMappings)
1766 List<Mapping> mappingsFromSequence = mapping
1767 .getMappingsFromSequence(dnaSeq);
1769 for (Mapping aMapping : mappingsFromSequence)
1771 MapList mapList = aMapping.getMap();
1772 if (mapList.getFromRatio() == 1)
1775 * not a dna-to-protein mapping (likely dna-to-cds)
1781 * skip if mapping is not to one of the target set of proteins
1783 SequenceI proteinProduct = aMapping.getTo();
1784 if (productSeqs != null && !productSeqs.contains(proteinProduct))
1790 * try to locate the CDS from the dataset mappings;
1791 * guard against duplicate results (for the case that protein has
1792 * dbrefs to both dna and cds sequences)
1794 SequenceI cdsSeq = findCdsForProtein(mappings, dnaSeq,
1795 seqMappings, aMapping);
1798 if (!foundSeqs.contains(cdsSeq))
1800 foundSeqs.add(cdsSeq);
1801 SequenceI derivedSequence = cdsSeq.deriveSequence();
1802 cdsSeqs.add(derivedSequence);
1803 if (!dataset.getSequences().contains(cdsSeq))
1805 dataset.addSequence(cdsSeq);
1812 * didn't find mapped CDS sequence - construct it and add
1813 * its dataset sequence to the dataset
1815 cdsSeq = makeCdsSequence(dnaSeq.getDatasetSequence(), aMapping,
1816 dataset).deriveSequence();
1817 // cdsSeq has a name constructed as CDS|<dbref>
1818 // <dbref> will be either the accession for the coding sequence,
1819 // marked in the /via/ dbref to the protein product accession
1820 // or it will be the original nucleotide accession.
1821 SequenceI cdsSeqDss = cdsSeq.getDatasetSequence();
1823 cdsSeqs.add(cdsSeq);
1826 * build the mapping from CDS to protein
1828 List<int[]> cdsRange = Collections
1829 .singletonList(new int[]
1830 { cdsSeq.getStart(),
1831 cdsSeq.getLength() + cdsSeq.getStart() - 1 });
1832 MapList cdsToProteinMap = new MapList(cdsRange,
1833 mapList.getToRanges(), mapList.getFromRatio(),
1834 mapList.getToRatio());
1836 if (!dataset.getSequences().contains(cdsSeqDss))
1839 * if this sequence is a newly created one, add it to the dataset
1840 * and made a CDS to protein mapping (if sequence already exists,
1841 * CDS-to-protein mapping _is_ the transcript-to-protein mapping)
1843 dataset.addSequence(cdsSeqDss);
1844 AlignedCodonFrame cdsToProteinMapping = new AlignedCodonFrame();
1845 cdsToProteinMapping.addMap(cdsSeqDss, proteinProduct,
1849 * guard against duplicating the mapping if repeating this action
1851 if (!mappings.contains(cdsToProteinMapping))
1853 mappings.add(cdsToProteinMapping);
1857 propagateDBRefsToCDS(cdsSeqDss, dnaSeq.getDatasetSequence(),
1858 proteinProduct, aMapping);
1860 * add another mapping from original 'from' range to CDS
1862 AlignedCodonFrame dnaToCdsMapping = new AlignedCodonFrame();
1863 final MapList dnaToCdsMap = new MapList(mapList.getFromRanges(),
1865 dnaToCdsMapping.addMap(dnaSeq.getDatasetSequence(), cdsSeqDss,
1867 if (!mappings.contains(dnaToCdsMapping))
1869 mappings.add(dnaToCdsMapping);
1873 * transfer dna chromosomal loci (if known) to the CDS
1874 * sequence (via the mapping)
1876 final MapList cdsToDnaMap = dnaToCdsMap.getInverse();
1877 transferGeneLoci(dnaSeq, cdsToDnaMap, cdsSeq);
1880 * add DBRef with mapping from protein to CDS
1881 * (this enables Get Cross-References from protein alignment)
1882 * This is tricky because we can't have two DBRefs with the
1883 * same source and accession, so need a different accession for
1884 * the CDS from the dna sequence
1887 // specific use case:
1888 // Genomic contig ENSCHR:1, contains coding regions for ENSG01,
1889 // ENSG02, ENSG03, with transcripts and products similarly named.
1890 // cannot add distinct dbrefs mapping location on ENSCHR:1 to ENSG01
1892 // JBPNote: ?? can't actually create an example that demonstrates we
1894 // synthesize an xref.
1896 List<DBRefEntry> primrefs = dnaDss.getPrimaryDBRefs();
1897 for (int ip = 0, np = primrefs.size(); ip < np; ip++)
1899 DBRefEntry primRef = primrefs.get(ip);
1901 * create a cross-reference from CDS to the source sequence's
1902 * primary reference and vice versa
1904 String source = primRef.getSource();
1905 String version = primRef.getVersion();
1906 DBRefEntry cdsCrossRef = new DBRefEntry(source,
1907 source + ":" + version, primRef.getAccessionId());
1909 .setMap(new Mapping(dnaDss, new MapList(cdsToDnaMap)));
1910 cdsSeqDss.addDBRef(cdsCrossRef);
1912 dnaSeq.addDBRef(new DBRefEntry(source, version,
1913 cdsSeq.getName(), new Mapping(cdsSeqDss, dnaToCdsMap)));
1914 // problem here is that the cross-reference is synthesized -
1915 // cdsSeq.getName() may be like 'CDS|dnaaccession' or
1917 // assuming cds version same as dna ?!?
1919 DBRefEntry proteinToCdsRef = new DBRefEntry(source, version,
1922 proteinToCdsRef.setMap(
1923 new Mapping(cdsSeqDss, cdsToProteinMap.getInverse()));
1924 proteinProduct.addDBRef(proteinToCdsRef);
1927 * transfer any features on dna that overlap the CDS
1929 transferFeatures(dnaSeq, cdsSeq, dnaToCdsMap, null,
1930 SequenceOntologyI.CDS);
1935 AlignmentI cds = new Alignment(
1936 cdsSeqs.toArray(new SequenceI[cdsSeqs.size()]));
1937 cds.setDataset(dataset);
1943 * Tries to transfer gene loci (dbref to chromosome positions) from fromSeq to
1944 * toSeq, mediated by the given mapping between the sequences
1947 * @param targetToFrom
1951 protected static void transferGeneLoci(SequenceI fromSeq,
1952 MapList targetToFrom, SequenceI targetSeq)
1954 if (targetSeq.getGeneLoci() != null)
1956 // already have - don't override
1959 GeneLociI fromLoci = fromSeq.getGeneLoci();
1960 if (fromLoci == null)
1965 MapList newMap = targetToFrom.traverse(fromLoci.getMapping());
1969 targetSeq.setGeneLoci(fromLoci.getSpeciesId(),
1970 fromLoci.getAssemblyId(), fromLoci.getChromosomeId(), newMap);
1975 * A helper method that finds a CDS sequence in the alignment dataset that is
1976 * mapped to the given protein sequence, and either is, or has a mapping from,
1977 * the given dna sequence.
1980 * set of all mappings on the dataset
1982 * a dna (or cds) sequence we are searching from
1983 * @param seqMappings
1984 * the set of mappings involving dnaSeq
1986 * a transcript-to-peptide mapping
1989 static SequenceI findCdsForProtein(List<AlignedCodonFrame> mappings,
1990 SequenceI dnaSeq, List<AlignedCodonFrame> seqMappings,
1994 * TODO a better dna-cds-protein mapping data representation to allow easy
1995 * navigation; until then this clunky looping around lists of mappings
1997 SequenceI seqDss = dnaSeq.getDatasetSequence() == null ? dnaSeq
1998 : dnaSeq.getDatasetSequence();
1999 SequenceI proteinProduct = aMapping.getTo();
2002 * is this mapping from the whole dna sequence (i.e. CDS)?
2003 * allowing for possible stop codon on dna but not peptide
2005 int mappedFromLength = MappingUtils
2006 .getLength(aMapping.getMap().getFromRanges());
2007 int dnaLength = seqDss.getLength();
2008 if (mappedFromLength == dnaLength
2009 || mappedFromLength == dnaLength - CODON_LENGTH)
2012 * if sequence has CDS features, this is a transcript with no UTR
2013 * - do not take this as the CDS sequence! (JAL-2789)
2015 if (seqDss.getFeatures().getFeaturesByOntology(SequenceOntologyI.CDS)
2023 * looks like we found the dna-to-protein mapping; search for the
2024 * corresponding cds-to-protein mapping
2026 List<AlignedCodonFrame> mappingsToPeptide = MappingUtils
2027 .findMappingsForSequence(proteinProduct, mappings);
2028 for (AlignedCodonFrame acf : mappingsToPeptide)
2030 for (SequenceToSequenceMapping map : acf.getMappings())
2032 Mapping mapping = map.getMapping();
2033 if (mapping != aMapping
2034 && mapping.getMap().getFromRatio() == CODON_LENGTH
2035 && proteinProduct == mapping.getTo()
2036 && seqDss != map.getFromSeq())
2038 mappedFromLength = MappingUtils
2039 .getLength(mapping.getMap().getFromRanges());
2040 if (mappedFromLength == map.getFromSeq().getLength())
2043 * found a 3:1 mapping to the protein product which covers
2044 * the whole dna sequence i.e. is from CDS; finally check the CDS
2045 * is mapped from the given dna start sequence
2047 SequenceI cdsSeq = map.getFromSeq();
2048 // todo this test is weak if seqMappings contains multiple mappings;
2049 // we get away with it if transcript:cds relationship is 1:1
2050 List<AlignedCodonFrame> dnaToCdsMaps = MappingUtils
2051 .findMappingsForSequence(cdsSeq, seqMappings);
2052 if (!dnaToCdsMaps.isEmpty())
2064 * Helper method that makes a CDS sequence as defined by the mappings from the
2065 * given sequence i.e. extracts the 'mapped from' ranges (which may be on
2066 * forward or reverse strand).
2071 * - existing dataset. We check for sequences that look like the CDS
2072 * we are about to construct, if one exists already, then we will
2073 * just return that one.
2074 * @return CDS sequence (as a dataset sequence)
2076 static SequenceI makeCdsSequence(SequenceI seq, Mapping mapping,
2080 * construct CDS sequence name as "CDS|" with 'from id' held in the mapping
2081 * if set (e.g. EMBL protein_id), else sequence name appended
2083 String mapFromId = mapping.getMappedFromId();
2084 final String seqId = "CDS|"
2085 + (mapFromId != null ? mapFromId : seq.getName());
2087 SequenceI newSeq = null;
2090 * construct CDS sequence by splicing mapped from ranges
2092 char[] seqChars = seq.getSequence();
2093 List<int[]> fromRanges = mapping.getMap().getFromRanges();
2094 int cdsWidth = MappingUtils.getLength(fromRanges);
2095 char[] newSeqChars = new char[cdsWidth];
2098 for (int[] range : fromRanges)
2100 if (range[0] <= range[1])
2102 // forward strand mapping - just copy the range
2103 int length = range[1] - range[0] + 1;
2104 System.arraycopy(seqChars, range[0] - 1, newSeqChars, newPos,
2110 // reverse strand mapping - copy and complement one by one
2111 for (int i = range[0]; i >= range[1]; i--)
2113 newSeqChars[newPos++] = Dna.getComplement(seqChars[i - 1]);
2117 newSeq = new Sequence(seqId, newSeqChars, 1, newPos);
2120 if (dataset != null)
2122 SequenceI[] matches = dataset.findSequenceMatch(newSeq.getName());
2123 if (matches != null)
2125 boolean matched = false;
2126 for (SequenceI mtch : matches)
2128 if (mtch.getStart() != newSeq.getStart())
2132 if (mtch.getEnd() != newSeq.getEnd())
2136 if (!Arrays.equals(mtch.getSequence(), newSeq.getSequence()))
2148 "JAL-2154 regression: warning - found (and ignored) a duplicate CDS sequence:"
2154 // newSeq.setDescription(mapFromId);
2160 * Adds any DBRefEntrys to cdsSeq from contig that have a Mapping congruent to
2161 * the given mapping.
2165 * @param proteinProduct
2167 * @return list of DBRefEntrys added
2169 protected static List<DBRefEntry> propagateDBRefsToCDS(SequenceI cdsSeq,
2170 SequenceI contig, SequenceI proteinProduct, Mapping mapping)
2173 // gather direct refs from contig congruent with mapping
2174 List<DBRefEntry> direct = new ArrayList<>();
2175 HashSet<String> directSources = new HashSet<>();
2177 List<DBRefEntry> refs = contig.getDBRefs();
2180 for (int ib = 0, nb = refs.size(); ib < nb; ib++)
2182 DBRefEntry dbr = refs.get(ib);
2184 if (dbr.hasMap() && (map = dbr.getMap().getMap()).isTripletMap())
2186 // check if map is the CDS mapping
2187 if (mapping.getMap().equals(map))
2190 directSources.add(dbr.getSource());
2195 List<DBRefEntry> onSource = DBRefUtils.selectRefs(
2196 proteinProduct.getDBRefs(),
2197 directSources.toArray(new String[0]));
2198 List<DBRefEntry> propagated = new ArrayList<>();
2200 // and generate appropriate mappings
2201 for (int ic = 0, nc = direct.size(); ic < nc; ic++)
2203 DBRefEntry cdsref = direct.get(ic);
2204 Mapping m = cdsref.getMap();
2205 // clone maplist and mapping
2206 MapList cdsposmap = new MapList(
2207 Arrays.asList(new int[][]
2208 { new int[] { cdsSeq.getStart(), cdsSeq.getEnd() } }),
2209 m.getMap().getToRanges(), 3, 1);
2210 Mapping cdsmap = new Mapping(m.getTo(), m.getMap());
2213 DBRefEntry newref = new DBRefEntry(cdsref.getSource(),
2214 cdsref.getVersion(), cdsref.getAccessionId(),
2215 new Mapping(cdsmap.getTo(), cdsposmap));
2217 // and see if we can map to the protein product for this mapping.
2218 // onSource is the filtered set of accessions on protein that we are
2219 // tranferring, so we assume accession is the same.
2220 if (cdsmap.getTo() == null && onSource != null)
2222 List<DBRefEntry> sourceRefs = DBRefUtils.searchRefs(onSource,
2223 cdsref.getAccessionId());
2224 if (sourceRefs != null)
2226 for (DBRefEntry srcref : sourceRefs)
2228 if (srcref.getSource().equalsIgnoreCase(cdsref.getSource()))
2230 // we have found a complementary dbref on the protein product, so
2231 // update mapping's getTo
2232 newref.getMap().setTo(proteinProduct);
2237 cdsSeq.addDBRef(newref);
2238 propagated.add(newref);
2244 * Transfers co-located features on 'fromSeq' to 'toSeq', adjusting the
2245 * feature start/end ranges, optionally omitting specified feature types.
2246 * Returns the number of features copied.
2251 * the mapping from 'fromSeq' to 'toSeq'
2253 * if not null, only features of this type are copied (including
2254 * subtypes in the Sequence Ontology)
2257 protected static int transferFeatures(SequenceI fromSeq, SequenceI toSeq,
2258 MapList mapping, String select, String... omitting)
2260 SequenceI copyTo = toSeq;
2261 while (copyTo.getDatasetSequence() != null)
2263 copyTo = copyTo.getDatasetSequence();
2265 if (fromSeq == copyTo || fromSeq.getDatasetSequence() == copyTo)
2267 return 0; // shared dataset sequence
2271 * get features, optionally restricted by an ontology term
2273 List<SequenceFeature> sfs = select == null
2274 ? fromSeq.getFeatures().getPositionalFeatures()
2275 : fromSeq.getFeatures().getFeaturesByOntology(select);
2278 for (SequenceFeature sf : sfs)
2280 String type = sf.getType();
2281 boolean omit = false;
2282 for (String toOmit : omitting)
2284 if (type.equals(toOmit))
2295 * locate the mapped range - null if either start or end is
2296 * not mapped (no partial overlaps are calculated)
2298 int start = sf.getBegin();
2299 int end = sf.getEnd();
2300 int[] mappedTo = mapping.locateInTo(start, end);
2302 * if whole exon range doesn't map, try interpreting it
2303 * as 5' or 3' exon overlapping the CDS range
2305 if (mappedTo == null)
2307 mappedTo = mapping.locateInTo(end, end);
2308 if (mappedTo != null)
2311 * end of exon is in CDS range - 5' overlap
2312 * to a range from the start of the peptide
2317 if (mappedTo == null)
2319 mappedTo = mapping.locateInTo(start, start);
2320 if (mappedTo != null)
2323 * start of exon is in CDS range - 3' overlap
2324 * to a range up to the end of the peptide
2326 mappedTo[1] = toSeq.getLength();
2329 if (mappedTo != null)
2331 int newBegin = Math.min(mappedTo[0], mappedTo[1]);
2332 int newEnd = Math.max(mappedTo[0], mappedTo[1]);
2333 SequenceFeature copy = new SequenceFeature(sf, newBegin, newEnd,
2334 sf.getFeatureGroup(), sf.getScore());
2335 copyTo.addSequenceFeature(copy);
2343 * Returns a mapping from dna to protein by inspecting sequence features of
2344 * type "CDS" on the dna. A mapping is constructed if the total CDS feature
2345 * length is 3 times the peptide length (optionally after dropping a trailing
2346 * stop codon). This method does not check whether the CDS nucleotide sequence
2347 * translates to the peptide sequence.
2353 public static MapList mapCdsToProtein(SequenceI dnaSeq,
2354 SequenceI proteinSeq)
2356 List<int[]> ranges = findCdsPositions(dnaSeq);
2357 int mappedDnaLength = MappingUtils.getLength(ranges);
2360 * if not a whole number of codons, truncate mapping
2362 int codonRemainder = mappedDnaLength % CODON_LENGTH;
2363 if (codonRemainder > 0)
2365 mappedDnaLength -= codonRemainder;
2366 MappingUtils.removeEndPositions(codonRemainder, ranges);
2369 int proteinLength = proteinSeq.getLength();
2370 int proteinStart = proteinSeq.getStart();
2371 int proteinEnd = proteinSeq.getEnd();
2374 * incomplete start codon may mean X at start of peptide
2375 * we ignore both for mapping purposes
2377 if (proteinSeq.getCharAt(0) == 'X')
2379 // todo JAL-2022 support startPhase > 0
2383 List<int[]> proteinRange = new ArrayList<>();
2386 * dna length should map to protein (or protein plus stop codon)
2388 int codesForResidues = mappedDnaLength / CODON_LENGTH;
2389 if (codesForResidues == (proteinLength + 1))
2391 // assuming extra codon is for STOP and not in peptide
2392 // todo: check trailing codon is indeed a STOP codon
2394 mappedDnaLength -= CODON_LENGTH;
2395 MappingUtils.removeEndPositions(CODON_LENGTH, ranges);
2398 if (codesForResidues == proteinLength)
2400 proteinRange.add(new int[] { proteinStart, proteinEnd });
2401 return new MapList(ranges, proteinRange, CODON_LENGTH, 1);
2407 * Returns a list of CDS ranges found (as sequence positions base 1), i.e. of
2408 * [start, end] positions of sequence features of type "CDS" (or a sub-type of
2409 * CDS in the Sequence Ontology). The ranges are sorted into ascending start
2410 * position order, so this method is only valid for linear CDS in the same
2411 * sense as the protein product.
2416 protected static List<int[]> findCdsPositions(SequenceI dnaSeq)
2418 List<int[]> result = new ArrayList<>();
2420 List<SequenceFeature> sfs = dnaSeq.getFeatures()
2421 .getFeaturesByOntology(SequenceOntologyI.CDS);
2426 SequenceFeatures.sortFeatures(sfs, true);
2428 for (SequenceFeature sf : sfs)
2433 String s = sf.getPhase();
2436 phase = Integer.parseInt(s);
2438 } catch (NumberFormatException e)
2443 * phase > 0 on first codon means 5' incomplete - skip to the start
2444 * of the next codon; example ENST00000496384
2446 int begin = sf.getBegin();
2447 int end = sf.getEnd();
2448 if (result.isEmpty() && phase > 0)
2453 // shouldn't happen!
2455 .println("Error: start phase extends beyond start CDS in "
2456 + dnaSeq.getName());
2459 result.add(new int[] { begin, end });
2463 * Finally sort ranges by start position. This avoids a dependency on
2464 * keeping features in order on the sequence (if they are in order anyway,
2465 * the sort will have almost no work to do). The implicit assumption is CDS
2466 * ranges are assembled in order. Other cases should not use this method,
2467 * but instead construct an explicit mapping for CDS (e.g. EMBL parsing).
2469 Collections.sort(result, IntRangeComparator.ASCENDING);
2474 * Makes an alignment with a copy of the given sequences, adding in any
2475 * non-redundant sequences which are mapped to by the cross-referenced
2481 * the alignment dataset shared by the new copy
2484 public static AlignmentI makeCopyAlignment(SequenceI[] seqs,
2485 SequenceI[] xrefs, AlignmentI dataset)
2487 AlignmentI copy = new Alignment(new Alignment(seqs));
2488 copy.setDataset(dataset);
2489 boolean isProtein = !copy.isNucleotide();
2490 SequenceIdMatcher matcher = new SequenceIdMatcher(seqs);
2493 // BH 2019.01.25 recoded to remove iterators
2495 for (int ix = 0, nx = xrefs.length; ix < nx; ix++)
2497 SequenceI xref = xrefs[ix];
2498 List<DBRefEntry> dbrefs = xref.getDBRefs();
2501 for (int ir = 0, nir = dbrefs.size(); ir < nir; ir++)
2503 DBRefEntry dbref = dbrefs.get(ir);
2504 Mapping map = dbref.getMap();
2506 if (map == null || (mto = map.getTo()) == null
2507 || mto.isProtein() != isProtein)
2511 SequenceI mappedTo = mto;
2512 SequenceI match = matcher.findIdMatch(mappedTo);
2515 matcher.add(mappedTo);
2516 copy.addSequence(mappedTo);
2526 * Try to align sequences in 'unaligned' to match the alignment of their
2527 * mapped regions in 'aligned'. For example, could use this to align CDS
2528 * sequences which are mapped to their parent cDNA sequences.
2530 * This method handles 1:1 mappings (dna-to-dna or protein-to-protein). For
2531 * dna-to-protein or protein-to-dna use alternative methods.
2534 * sequences to be aligned
2536 * holds aligned sequences and their mappings
2539 public static int alignAs(AlignmentI unaligned, AlignmentI aligned)
2542 * easy case - aligning a copy of aligned sequences
2544 if (alignAsSameSequences(unaligned, aligned))
2546 return unaligned.getHeight();
2550 * fancy case - aligning via mappings between sequences
2552 List<SequenceI> unmapped = new ArrayList<>();
2553 Map<Integer, Map<SequenceI, Character>> columnMap = buildMappedColumnsMap(
2554 unaligned, aligned, unmapped);
2555 int width = columnMap.size();
2556 char gap = unaligned.getGapCharacter();
2557 int realignedCount = 0;
2558 // TODO: verify this loop scales sensibly for very wide/high alignments
2560 for (SequenceI seq : unaligned.getSequences())
2562 if (!unmapped.contains(seq))
2564 char[] newSeq = new char[width];
2565 Arrays.fill(newSeq, gap); // JBPComment - doubt this is faster than the
2566 // Integer iteration below
2571 * traverse the map to find columns populated
2574 for (Integer column : columnMap.keySet())
2576 Character c = columnMap.get(column).get(seq);
2580 * sequence has a character at this position
2590 * trim trailing gaps
2592 if (lastCol < width)
2594 char[] tmp = new char[lastCol + 1];
2595 System.arraycopy(newSeq, 0, tmp, 0, lastCol + 1);
2598 // TODO: optimise SequenceI to avoid char[]->String->char[]
2599 seq.setSequence(String.valueOf(newSeq));
2603 return realignedCount;
2607 * If unaligned and aligned sequences share the same dataset sequences, then
2608 * simply copies the aligned sequences to the unaligned sequences and returns
2609 * true; else returns false
2612 * - sequences to be aligned based on aligned
2614 * - 'guide' alignment containing sequences derived from same dataset
2618 static boolean alignAsSameSequences(AlignmentI unaligned,
2621 if (aligned.getDataset() == null || unaligned.getDataset() == null)
2623 return false; // should only pass alignments with datasets here
2626 // map from dataset sequence to alignment sequence(s)
2627 Map<SequenceI, List<SequenceI>> alignedDatasets = new HashMap<>();
2628 for (SequenceI seq : aligned.getSequences())
2630 SequenceI ds = seq.getDatasetSequence();
2631 if (alignedDatasets.get(ds) == null)
2633 alignedDatasets.put(ds, new ArrayList<SequenceI>());
2635 alignedDatasets.get(ds).add(seq);
2639 * first pass - check whether all sequences to be aligned share a
2640 * dataset sequence with an aligned sequence; also note the leftmost
2641 * ungapped column from which to copy
2643 int leftmost = Integer.MAX_VALUE;
2644 for (SequenceI seq : unaligned.getSequences())
2646 final SequenceI ds = seq.getDatasetSequence();
2647 if (!alignedDatasets.containsKey(ds))
2651 SequenceI alignedSeq = alignedDatasets.get(ds).get(0);
2652 int startCol = alignedSeq.findIndex(seq.getStart()); // 1..
2653 leftmost = Math.min(leftmost, startCol);
2657 * second pass - copy aligned sequences;
2658 * heuristic rule: pair off sequences in order for the case where
2659 * more than one shares the same dataset sequence
2661 final char gapCharacter = aligned.getGapCharacter();
2662 for (SequenceI seq : unaligned.getSequences())
2664 List<SequenceI> alignedSequences = alignedDatasets
2665 .get(seq.getDatasetSequence());
2666 if (alignedSequences.isEmpty())
2669 * defensive check - shouldn't happen! (JAL-3536)
2673 SequenceI alignedSeq = alignedSequences.get(0);
2676 * gap fill for leading (5') UTR if any
2678 // TODO this copies intron columns - wrong!
2679 int startCol = alignedSeq.findIndex(seq.getStart()); // 1..
2680 int endCol = alignedSeq.findIndex(seq.getEnd());
2681 char[] seqchars = new char[endCol - leftmost + 1];
2682 Arrays.fill(seqchars, gapCharacter);
2683 char[] toCopy = alignedSeq.getSequence(startCol - 1, endCol);
2684 System.arraycopy(toCopy, 0, seqchars, startCol - leftmost,
2686 seq.setSequence(String.valueOf(seqchars));
2687 if (alignedSequences.size() > 0)
2689 // pop off aligned sequences (except the last one)
2690 alignedSequences.remove(0);
2695 * finally remove gapped columns (e.g. introns)
2697 new RemoveGapColCommand("", unaligned.getSequencesArray(), 0,
2698 unaligned.getWidth() - 1, unaligned);
2704 * Returns a map whose key is alignment column number (base 1), and whose
2705 * values are a map of sequence characters in that column.
2712 static SortedMap<Integer, Map<SequenceI, Character>> buildMappedColumnsMap(
2713 AlignmentI unaligned, AlignmentI aligned,
2714 List<SequenceI> unmapped)
2717 * Map will hold, for each aligned column position, a map of
2718 * {unalignedSequence, characterPerSequence} at that position.
2719 * TreeMap keeps the entries in ascending column order.
2721 SortedMap<Integer, Map<SequenceI, Character>> map = new TreeMap<>();
2724 * record any sequences that have no mapping so can't be realigned
2726 unmapped.addAll(unaligned.getSequences());
2728 List<AlignedCodonFrame> mappings = aligned.getCodonFrames();
2730 for (SequenceI seq : unaligned.getSequences())
2732 for (AlignedCodonFrame mapping : mappings)
2734 SequenceI fromSeq = mapping.findAlignedSequence(seq, aligned);
2735 if (fromSeq != null)
2737 Mapping seqMap = mapping.getMappingBetween(fromSeq, seq);
2738 if (addMappedPositions(seq, fromSeq, seqMap, map))
2740 unmapped.remove(seq);
2749 * Helper method that adds to a map the mapped column positions of a sequence.
2751 * For example if aaTT-Tg-gAAA is mapped to TTTAAA then the map should record
2752 * that columns 3,4,6,10,11,12 map to characters T,T,T,A,A,A of the mapped to
2756 * the sequence whose column positions we are recording
2758 * a sequence that is mapped to the first sequence
2760 * the mapping from 'fromSeq' to 'seq'
2762 * a map to add the column positions (in fromSeq) of the mapped
2766 static boolean addMappedPositions(SequenceI seq, SequenceI fromSeq,
2767 Mapping seqMap, Map<Integer, Map<SequenceI, Character>> map)
2775 * invert mapping if it is from unaligned to aligned sequence
2777 if (seqMap.getTo() == fromSeq.getDatasetSequence())
2779 seqMap = new Mapping(seq.getDatasetSequence(),
2780 seqMap.getMap().getInverse());
2783 int toStart = seq.getStart();
2786 * traverse [start, end, start, end...] ranges in fromSeq
2788 for (int[] fromRange : seqMap.getMap().getFromRanges())
2790 for (int i = 0; i < fromRange.length - 1; i += 2)
2792 boolean forward = fromRange[i + 1] >= fromRange[i];
2795 * find the range mapped to (sequence positions base 1)
2797 int[] range = seqMap.locateMappedRange(fromRange[i],
2801 jalview.bin.Console.errPrintln("Error in mapping " + seqMap
2802 + " from " + fromSeq.getName());
2805 int fromCol = fromSeq.findIndex(fromRange[i]);
2806 int mappedCharPos = range[0];
2809 * walk over the 'from' aligned sequence in forward or reverse
2810 * direction; when a non-gap is found, record the column position
2811 * of the next character of the mapped-to sequence; stop when all
2812 * the characters of the range have been counted
2814 while (mappedCharPos <= range[1] && fromCol <= fromSeq.getLength()
2817 if (!Comparison.isGap(fromSeq.getCharAt(fromCol - 1)))
2820 * mapped from sequence has a character in this column
2821 * record the column position for the mapped to character
2823 Map<SequenceI, Character> seqsMap = map.get(fromCol);
2824 if (seqsMap == null)
2826 seqsMap = new HashMap<>();
2827 map.put(fromCol, seqsMap);
2829 seqsMap.put(seq, seq.getCharAt(mappedCharPos - toStart));
2832 fromCol += (forward ? 1 : -1);
2839 // strictly temporary hack until proper criteria for aligning protein to cds
2840 // are in place; this is so Ensembl -> fetch xrefs Uniprot aligns the Uniprot
2841 public static boolean looksLikeEnsembl(AlignmentI alignment)
2843 for (SequenceI seq : alignment.getSequences())
2845 String name = seq.getName();
2846 if (!name.startsWith("ENSG") && !name.startsWith("ENST"))