X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fanalysis%2FAlignmentUtils.java;h=2946ba2416922a157fdcebf25164c07f9b77c24a;hb=05951a61acec617599e9a92828b8a498fab6ae32;hp=bef667dfb140e63241fe91059d1d0cac4bd78cb9;hpb=06de78be50c3934158fa1d35ec92ad86b54e959f;p=jalview.git diff --git a/src/jalview/analysis/AlignmentUtils.java b/src/jalview/analysis/AlignmentUtils.java index bef667d..2946ba2 100644 --- a/src/jalview/analysis/AlignmentUtils.java +++ b/src/jalview/analysis/AlignmentUtils.java @@ -74,12 +74,15 @@ import java.util.TreeMap; */ public class AlignmentUtils { - private static final int CODON_LENGTH = 3; private static final String SEQUENCE_VARIANT = "sequence_variant:"; - private static final String ID = "ID"; + /* + * the 'id' attribute is provided for variant features fetched from + * Ensembl using its REST service with JSON format + */ + public static final String VARIANT_ID = "id"; /** * A data model to hold the 'normal' base value at a position, and an optional @@ -128,7 +131,7 @@ public class AlignmentUtils */ public static AlignmentI expandContext(AlignmentI core, int flankSize) { - List sq = new ArrayList(); + List sq = new ArrayList<>(); int maxoffset = 0; for (SequenceI s : core.getSequences()) { @@ -258,7 +261,7 @@ public class AlignmentUtils public static Map> getSequencesByName( AlignmentI al) { - Map> theMap = new LinkedHashMap>(); + Map> theMap = new LinkedHashMap<>(); for (SequenceI seq : al.getSequences()) { String name = seq.getName(); @@ -267,7 +270,7 @@ public class AlignmentUtils List seqs = theMap.get(name); if (seqs == null) { - seqs = new ArrayList(); + seqs = new ArrayList<>(); theMap.put(name, seqs); } seqs.add(seq); @@ -294,8 +297,8 @@ public class AlignmentUtils return false; } - Set mappedDna = new HashSet(); - Set mappedProtein = new HashSet(); + Set mappedDna = new HashSet<>(); + Set mappedProtein = new HashSet<>(); /* * First pass - map sequences where cross-references exist. This include @@ -465,7 +468,7 @@ public class AlignmentUtils { String lastCodon = String.valueOf(cdnaSeqChars, cdnaLength - CODON_LENGTH, CODON_LENGTH).toUpperCase(); - for (String stop : ResidueProperties.STOP) + for (String stop : ResidueProperties.STOP_CODONS) { if (lastCodon.equals(stop)) { @@ -536,7 +539,8 @@ public class AlignmentUtils * allow * in protein to match untranslatable in dna */ final char aaRes = aaSeqChars[aaPos]; - if ((translated == null || "STOP".equals(translated)) && aaRes == '*') + if ((translated == null || ResidueProperties.STOP.equals(translated)) + && aaRes == '*') { continue; } @@ -568,7 +572,8 @@ public class AlignmentUtils if (dnaPos == cdnaSeqChars.length - CODON_LENGTH) { String codon = String.valueOf(cdnaSeqChars, dnaPos, CODON_LENGTH); - if ("STOP".equals(ResidueProperties.codonTranslate(codon))) + if (ResidueProperties.STOP + .equals(ResidueProperties.codonTranslate(codon))) { return true; } @@ -881,7 +886,7 @@ public class AlignmentUtils System.err.println("Wrong alignment type in alignProteinAsDna"); return 0; } - List unmappedProtein = new ArrayList(); + List unmappedProtein = new ArrayList<>(); Map> alignedCodons = buildCodonColumnsMap( protein, dna, unmappedProtein); return alignProteinAs(protein, alignedCodons, unmappedProtein); @@ -1092,7 +1097,7 @@ public class AlignmentUtils * {dnaSequence, {proteinSequence, codonProduct}} at that position. The * comparator keeps the codon positions ordered. */ - Map> alignedCodons = new TreeMap>( + Map> alignedCodons = new TreeMap<>( new CodonComparator()); for (SequenceI dnaSeq : dna.getSequences()) @@ -1138,9 +1143,9 @@ public class AlignmentUtils // TODO delete this ugly hack once JAL-2022 is resolved // i.e. we can model startPhase > 0 (incomplete start codon) - List sequencesChecked = new ArrayList(); + List sequencesChecked = new ArrayList<>(); AlignedCodon lastCodon = null; - Map toAdd = new HashMap(); + Map toAdd = new HashMap<>(); for (Entry> entry : alignedCodons .entrySet()) @@ -1319,7 +1324,7 @@ public class AlignmentUtils Map seqProduct = alignedCodons.get(codon); if (seqProduct == null) { - seqProduct = new HashMap(); + seqProduct = new HashMap<>(); alignedCodons.put(codon, seqProduct); } seqProduct.put(protein, codon); @@ -1456,7 +1461,7 @@ public class AlignmentUtils { continue; } - final List result = new ArrayList(); + final List result = new ArrayList<>(); for (AlignmentAnnotation dsann : datasetAnnotations) { /* @@ -1597,11 +1602,12 @@ public class AlignmentUtils return false; } String name = seq2.getName(); - final DBRefEntry[] xrefs = seq1.getDBRefs(); + final List xrefs = seq1.getDBRefs(); if (xrefs != null) { - for (DBRefEntry xref : xrefs) + for (int ix = 0, nx = xrefs.size(); ix < nx; ix++) { + DBRefEntry xref = xrefs.get(ix); String xrefName = xref.getSource() + "|" + xref.getAccessionId(); // case-insensitive test, consistent with DBRefEntry.equalRef() if (xrefName.equalsIgnoreCase(name)) @@ -1638,13 +1644,13 @@ public class AlignmentUtils throw new IllegalArgumentException( "IMPLEMENTATION ERROR: dataset.getDataset() must be null!"); } - List foundSeqs = new ArrayList(); - List cdsSeqs = new ArrayList(); + List foundSeqs = new ArrayList<>(); + List cdsSeqs = new ArrayList<>(); List mappings = dataset.getCodonFrames(); HashSet productSeqs = null; if (products != null) { - productSeqs = new HashSet(); + productSeqs = new HashSet<>(); for (SequenceI seq : products) { productSeqs.add(seq.getDatasetSequence() == null ? seq : seq @@ -1797,8 +1803,10 @@ public class AlignmentUtils // need to // synthesize an xref. - for (DBRefEntry primRef : dnaDss.getPrimaryDBRefs()) + List primrefs = dnaDss.getPrimaryDBRefs(); + for (int ip = 0, np = primrefs.size(); ip < np; ip++) { + DBRefEntry primRef = primrefs.get(ip); /* * create a cross-reference from CDS to the source sequence's * primary reference and vice versa @@ -1812,7 +1820,6 @@ public class AlignmentUtils dnaSeq.addDBRef(new DBRefEntry(source, version, cdsSeq .getName(), new Mapping(cdsSeqDss, dnaToCdsMap))); - // problem here is that the cross-reference is synthesized - // cdsSeq.getName() may be like 'CDS|dnaaccession' or // 'CDS|emblcdsacc' @@ -1825,7 +1832,6 @@ public class AlignmentUtils .getInverse())); proteinProduct.addDBRef(proteinToCdsRef); } - /* * transfer any features on dna that overlap the CDS */ @@ -1886,7 +1892,7 @@ public class AlignmentUtils * @param seqMappings * the set of mappings involving dnaSeq * @param aMapping - * an initial candidate from seqMappings + * a transcript-to-peptide mapping * @return */ static SequenceI findCdsForProtein(List mappings, @@ -1911,7 +1917,15 @@ public class AlignmentUtils if (mappedFromLength == dnaLength || mappedFromLength == dnaLength - CODON_LENGTH) { - return seqDss; + /* + * if sequence has CDS features, this is a transcript with no UTR + * - do not take this as the CDS sequence! (JAL-2789) + */ + if (seqDss.getFeatures().getFeaturesByOntology(SequenceOntologyI.CDS) + .isEmpty()) + { + return seqDss; + } } /* @@ -1936,10 +1950,12 @@ public class AlignmentUtils { /* * found a 3:1 mapping to the protein product which covers - * the whole dna sequence i.e. is from CDS; finally check it - * is from the dna start sequence + * the whole dna sequence i.e. is from CDS; finally check the CDS + * is mapped from the given dna start sequence */ SequenceI cdsSeq = map.getFromSeq(); + // todo this test is weak if seqMappings contains multiple mappings; + // we get away with it if transcript:cds relationship is 1:1 List dnaToCdsMaps = MappingUtils .findMappingsForSequence(cdsSeq, seqMappings); if (!dnaToCdsMaps.isEmpty()) @@ -2054,16 +2070,20 @@ public class AlignmentUtils protected static List propagateDBRefsToCDS(SequenceI cdsSeq, SequenceI contig, SequenceI proteinProduct, Mapping mapping) { + // gather direct refs from contig congruent with mapping - List direct = new ArrayList(); - HashSet directSources = new HashSet(); - if (contig.getDBRefs() != null) + List direct = new ArrayList<>(); + HashSet directSources = new HashSet<>(); + + List refs = contig.getDBRefs(); + if (refs != null) { - for (DBRefEntry dbr : contig.getDBRefs()) + for (int ib = 0, nb = refs.size(); ib < nb; ib++) { - if (dbr.hasMap() && dbr.getMap().getMap().isTripletMap()) + DBRefEntry dbr = refs.get(ib); + MapList map; + if (dbr.hasMap() && (map = dbr.getMap().getMap()).isTripletMap()) { - MapList map = dbr.getMap().getMap(); // check if map is the CDS mapping if (mapping.getMap().equals(map)) { @@ -2073,21 +2093,22 @@ public class AlignmentUtils } } } - DBRefEntry[] onSource = DBRefUtils.selectRefs( + List onSource = DBRefUtils.selectRefs( proteinProduct.getDBRefs(), directSources.toArray(new String[0])); - List propagated = new ArrayList(); + List propagated = new ArrayList<>(); // and generate appropriate mappings - for (DBRefEntry cdsref : direct) + for (int ic = 0, nc = direct.size(); ic < nc; ic++) { + DBRefEntry cdsref = direct.get(ic); + Mapping m = cdsref.getMap(); // 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()); + m.getMap().getToRanges(), 3, 1); + Mapping cdsmap = new Mapping(m.getTo(),m.getMap()); // create dbref DBRefEntry newref = new DBRefEntry(cdsref.getSource(), @@ -2233,12 +2254,13 @@ public class AlignmentUtils int mappedDnaLength = MappingUtils.getLength(ranges); /* - * if not a whole number of codons, something is wrong, - * abort mapping + * if not a whole number of codons, truncate mapping */ - if (mappedDnaLength % CODON_LENGTH > 0) + int codonRemainder = mappedDnaLength % CODON_LENGTH; + if (codonRemainder > 0) { - return null; + mappedDnaLength -= codonRemainder; + MappingUtils.removeEndPositions(codonRemainder, ranges); } int proteinLength = proteinSeq.getLength(); @@ -2255,7 +2277,7 @@ public class AlignmentUtils proteinStart++; proteinLength--; } - List proteinRange = new ArrayList(); + List proteinRange = new ArrayList<>(); /* * dna length should map to protein (or protein plus stop codon) @@ -2290,7 +2312,7 @@ public class AlignmentUtils */ protected static List findCdsPositions(SequenceI dnaSeq) { - List result = new ArrayList(); + List result = new ArrayList<>(); List sfs = dnaSeq.getFeatures().getFeaturesByOntology( SequenceOntologyI.CDS); @@ -2305,10 +2327,14 @@ public class AlignmentUtils int phase = 0; try { - phase = Integer.parseInt(sf.getPhase()); + String s = sf.getPhase(); + if (s != null) + { + phase = Integer.parseInt(s); + } } catch (NumberFormatException e) { - // ignore + // SwingJS -- need to avoid these. } /* * phase > 0 on first codon means 5' incomplete - skip to the start @@ -2419,19 +2445,26 @@ public class AlignmentUtils /* * variants in first codon base */ - for (DnaVariant var : codonVariants[0]) + for (DnaVariant dnavar : codonVariants[0]) { - if (var.variant != null) + if (dnavar.variant != null) { - String alleles = (String) var.variant.getValue(Gff3Helper.ALLELES); + String alleles = (String) dnavar.variant.getValue(Gff3Helper.ALLELES); if (alleles != null) { for (String base : alleles.split(",")) { - String codon = base + base2 + base3; - if (addPeptideVariant(peptide, peptidePos, residue, var, codon)) + if (!base1.equalsIgnoreCase(base)) { - count++; + String codon = base.toUpperCase() + base2.toLowerCase() + + base3.toLowerCase(); + String canonical = base1.toUpperCase() + base2.toLowerCase() + + base3.toLowerCase(); + if (addPeptideVariant(peptide, peptidePos, residue, dnavar, + codon, canonical)) + { + count++; + } } } } @@ -2450,10 +2483,17 @@ public class AlignmentUtils { for (String base : alleles.split(",")) { - String codon = base1 + base + base3; - if (addPeptideVariant(peptide, peptidePos, residue, var, codon)) + if (!base2.equalsIgnoreCase(base)) { - count++; + String codon = base1.toLowerCase() + base.toUpperCase() + + base3.toLowerCase(); + String canonical = base1.toLowerCase() + base2.toUpperCase() + + base3.toLowerCase(); + if (addPeptideVariant(peptide, peptidePos, residue, var, + codon, canonical)) + { + count++; + } } } } @@ -2472,10 +2512,17 @@ public class AlignmentUtils { for (String base : alleles.split(",")) { - String codon = base1 + base2 + base; - if (addPeptideVariant(peptide, peptidePos, residue, var, codon)) + if (!base3.equalsIgnoreCase(base)) { - count++; + String codon = base1.toLowerCase() + base2.toLowerCase() + + base.toUpperCase(); + String canonical = base1.toLowerCase() + base2.toLowerCase() + + base3.toUpperCase(); + if (addPeptideVariant(peptide, peptidePos, residue, var, + codon, canonical)) + { + count++; + } } } } @@ -2486,20 +2533,22 @@ public class AlignmentUtils } /** - * Helper method that adds a peptide variant feature, provided the given codon - * translates to a value different to the current residue (is a non-synonymous - * variant). ID and clinical_significance attributes of the dna variant (if - * present) are copied to the new feature. + * Helper method that adds a peptide variant feature. ID and + * clinical_significance attributes of the dna variant (if present) are copied + * to the new feature. * * @param peptide * @param peptidePos * @param residue * @param var * @param codon + * the variant codon e.g. aCg + * @param canonical + * the 'normal' codon e.g. aTg * @return true if a feature was added, else false */ static boolean addPeptideVariant(SequenceI peptide, int peptidePos, - String residue, DnaVariant var, String codon) + String residue, DnaVariant var, String codon, String canonical) { /* * get peptide translation of codon e.g. GAT -> D @@ -2507,57 +2556,71 @@ public class AlignmentUtils * e.g. multibase variants or HGMD_MUTATION etc * are currently ignored here */ - String trans = codon.contains("-") ? "-" + String trans = codon.contains("-") ? null : (codon.length() > CODON_LENGTH ? null : ResidueProperties.codonTranslate(codon)); - if (trans != null && !trans.equals(residue)) + if (trans == null) + { + return false; + } + String desc = canonical + "/" + codon; + String featureType = ""; + if (trans.equals(residue)) + { + featureType = SequenceOntologyI.SYNONYMOUS_VARIANT; + } + else if (ResidueProperties.STOP.equals(trans)) + { + featureType = SequenceOntologyI.STOP_GAINED; + } + else { String residue3Char = StringUtils .toSentenceCase(ResidueProperties.aa2Triplet.get(residue)); String trans3Char = StringUtils .toSentenceCase(ResidueProperties.aa2Triplet.get(trans)); - String desc = "p." + residue3Char + peptidePos + trans3Char; - SequenceFeature sf = new SequenceFeature( - SequenceOntologyI.SEQUENCE_VARIANT, desc, peptidePos, - peptidePos, var.getSource()); - StringBuilder attributes = new StringBuilder(32); - String id = (String) var.variant.getValue(ID); - if (id != null) - { - if (id.startsWith(SEQUENCE_VARIANT)) - { - id = id.substring(SEQUENCE_VARIANT.length()); - } - sf.setValue(ID, id); - attributes.append(ID).append("=").append(id); - // TODO handle other species variants JAL-2064 - StringBuilder link = new StringBuilder(32); - try - { - 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) - { - // as if - } - } - String clinSig = (String) var.variant.getValue(CLINICAL_SIGNIFICANCE); - if (clinSig != null) + desc = "p." + residue3Char + peptidePos + trans3Char; + featureType = SequenceOntologyI.NONSYNONYMOUS_VARIANT; + } + SequenceFeature sf = new SequenceFeature(featureType, desc, peptidePos, + peptidePos, var.getSource()); + + StringBuilder attributes = new StringBuilder(32); + String id = (String) var.variant.getValue(VARIANT_ID); + if (id != null) + { + if (id.startsWith(SEQUENCE_VARIANT)) { - sf.setValue(CLINICAL_SIGNIFICANCE, clinSig); - attributes.append(";").append(CLINICAL_SIGNIFICANCE).append("=") - .append(clinSig); + id = id.substring(SEQUENCE_VARIANT.length()); } - peptide.addSequenceFeature(sf); - if (attributes.length() > 0) + sf.setValue(VARIANT_ID, id); + attributes.append(VARIANT_ID).append("=").append(id); + // TODO handle other species variants JAL-2064 + StringBuilder link = new StringBuilder(32); + try { - sf.setAttributes(attributes.toString()); + 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) + { + // as if } - return true; } - return false; + String clinSig = (String) var.variant.getValue(CLINICAL_SIGNIFICANCE); + if (clinSig != null) + { + sf.setValue(CLINICAL_SIGNIFICANCE, clinSig); + attributes.append(";").append(CLINICAL_SIGNIFICANCE).append("=") + .append(clinSig); + } + peptide.addSequenceFeature(sf); + if (attributes.length() > 0) + { + sf.setAttributes(attributes.toString()); + } + return true; } /** @@ -2579,7 +2642,7 @@ public class AlignmentUtils * map from peptide position to all variants of the codon which codes for it * LinkedHashMap ensures we keep the peptide features in sequence order */ - LinkedHashMap[]> variants = new LinkedHashMap[]>(); + LinkedHashMap[]> variants = new LinkedHashMap<>(); List dnaFeatures = dnaSeq.getFeatures() .getFeaturesByOntology(SequenceOntologyI.SEQUENCE_VARIANT); @@ -2638,9 +2701,9 @@ public class AlignmentUtils if (codonVariants == null) { codonVariants = new ArrayList[CODON_LENGTH]; - codonVariants[0] = new ArrayList(); - codonVariants[1] = new ArrayList(); - codonVariants[2] = new ArrayList(); + codonVariants[0] = new ArrayList<>(); + codonVariants[1] = new ArrayList<>(); + codonVariants[2] = new ArrayList<>(); variants.put(peptidePosition, codonVariants); } @@ -2711,19 +2774,25 @@ public class AlignmentUtils SequenceIdMatcher matcher = new SequenceIdMatcher(seqs); if (xrefs != null) { - for (SequenceI xref : xrefs) + // BH 2019.01.25 streamlined this triply nested loop to remove all iterators + + for (int ix = 0, nx = xrefs.length; ix < nx; ix++) { - DBRefEntry[] dbrefs = xref.getDBRefs(); + SequenceI xref = xrefs[ix]; + List dbrefs = xref.getDBRefs(); if (dbrefs != null) { - for (DBRefEntry dbref : dbrefs) + for (int ir = 0, nir = dbrefs.size(); ir < nir; ir++) { - if (dbref.getMap() == null || dbref.getMap().getTo() == null - || dbref.getMap().getTo().isProtein() != isProtein) + DBRefEntry dbref = dbrefs.get(ir); + Mapping map = dbref.getMap(); + SequenceI mto; + if (map == null || (mto = map.getTo()) == null + || mto.isProtein() != isProtein) { continue; } - SequenceI mappedTo = dbref.getMap().getTo(); + SequenceI mappedTo = mto; SequenceI match = matcher.findIdMatch(mappedTo); if (match == null) { @@ -2764,7 +2833,7 @@ public class AlignmentUtils /* * fancy case - aligning via mappings between sequences */ - List unmapped = new ArrayList(); + List unmapped = new ArrayList<>(); Map> columnMap = buildMappedColumnsMap( unaligned, aligned, unmapped); int width = columnMap.size(); @@ -2839,7 +2908,7 @@ public class AlignmentUtils } // map from dataset sequence to alignment sequence(s) - Map> alignedDatasets = new HashMap>(); + Map> alignedDatasets = new HashMap<>(); for (SequenceI seq : aligned.getSequences()) { SequenceI ds = seq.getDatasetSequence(); @@ -2902,7 +2971,7 @@ public class AlignmentUtils * {unalignedSequence, characterPerSequence} at that position. * TreeMap keeps the entries in ascending column order. */ - SortedMap> map = new TreeMap>(); + SortedMap> map = new TreeMap<>(); /* * record any sequences that have no mapping so can't be realigned @@ -3007,7 +3076,7 @@ public class AlignmentUtils Map seqsMap = map.get(fromCol); if (seqsMap == null) { - seqsMap = new HashMap(); + seqsMap = new HashMap<>(); map.put(fromCol, seqsMap); } seqsMap.put(seq, seq.getCharAt(mappedCharPos - toStart));