X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;ds=sidebyside;f=src%2Fjalview%2Fanalysis%2FAlignmentUtils.java;h=90d91970afeb520c60cdd342e6c68285695eecaa;hb=5eeb49c62b1004b1efe837a543c8ff84aa6d7faf;hp=b65096c224f947e86c10e59e8a5f24d7b50ebd46;hpb=ff450fad8709ae81919af7a15ea382af7292794c;p=jalview.git diff --git a/src/jalview/analysis/AlignmentUtils.java b/src/jalview/analysis/AlignmentUtils.java index b65096c..90d9197 100644 --- a/src/jalview/analysis/AlignmentUtils.java +++ b/src/jalview/analysis/AlignmentUtils.java @@ -170,10 +170,12 @@ public class AlignmentUtils } } // TODO use Character.toLowerCase to avoid creating String objects? - char[] upstream = new String(ds.getSequence(s.getStart() - 1 - - ustream_ds, s.getStart() - 1)).toLowerCase().toCharArray(); - char[] downstream = new String(ds.getSequence(s_end - 1, s_end - + dstream_ds)).toLowerCase().toCharArray(); + char[] upstream = new String(ds + .getSequence(s.getStart() - 1 - ustream_ds, s.getStart() - 1)) + .toLowerCase().toCharArray(); + char[] downstream = new String( + ds.getSequence(s_end - 1, s_end + dstream_ds)).toLowerCase() + .toCharArray(); char[] coreseq = s.getSequence(); char[] nseq = new char[offset + upstream.length + downstream.length + coreseq.length]; @@ -188,8 +190,8 @@ public class AlignmentUtils System.arraycopy(upstream, 0, nseq, p, upstream.length); System.arraycopy(coreseq, 0, nseq, p + upstream.length, coreseq.length); - System.arraycopy(downstream, 0, nseq, p + coreseq.length - + upstream.length, downstream.length); + System.arraycopy(downstream, 0, nseq, + p + coreseq.length + upstream.length, downstream.length); s.setSequence(new String(nseq)); s.setStart(s.getStart() - ustream_ds); s.setEnd(s_end + downstream.length); @@ -316,9 +318,9 @@ public class AlignmentUtils * @return */ protected static boolean mapProteinToCdna( - final AlignmentI proteinAlignment, - final AlignmentI cdnaAlignment, Set mappedDna, - Set mappedProtein, boolean xrefsOnly) + final AlignmentI proteinAlignment, final AlignmentI cdnaAlignment, + Set mappedDna, Set mappedProtein, + boolean xrefsOnly) { boolean mappingExistsOrAdded = false; List thisSeqs = proteinAlignment.getSequences(); @@ -347,9 +349,8 @@ public class AlignmentUtils * Don't map non-xrefd sequences more than once each. This heuristic * allows us to pair up similar sequences in ordered alignments. */ - if (!xrefsOnly - && (mappedProtein.contains(aaSeq) || mappedDna - .contains(cdnaSeq))) + if (!xrefsOnly && (mappedProtein.contains(aaSeq) + || mappedDna.contains(cdnaSeq))) { continue; } @@ -402,7 +403,8 @@ public class AlignmentUtils /** * Builds a mapping (if possible) of a cDNA to a protein sequence. *
    - *
  • first checks if the cdna translates exactly to the protein sequence
    • + *
  • first checks if the cdna translates exactly to the protein + * sequence
    • *
  • else checks for translation after removing a STOP codon
    • *
  • else checks for translation after removing a START codon
    • *
  • if that fails, inspect CDS features on the cDNA sequence
    • @@ -424,8 +426,9 @@ public class AlignmentUtils * String objects. */ final SequenceI proteinDataset = proteinSeq.getDatasetSequence(); - char[] aaSeqChars = proteinDataset != null ? proteinDataset - .getSequence() : proteinSeq.getSequence(); + char[] aaSeqChars = proteinDataset != null + ? proteinDataset.getSequence() + : proteinSeq.getSequence(); final SequenceI cdnaDataset = cdnaSeq.getDatasetSequence(); char[] cdnaSeqChars = cdnaDataset != null ? cdnaDataset.getSequence() : cdnaSeq.getSequence(); @@ -466,8 +469,7 @@ public class AlignmentUtils * If lengths still don't match, try ignoring start codon. */ int startOffset = 0; - if (cdnaLength != mappedLength - && cdnaLength > 2 + if (cdnaLength != mappedLength && cdnaLength > 2 && String.valueOf(cdnaSeqChars, 0, CODON_LENGTH).toUpperCase() .equals(ResidueProperties.START)) { @@ -481,8 +483,9 @@ public class AlignmentUtils /* * protein is translation of dna (+/- start/stop codons) */ - MapList map = new MapList(new int[] { cdnaStart, cdnaEnd }, new int[] - { proteinStart, proteinEnd }, CODON_LENGTH, 1); + MapList map = new MapList(new int[] { cdnaStart, cdnaEnd }, + new int[] + { proteinStart, proteinEnd }, CODON_LENGTH, 1); return map; } @@ -512,7 +515,8 @@ public class AlignmentUtils int aaPos = 0; int dnaPos = cdnaStart; - for (; dnaPos < cdnaSeqChars.length - 2 && aaPos < aaSeqChars.length; dnaPos += CODON_LENGTH, aaPos++) + for (; dnaPos < cdnaSeqChars.length - 2 + && aaPos < aaSeqChars.length; dnaPos += CODON_LENGTH, aaPos++) { String codon = String.valueOf(cdnaSeqChars, dnaPos, CODON_LENGTH); final String translated = ResidueProperties.codonTranslate(codon); @@ -631,10 +635,9 @@ public class AlignmentUtils * @param preserveUnmappedGaps * @param preserveMappedGaps */ - public static void alignSequenceAs(SequenceI alignTo, - SequenceI alignFrom, AlignedCodonFrame mapping, String myGap, - char sourceGap, boolean preserveMappedGaps, - boolean preserveUnmappedGaps) + public static void alignSequenceAs(SequenceI alignTo, SequenceI alignFrom, + AlignedCodonFrame mapping, String myGap, char sourceGap, + boolean preserveMappedGaps, boolean preserveUnmappedGaps) { // TODO generalise to work for Protein-Protein, dna-dna, dna-protein @@ -830,8 +833,9 @@ public class AlignmentUtils } else { - gapsToAdd = Math.min(intronLength + trailingGapLength - - sourceGapMappedLength, trailingGapLength); + gapsToAdd = Math.min( + intronLength + trailingGapLength - sourceGapMappedLength, + trailingGapLength); } } } @@ -930,7 +934,8 @@ public class AlignmentUtils * @return */ static boolean alignCdsSequenceAsProtein(SequenceI cdsSeq, - AlignmentI protein, List mappings, char gapChar) + AlignmentI protein, List mappings, + char gapChar) { SequenceI cdsDss = cdsSeq.getDatasetSequence(); if (cdsDss == null) @@ -963,14 +968,13 @@ public class AlignmentUtils .getLength(mapList.getToRanges()); boolean addStopCodon = (cdsLength == mappedFromLength * CODON_LENGTH + CODON_LENGTH) - || (peptide.getDatasetSequence().getLength() == mappedFromLength - 1); + || (peptide.getDatasetSequence() + .getLength() == mappedFromLength - 1); if (cdsLength != mappedToLength && !addStopCodon) { - System.err - .println(String - .format("Can't align cds as protein (length mismatch %d/%d): %s", - cdsLength, mappedToLength, - cdsSeq.getName())); + System.err.println(String.format( + "Can't align cds as protein (length mismatch %d/%d): %s", + cdsLength, mappedToLength, cdsSeq.getName())); } /* @@ -1088,8 +1092,8 @@ public class AlignmentUtils if (prot != null) { Mapping seqMap = mapping.getMappingForSequence(dnaSeq); - addCodonPositions(dnaSeq, prot, protein.getGapCharacter(), - seqMap, alignedCodons); + addCodonPositions(dnaSeq, prot, protein.getGapCharacter(), seqMap, + alignedCodons); unmappedProtein.remove(prot); } } @@ -1142,8 +1146,8 @@ public class AlignmentUtils AlignedCodon codon = sequenceCodon.getValue(); if (codon.peptideCol > 1) { - System.err - .println("Problem mapping protein with >1 unmapped start positions: " + System.err.println( + "Problem mapping protein with >1 unmapped start positions: " + seq.getName()); } else if (codon.peptideCol == 1) @@ -1154,8 +1158,8 @@ public class AlignmentUtils if (lastCodon != null) { AlignedCodon firstPeptide = new AlignedCodon(lastCodon.pos1, - lastCodon.pos2, lastCodon.pos3, String.valueOf(seq - .getCharAt(0)), 0); + lastCodon.pos2, lastCodon.pos3, + String.valueOf(seq.getCharAt(0)), 0); toAdd.put(seq, firstPeptide); } else @@ -1317,7 +1321,8 @@ public class AlignmentUtils *
      *
    • One alignment must be nucleotide, and the other protein
      • *
    • At least one pair of sequences must be already mapped, or mappable
      • - *
    • Mappable means the nucleotide translation matches the protein sequence
      • + *
    • Mappable means the nucleotide translation matches the protein + * sequence
      • *
    • The translation may ignore start and stop codons if present in the * nucleotide
      • *
    @@ -1373,9 +1378,10 @@ public class AlignmentUtils return false; } - SequenceI dnaDs = dnaSeq.getDatasetSequence() == null ? dnaSeq : dnaSeq - .getDatasetSequence(); - SequenceI proteinDs = proteinSeq.getDatasetSequence() == null ? proteinSeq + SequenceI dnaDs = dnaSeq.getDatasetSequence() == null ? dnaSeq + : dnaSeq.getDatasetSequence(); + SequenceI proteinDs = proteinSeq.getDatasetSequence() == null + ? proteinSeq : proteinSeq.getDatasetSequence(); for (AlignedCodonFrame mapping : mappings) @@ -1412,8 +1418,7 @@ public class AlignmentUtils * the alignment to check for presence of annotations */ public static void findAddableReferenceAnnotations( - List sequenceScope, - Map labelForCalcId, + List sequenceScope, Map labelForCalcId, final Map> candidates, AlignmentI al) { @@ -1517,8 +1522,8 @@ public class AlignmentUtils /** * Set visibility of alignment annotations of specified types (labels), for - * specified sequences. This supports controls like - * "Show all secondary structure", "Hide all Temp factor", etc. + * specified sequences. This supports controls like "Show all secondary + * structure", "Hide all Temp factor", etc. * * @al the alignment to scan for annotations * @param types @@ -1542,9 +1547,8 @@ public class AlignmentUtils { if (anyType || types.contains(aa.label)) { - if ((aa.sequenceRef != null) - && (forSequences == null || forSequences - .contains(aa.sequenceRef))) + if ((aa.sequenceRef != null) && (forSequences == null + || forSequences.contains(aa.sequenceRef))) { aa.visible = doShow; } @@ -1632,8 +1636,8 @@ public class AlignmentUtils productSeqs = new HashSet(); for (SequenceI seq : products) { - productSeqs.add(seq.getDatasetSequence() == null ? seq : seq - .getDatasetSequence()); + productSeqs.add(seq.getDatasetSequence() == null ? seq + : seq.getDatasetSequence()); } } @@ -1726,8 +1730,9 @@ public class AlignmentUtils /* * add a mapping from CDS to the (unchanged) mapped to range */ - List cdsRange = Collections.singletonList(new int[] { 1, - cdsSeq.getLength() }); + List cdsRange = Collections + .singletonList(new int[] + { 1, cdsSeq.getLength() }); MapList cdsToProteinMap = new MapList(cdsRange, mapList.getToRanges(), mapList.getFromRatio(), mapList.getToRatio()); @@ -1792,12 +1797,11 @@ public class AlignmentUtils // 'CDS|emblcdsacc' // assuming cds version same as dna ?!? - DBRefEntry proteinToCdsRef = new DBRefEntry( - primRef.getSource(), primRef.getVersion(), - cdsSeq.getName()); + DBRefEntry proteinToCdsRef = new DBRefEntry(primRef.getSource(), + primRef.getVersion(), cdsSeq.getName()); // - proteinToCdsRef.setMap(new Mapping(cdsSeqDss, cdsToProteinMap - .getInverse())); + proteinToCdsRef.setMap( + new Mapping(cdsSeqDss, cdsToProteinMap.getInverse())); proteinProduct.addDBRef(proteinToCdsRef); } @@ -1810,8 +1814,8 @@ public class AlignmentUtils } } - AlignmentI cds = new Alignment(cdsSeqs.toArray(new SequenceI[cdsSeqs - .size()])); + AlignmentI cds = new Alignment( + cdsSeqs.toArray(new SequenceI[cdsSeqs.size()])); cds.setDataset(dataset); return cds; @@ -1848,8 +1852,8 @@ public class AlignmentUtils * is this mapping from the whole dna sequence (i.e. CDS)? * allowing for possible stop codon on dna but not peptide */ - int mappedFromLength = MappingUtils.getLength(aMapping.getMap() - .getFromRanges()); + int mappedFromLength = MappingUtils + .getLength(aMapping.getMap().getFromRanges()); int dnaLength = seqDss.getLength(); if (mappedFromLength == dnaLength || mappedFromLength == dnaLength - CODON_LENGTH) @@ -1873,8 +1877,8 @@ public class AlignmentUtils && proteinProduct == mapping.getTo() && seqDss != map.getFromSeq()) { - mappedFromLength = MappingUtils.getLength(mapping.getMap() - .getFromRanges()); + mappedFromLength = MappingUtils + .getLength(mapping.getMap().getFromRanges()); if (mappedFromLength == map.getFromSeq().getLength()) { /* @@ -1972,8 +1976,8 @@ public class AlignmentUtils } else { - System.err - .println("JAL-2154 regression: warning - found (and ignnored a duplicate CDS sequence):" + System.err.println( + "JAL-2154 regression: warning - found (and ignnored a duplicate CDS sequence):" + mtch.toString()); } } @@ -2025,16 +2029,17 @@ public class AlignmentUtils for (DBRefEntry cdsref : direct) { // clone maplist and mapping - MapList cdsposmap = new MapList(Arrays.asList(new int[][] { new int[] - { cdsSeq.getStart(), cdsSeq.getEnd() } }), cdsref.getMap().getMap() - .getToRanges(), 3, 1); - Mapping cdsmap = new Mapping(cdsref.getMap().getTo(), cdsref.getMap() - .getMap()); + MapList cdsposmap = new MapList( + Arrays.asList(new int[][] + { new int[] { cdsSeq.getStart(), cdsSeq.getEnd() } }), + cdsref.getMap().getMap().getToRanges(), 3, 1); + Mapping cdsmap = new Mapping(cdsref.getMap().getTo(), + cdsref.getMap().getMap()); // create dbref DBRefEntry newref = new DBRefEntry(cdsref.getSource(), - cdsref.getVersion(), cdsref.getAccessionId(), new Mapping( - cdsmap.getTo(), cdsposmap)); + cdsref.getVersion(), cdsref.getAccessionId(), + new Mapping(cdsmap.getTo(), cdsposmap)); // and see if we can map to the protein product for this mapping. // onSource is the filtered set of accessions on protein that we are @@ -2159,7 +2164,10 @@ public class AlignmentUtils /** * Returns a mapping from dna to protein by inspecting sequence features of - * type "CDS" on the dna. + * type "CDS" on the dna. A mapping is constructed if the total CDS feature + * length is 3 times the peptide length (optionally after dropping a trailing + * stop codon). This method does not check whether the CDS nucleotide sequence + * translates to the peptide sequence. * * @param dnaSeq * @param proteinSeq @@ -2171,6 +2179,15 @@ public class AlignmentUtils List ranges = findCdsPositions(dnaSeq); int mappedDnaLength = MappingUtils.getLength(ranges); + /* + * if not a whole number of codons, something is wrong, + * abort mapping + */ + if (mappedDnaLength % CODON_LENGTH > 0) + { + return null; + } + int proteinLength = proteinSeq.getLength(); int proteinStart = proteinSeq.getStart(); int proteinEnd = proteinSeq.getEnd(); @@ -2194,8 +2211,12 @@ public class AlignmentUtils if (codesForResidues == (proteinLength + 1)) { // assuming extra codon is for STOP and not in peptide + // todo: check trailing codon is indeed a STOP codon codesForResidues--; + mappedDnaLength -= CODON_LENGTH; + MappingUtils.removeEndPositions(CODON_LENGTH, ranges); } + if (codesForResidues == proteinLength) { proteinRange.add(new int[] { proteinStart, proteinEnd }); @@ -2206,7 +2227,7 @@ public class AlignmentUtils /** * Returns a list of CDS ranges found (as sequence positions base 1), i.e. of - * start/end positions of sequence features of type "CDS" (or a sub-type of + * [start, end] positions of sequence features of type "CDS" (or a sub-type of * CDS in the Sequence Ontology). The ranges are sorted into ascending start * position order, so this method is only valid for linear CDS in the same * sense as the protein product. @@ -2225,7 +2246,6 @@ public class AlignmentUtils return result; } SequenceFeatures.sortFeatures(sfs, true); - int startPhase = 0; for (SequenceFeature sf : sfs) { @@ -2243,7 +2263,7 @@ public class AlignmentUtils */ int begin = sf.getBegin(); int end = sf.getEnd(); - if (result.isEmpty()) + if (result.isEmpty() && phase > 0) { begin += phase; if (begin > end) @@ -2258,16 +2278,6 @@ public class AlignmentUtils } /* - * remove 'startPhase' positions (usually 0) from the first range - * so we begin at the start of a complete codon - */ - if (!result.isEmpty()) - { - // TODO JAL-2022 correctly model start phase > 0 - result.get(0)[0] += startPhase; - } - - /* * Finally sort ranges by start position. This avoids a dependency on * keeping features in order on the sequence (if they are in order anyway, * the sort will have almost no work to do). The implicit assumption is CDS @@ -2445,8 +2455,8 @@ public class AlignmentUtils * are currently ignored here */ String trans = codon.contains("-") ? "-" - : (codon.length() > CODON_LENGTH ? null : ResidueProperties - .codonTranslate(codon)); + : (codon.length() > CODON_LENGTH ? null + : ResidueProperties.codonTranslate(codon)); if (trans != null && !trans.equals(residue)) { String residue3Char = StringUtils @@ -2471,10 +2481,8 @@ public class AlignmentUtils StringBuilder link = new StringBuilder(32); try { - link.append(desc) - .append(" ") - .append(id) - .append("|http://www.ensembl.org/Homo_sapiens/Variation/Summary?v=") + link.append(desc).append(" ").append(id).append( + "|http://www.ensembl.org/Homo_sapiens/Variation/Summary?v=") .append(URLEncoder.encode(id, "UTF-8")); sf.addLink(link.toString()); } catch (UnsupportedEncodingException e) @@ -2796,8 +2804,8 @@ public class AlignmentUtils */ for (SequenceI seq : unaligned.getSequences()) { - List alignedSequences = alignedDatasets.get(seq - .getDatasetSequence()); + List alignedSequences = alignedDatasets + .get(seq.getDatasetSequence()); // TODO: getSequenceAsString() will be deprecated in the future // TODO: need to leave to SequenceI implementor to update gaps seq.setSequence(alignedSequences.get(0).getSequenceAsString()); @@ -2821,7 +2829,8 @@ public class AlignmentUtils * @return */ static SortedMap> buildMappedColumnsMap( - AlignmentI unaligned, AlignmentI aligned, List unmapped) + AlignmentI unaligned, AlignmentI aligned, + List unmapped) { /* * Map will hold, for each aligned column position, a map of @@ -2856,7 +2865,8 @@ public class AlignmentUtils } /** - * Helper method that adds to a map the mapped column positions of a sequence.
    + * Helper method that adds to a map the mapped column positions of a sequence. + *
    * For example if aaTT-Tg-gAAA is mapped to TTTAAA then the map should record * that columns 3,4,6,10,11,12 map to characters T,T,T,A,A,A of the mapped to * sequence. @@ -2885,8 +2895,8 @@ public class AlignmentUtils */ if (seqMap.getTo() == fromSeq.getDatasetSequence()) { - seqMap = new Mapping(seq.getDatasetSequence(), seqMap.getMap() - .getInverse()); + seqMap = new Mapping(seq.getDatasetSequence(), + seqMap.getMap().getInverse()); } int toStart = seq.getStart();