X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fext%2Fensembl%2FEnsemblSeqProxy.java;h=9109fb22f7eef8756c465a02fc1ac56bd229eaaf;hb=853624fb32058cccc544ae7d13af6ad4b0800b6c;hp=0bfeda1121211a99e65976ef798a8853065f2182;hpb=949fed115506ff00c221669e096546f9c39a6ace;p=jalview.git diff --git a/src/jalview/ext/ensembl/EnsemblSeqProxy.java b/src/jalview/ext/ensembl/EnsemblSeqProxy.java index 0bfeda1..9109fb2 100644 --- a/src/jalview/ext/ensembl/EnsemblSeqProxy.java +++ b/src/jalview/ext/ensembl/EnsemblSeqProxy.java @@ -1,6 +1,8 @@ package jalview.ext.ensembl; import jalview.analysis.AlignmentUtils; +import jalview.analysis.Dna; +import jalview.bin.Cache; import jalview.datamodel.Alignment; import jalview.datamodel.AlignmentI; import jalview.datamodel.DBRefEntry; @@ -11,12 +13,11 @@ import jalview.datamodel.SequenceI; import jalview.exceptions.JalviewException; import jalview.io.FastaFile; import jalview.io.FileParse; -import jalview.io.gff.SequenceOntology; -import jalview.schemes.ResidueProperties; +import jalview.io.gff.SequenceOntologyFactory; +import jalview.io.gff.SequenceOntologyI; +import jalview.util.Comparison; import jalview.util.DBRefUtils; import jalview.util.MapList; -import jalview.util.MappingUtils; -import jalview.util.StringUtils; import java.io.IOException; import java.net.MalformedURLException; @@ -25,53 +26,48 @@ import java.util.ArrayList; import java.util.Arrays; import java.util.Collections; import java.util.Comparator; -import java.util.LinkedHashMap; import java.util.List; -import java.util.Map.Entry; /** * Base class for Ensembl sequence fetchers * + * @see http://rest.ensembl.org/documentation/info/sequence_id * @author gmcarstairs */ public abstract class EnsemblSeqProxy extends EnsemblRestClient { - private static final List CROSS_REFERENCES = Arrays - .asList(new String[] { "CCDS" }); - - protected static final String CONSEQUENCE_TYPE = "consequence_type"; + private static final String ALLELES = "alleles"; protected static final String PARENT = "Parent"; protected static final String ID = "ID"; - /* - * this needs special handling, as it isA sequence_variant in the - * Sequence Ontology, but behaves in Ensembl as if it isA transcript - */ - protected static final String NMD_VARIANT = "NMD_transcript_variant"; - protected static final String NAME = "Name"; + protected static final String DESCRIPTION = "description"; + + /* + * enum for 'type' parameter to the /sequence REST service + */ public enum EnsemblSeqType { /** - * type=genomic for the full dna including introns + * type=genomic to fetch full dna including introns */ GENOMIC("genomic"), /** - * type=cdna for transcribed dna including UTRs + * type=cdna to fetch coding dna including UTRs */ CDNA("cdna"), /** - * type=cds for coding dna excluding UTRs + * type=cds to fetch coding dna excluding UTRs */ CDS("cds"), /** - * type=protein for the peptide product sequence + * type=protein to fetch peptide product sequence */ PROTEIN("protein"); @@ -114,10 +110,19 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient } /** - * Constructor + * Default constructor (to use rest.ensembl.org) */ public EnsemblSeqProxy() { + super(); + } + + /** + * Constructor given the target domain to fetch data from + */ + public EnsemblSeqProxy(String d) + { + super(d); } /** @@ -127,7 +132,6 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient @Override public AlignmentI getSequenceRecords(String query) throws Exception { - long now = System.currentTimeMillis(); // TODO use a String... query vararg instead? // danger: accession separator used as a regex here, a string elsewhere @@ -156,17 +160,16 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient + " chunks. Unexpected problem (" + r.getLocalizedMessage() + ")"; System.err.println(msg); - if (alignment != null) - { - break; // return what we got - } - else - { - throw new JalviewException(msg, r); - } + r.printStackTrace(); + break; } } + if (alignment == null) + { + return null; + } + /* * fetch and transfer genomic sequence features, * fetch protein product and add as cross-reference @@ -181,8 +184,6 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient getCrossReferences(seq); } - System.out.println(getClass().getName() + " took " - + (System.currentTimeMillis() - now) + "ms to fetch"); return alignment; } @@ -208,7 +209,7 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient * get 'dummy' genomic sequence with exon, cds and variation features */ SequenceI genomicSequence = null; - EnsemblOverlap gffFetcher = new EnsemblOverlap(); + EnsemblFeatures gffFetcher = new EnsemblFeatures(getDomain()); EnsemblFeatureType[] features = getFeaturesToFetch(); AlignmentI geneFeatures = gffFetcher.getSequenceRecords(accId, features); @@ -258,10 +259,11 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient String accId = querySeq.getName(); try { - AlignmentI protein = new EnsemblProtein().getSequenceRecords(accId); + AlignmentI protein = new EnsemblProtein(getDomain()) + .getSequenceRecords(accId); if (protein == null || protein.getHeight() == 0) { - System.out.println("Failed to retrieve protein for " + accId); + System.out.println("No protein product found for " + accId); return; } SequenceI proteinSeq = protein.getSequenceAt(0); @@ -272,19 +274,63 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient proteinSeq.createDatasetSequence(); querySeq.createDatasetSequence(); - MapList mapList = mapCdsToProtein(querySeq, proteinSeq); + MapList mapList = AlignmentUtils + .mapCdsToProtein(querySeq, proteinSeq); if (mapList != null) { - Mapping map = new Mapping(proteinSeq.getDatasetSequence(), mapList); - DBRefEntry dbr = new DBRefEntry(getDbSource(), getDbVersion(), - accId, map); + // clunky: ensure Uniprot xref if we have one is on mapped sequence + SequenceI ds = proteinSeq.getDatasetSequence(); + // TODO: Verify ensp primary ref is on proteinSeq.getDatasetSequence() + Mapping map = new Mapping(ds, mapList); + DBRefEntry dbr = new DBRefEntry(getDbSource(), + getEnsemblDataVersion(), proteinSeq.getName(), map); querySeq.getDatasetSequence().addDBRef(dbr); - + DBRefEntry[] uprots = DBRefUtils.selectRefs(ds.getDBRefs(), + new String[] { DBRefSource.UNIPROT }); + DBRefEntry[] upxrefs = DBRefUtils.selectRefs(querySeq.getDBRefs(), + new String[] { DBRefSource.UNIPROT }); + if (uprots != null) + { + for (DBRefEntry up : uprots) + { + // locate local uniprot ref and map + List upx = DBRefUtils.searchRefs(upxrefs, + up.getAccessionId()); + DBRefEntry upxref; + if (upx.size() != 0) + { + upxref = upx.get(0); + + if (upx.size() > 1) + { + Cache.log + .warn("Implementation issue - multiple uniprot acc on product sequence."); + } + } + else + { + upxref = new DBRefEntry(DBRefSource.UNIPROT, + getEnsemblDataVersion(), up.getAccessionId()); + } + + Mapping newMap = new Mapping(ds, mapList); + upxref.setVersion(getEnsemblDataVersion()); + upxref.setMap(newMap); + if (upx.size() == 0) + { + // add the new uniprot ref + querySeq.getDatasetSequence().addDBRef(upxref); + } + + } + } + /* - * compute peptide variants from dna variants and add as - * sequence features on the protein sequence ta-da + * copy exon features to protein, compute peptide variants from dna + * variants and add as features on the protein sequence ta-da */ - computeProteinFeatures(querySeq, proteinSeq, mapList); + AlignmentUtils + .computeProteinFeatures(querySeq, proteinSeq, mapList); } } catch (Exception e) { @@ -295,8 +341,7 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient } /** - * Get Uniprot and PDB xrefs from Ensembl, and attach them to the protein - * sequence + * Get database xrefs from Ensembl, and attach them to the sequence * * @param seq */ @@ -307,118 +352,20 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient seq = seq.getDatasetSequence(); } - EnsemblXref xrefFetcher = new EnsemblXref(); - List xrefs = xrefFetcher.getCrossReferences(seq.getName(), - getCrossReferenceDatabases()); + EnsemblXref xrefFetcher = new EnsemblXref(getDomain(), getDbSource(), + getEnsemblDataVersion()); + List xrefs = xrefFetcher.getCrossReferences(seq.getName()); for (DBRefEntry xref : xrefs) { seq.addDBRef(xref); } - } - - /** - * Returns a list of database names to be used when fetching cross-references. - * - * @return - */ - protected List getCrossReferenceDatabases() - { - return CROSS_REFERENCES; - } - - /** - * Returns a mapping from dna to protein by inspecting sequence features of - * type "CDS" on the dna. - * - * @param dnaSeq - * @param proteinSeq - * @return - */ - protected MapList mapCdsToProtein(SequenceI dnaSeq, SequenceI proteinSeq) - { - List ranges = new ArrayList(50); - - int mappedDnaLength = getCdsRanges(dnaSeq, ranges); - - int proteinLength = proteinSeq.getLength(); - List proteinRange = new ArrayList(); - int proteinStart = 1; - - /* - * incomplete start codon may mean X at start of peptide - * we ignore both for mapping purposes - */ - if (proteinSeq.getCharAt(0) == 'X') - { - proteinStart = 2; - proteinLength--; - } - proteinRange.add(new int[] { proteinStart, proteinLength }); /* - * dna length should map to protein (or protein plus stop codon) + * and add a reference to itself */ - int codesForResidues = mappedDnaLength / 3; - if (codesForResidues == proteinLength - || codesForResidues == (proteinLength + 1)) - { - return new MapList(ranges, proteinRange, 3, 1); - } - return null; - } - - /** - * Adds CDS ranges to the ranges list, and returns the total length mapped. - * - * No need to worry about reverse strand dna here since the retrieved sequence - * is as transcribed (reverse complement for reverse strand), i.e in the same - * sense as the peptide. - * - * @param dnaSeq - * @param ranges - * @return - */ - protected int getCdsRanges(SequenceI dnaSeq, List ranges) - { - SequenceFeature[] sfs = dnaSeq.getSequenceFeatures(); - if (sfs == null) - { - return 0; - } - int mappedDnaLength = 0; - for (SequenceFeature sf : sfs) - { - /* - * process a CDS feature (or a sub-type of CDS) - */ - if (SequenceOntology.getInstance().isA(sf.getType(), SequenceOntology.CDS)) - { - int phase = 0; - try { - phase = Integer.parseInt(sf.getPhase()); - } catch (NumberFormatException e) - { - // ignore - } - /* - * phase > 0 on first codon means 5' incomplete - skip to the start - * of the next codon; example ENST00000496384 - */ - int begin = sf.getBegin(); - int end = sf.getEnd(); - if (ranges.isEmpty() && phase > 0) - { - begin += phase; - if (begin > end) - { - continue; // shouldn't happen? - } - } - ranges.add(new int[] { begin, end }); - mappedDnaLength += Math.abs(end - begin) + 1; - } - } - return mappedDnaLength; + DBRefEntry self = new DBRefEntry(getDbSource(), + getEnsemblDataVersion(), seq.getName()); + seq.addDBRef(self); } /** @@ -440,6 +387,11 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient throw new JalviewException("ENSEMBL Rest API not available."); } FileParse fp = getSequenceReader(ids); + if (fp == null) + { + return alignment; + } + FastaFile fr = new FastaFile(fp); if (fr.hasWarningMessage()) { @@ -464,9 +416,8 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient if (fr.getSeqs().size() > 0) { - AlignmentI seqal = new Alignment( - fr.getSeqsAsArray()); - for (SequenceI sq:seqal.getSequences()) + AlignmentI seqal = new Alignment(fr.getSeqsAsArray()); + for (SequenceI sq : seqal.getSequences()) { if (sq.getDescription() == null) { @@ -476,7 +427,9 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient if (ids.contains(name) || ids.contains(name.replace("ENSP", "ENST"))) { - DBRefUtils.parseToDbRef(sq, DBRefSource.ENSEMBL, "0", name); + DBRefEntry dbref = DBRefUtils.parseToDbRef(sq, getDbSource(), + getEnsemblDataVersion(), name); + sq.addDBRef(dbref); } } if (alignment == null) @@ -505,7 +458,7 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient * multiple ids go in the POST body instead */ StringBuffer urlstring = new StringBuffer(128); - urlstring.append(SEQUENCE_ID_URL); + urlstring.append(getDomain() + "/sequence/id"); if (ids.size() == 1) { urlstring.append("/").append(ids.get(0)); @@ -579,7 +532,7 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient * the start position of the sequence we are mapping to * @return */ - protected MapList getGenomicRanges(SequenceI sourceSequence, + protected MapList getGenomicRangesFromFeatures(SequenceI sourceSequence, String accId, int start) { SequenceFeature[] sfs = sourceSequence.getSequenceFeatures(); @@ -596,7 +549,7 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient int mappedLength = 0; int direction = 1; // forward boolean directionSet = false; - + for (SequenceFeature sf : sfs) { /* @@ -605,27 +558,28 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient */ if (identifiesSequence(sf, accId)) { - int strand = sf.getStrand(); - - if (directionSet && strand != direction) - { - // abort - mix of forward and backward + int strand = sf.getStrand(); + strand = strand == 0 ? 1 : strand; // treat unknown as forward + + if (directionSet && strand != direction) + { + // abort - mix of forward and backward System.err.println("Error: forward and backward strand for " + accId); - return null; - } - direction = strand; - directionSet = true; - - /* - * add to CDS ranges, semi-sorted forwards/backwards - */ - if (strand < 0) - { - regions.add(0, new int[] { sf.getEnd(), sf.getBegin() }); - } - else - { + return null; + } + direction = strand; + directionSet = true; + + /* + * add to CDS ranges, semi-sorted forwards/backwards + */ + if (strand < 0) + { + regions.add(0, new int[] { sf.getEnd(), sf.getBegin() }); + } + else + { regions.add(new int[] { sf.getBegin(), sf.getEnd() }); } mappedLength += Math.abs(sf.getEnd() - sf.getBegin() + 1); @@ -640,7 +594,7 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient } } } - + if (regions.isEmpty()) { System.out.println("Failed to identify target sequence for " + accId @@ -653,10 +607,10 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient * (havana / ensembl_havana) */ Collections.sort(regions, new RangeSorter(direction == 1)); - - List to = new ArrayList(); - to.add(new int[] { start, start + mappedLength - 1 }); - + + List to = Arrays.asList(new int[] { start, + start + mappedLength - 1 }); + return new MapList(regions, to, 1, 1); } @@ -692,34 +646,102 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient * @param mapping * mapping from the sequence feature's coordinates to the target * sequence + * @param forwardStrand */ protected void transferFeature(SequenceFeature sf, - SequenceI targetSequence, MapList mapping) + SequenceI targetSequence, MapList mapping, boolean forwardStrand) { int start = sf.getBegin(); int end = sf.getEnd(); int[] mappedRange = mapping.locateInTo(start, end); - + if (mappedRange != null) { SequenceFeature copy = new SequenceFeature(sf); copy.setBegin(Math.min(mappedRange[0], mappedRange[1])); copy.setEnd(Math.max(mappedRange[0], mappedRange[1])); + if (".".equals(copy.getFeatureGroup())) + { + copy.setFeatureGroup(getDbSource()); + } targetSequence.addSequenceFeature(copy); /* - * for sequence_variant, make an additional feature with consequence + * for sequence_variant on reverse strand, have to convert the allele + * values to their complements */ - if (SequenceOntology.getInstance().isSequenceVariant(sf.getType())) + if (!forwardStrand + && SequenceOntologyFactory.getInstance().isA(sf.getType(), + SequenceOntologyI.SEQUENCE_VARIANT)) { - String consequence = (String) sf.getValue(CONSEQUENCE_TYPE); - if (consequence != null) - { - SequenceFeature sf2 = new SequenceFeature("consequence", - consequence, copy.getBegin(), copy.getEnd(), 0f, - null); - targetSequence.addSequenceFeature(sf2); - } + reverseComplementAlleles(copy); + } + } + } + + /** + * Change the 'alleles' value of a feature by converting to complementary + * bases, and also update the feature description to match + * + * @param sf + */ + static void reverseComplementAlleles(SequenceFeature sf) + { + final String alleles = (String) sf.getValue(ALLELES); + if (alleles == null) + { + return; + } + StringBuilder complement = new StringBuilder(alleles.length()); + for (String allele : alleles.split(",")) + { + reverseComplementAllele(complement, allele); + } + String comp = complement.toString(); + sf.setValue(ALLELES, comp); + sf.setDescription(comp); + + /* + * replace value of "alleles=" in sf.ATTRIBUTES as well + * so 'output as GFF' shows reverse complement alleles + */ + String atts = sf.getAttributes(); + if (atts != null) + { + atts = atts.replace(ALLELES + "=" + alleles, ALLELES + "=" + comp); + sf.setAttributes(atts); + } + } + + /** + * Makes the 'reverse complement' of the given allele and appends it to the + * buffer, after a comma separator if not the first + * + * @param complement + * @param allele + */ + static void reverseComplementAllele(StringBuilder complement, + String allele) + { + if (complement.length() > 0) + { + complement.append(","); + } + + /* + * some 'alleles' are actually descriptive terms + * e.g. HGMD_MUTATION, PhenCode_variation + * - we don't want to 'reverse complement' these + */ + if (!Comparison.isNucleotideSequence(allele, true)) + { + complement.append(allele); + } + else + { + for (int i = allele.length() - 1; i >= 0; i--) + { + complement.append(Dna.getComplement(allele.charAt(i))); } } } @@ -740,15 +762,22 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient return false; } + // long start = System.currentTimeMillis(); SequenceFeature[] sfs = sourceSequence.getSequenceFeatures(); - MapList mapping = getGenomicRanges(sourceSequence, accessionId, - targetSequence.getStart()); + MapList mapping = getGenomicRangesFromFeatures(sourceSequence, + accessionId, targetSequence.getStart()); if (mapping == null) { return false; } - return transferFeatures(sfs, targetSequence, mapping, accessionId); + boolean result = transferFeatures(sfs, targetSequence, mapping, + accessionId); + // System.out.println("transferFeatures (" + (sfs.length) + " --> " + // + targetSequence.getSequenceFeatures().length + ") to " + // + targetSequence.getName() + // + " took " + (System.currentTimeMillis() - start) + "ms"); + return result; } /** @@ -768,25 +797,18 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient final boolean forwardStrand = mapping.isFromForwardStrand(); /* - * sort features by start position (descending if reverse strand) - * before transferring (in forwards order) to the target sequence + * sort features by start position (which corresponds to end + * position descending if reverse strand) so as to add them in + * 'forwards' order to the target sequence */ - Arrays.sort(features, new Comparator() - { - @Override - public int compare(SequenceFeature o1, SequenceFeature o2) - { - int c = Integer.compare(o1.getBegin(), o2.getBegin()); - return forwardStrand ? c : -c; - } - }); + sortFeatures(features, forwardStrand); boolean transferred = false; for (SequenceFeature sf : features) { if (retainFeature(sf, parentId)) { - transferFeature(sf, targetSequence, mapping); + transferFeature(sf, targetSequence, mapping, forwardStrand); transferred = true; } } @@ -794,6 +816,33 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient } /** + * Sort features by start position ascending (if on forward strand), or end + * position descending (if on reverse strand) + * + * @param features + * @param forwardStrand + */ + protected static void sortFeatures(SequenceFeature[] features, + final boolean forwardStrand) + { + Arrays.sort(features, new Comparator() + { + @Override + public int compare(SequenceFeature o1, SequenceFeature o2) + { + if (forwardStrand) + { + return Integer.compare(o1.getBegin(), o2.getBegin()); + } + else + { + return Integer.compare(o2.getEnd(), o1.getEnd()); + } + } + }); + } + + /** * Answers true if the feature type is one we want to keep for the sequence. * Some features are only retrieved in order to identify the sequence range, * and may then be discarded as redundant information (e.g. "CDS" feature for @@ -847,11 +896,13 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient String type, String parentId) { List result = new ArrayList(); - + SequenceFeature[] sfs = sequence.getSequenceFeatures(); - if (sfs != null) { - SequenceOntology so = SequenceOntology.getInstance(); - for (SequenceFeature sf :sfs) { + if (sfs != null) + { + SequenceOntologyI so = SequenceOntologyFactory.getInstance(); + for (SequenceFeature sf : sfs) + { if (so.isA(sf.getType(), type)) { String parent = (String) sf.getValue(PARENT); @@ -866,224 +917,6 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient } /** - * Maps exon features from dna to protein, and computes variants in peptide - * product generated by variants in dna, and adds them as sequence_variant - * features on the protein sequence. Returns the number of variant features - * added. - * - * @param dnaSeq - * @param peptide - * @param dnaToProtein - */ - static int computeProteinFeatures(SequenceI dnaSeq, - SequenceI peptide, MapList dnaToProtein) - { - while (dnaSeq.getDatasetSequence() != null) - { - dnaSeq = dnaSeq.getDatasetSequence(); - } - while (peptide.getDatasetSequence() != null) - { - peptide = peptide.getDatasetSequence(); - } - - AlignmentUtils.transferFeatures(dnaSeq, peptide, dnaToProtein, - SequenceOntology.EXON); - - LinkedHashMap variants = buildDnaVariantsMap( - dnaSeq, dnaToProtein); - - /* - * scan codon variations, compute peptide variants and add to peptide sequence - */ - int count = 0; - for (Entry variant : variants.entrySet()) - { - int peptidePos = variant.getKey(); - String[][] codonVariants = variant.getValue(); - String residue = String.valueOf(peptide.getCharAt(peptidePos - 1)); // 0-based - List peptideVariants = computePeptideVariants(codonVariants, - residue); - if (!peptideVariants.isEmpty()) - { - String desc = StringUtils.listToDelimitedString(peptideVariants, - ", "); - SequenceFeature sf = new SequenceFeature( - SequenceOntology.SEQUENCE_VARIANT, desc, peptidePos, - peptidePos, 0f, null); - peptide.addSequenceFeature(sf); - count++; - } - } - return count; - } - - /** - * Builds a map whose key is position in the protein sequence, and value is an - * array of all variants for the coding codon positions - * - * @param dnaSeq - * @param dnaToProtein - * @return - */ - static LinkedHashMap buildDnaVariantsMap( - SequenceI dnaSeq, MapList dnaToProtein) - { - /* - * map from peptide position to all variant features of the codon for it - * LinkedHashMap ensures we add the peptide features in sequence order - */ - LinkedHashMap variants = new LinkedHashMap(); - SequenceOntology so = SequenceOntology.getInstance(); - - SequenceFeature[] dnaFeatures = dnaSeq.getSequenceFeatures(); - if (dnaFeatures == null) - { - return variants; - } - - int dnaStart = dnaSeq.getStart(); - int[] lastCodon = null; - int lastPeptidePostion = 0; - - /* - * build a map of codon variations for peptides - */ - for (SequenceFeature sf : dnaFeatures) - { - int dnaCol = sf.getBegin(); - if (dnaCol != sf.getEnd()) - { - // not handling multi-locus variant features - continue; - } - if (so.isSequenceVariant(sf.getType())) - { - int[] mapsTo = dnaToProtein.locateInTo(dnaCol, dnaCol); - if (mapsTo == null) - { - // feature doesn't lie within coding region - continue; - } - int peptidePosition = mapsTo[0]; - String[][] codonVariants = variants.get(peptidePosition); - if (codonVariants == null) - { - codonVariants = new String[3][]; - variants.put(peptidePosition, codonVariants); - } - - /* - * extract dna variants to a string array - */ - String alls = (String) sf.getValue("alleles"); - if (alls == null) - { - continue; - } - String[] alleles = alls.split(","); - - /* - * get this peptides codon positions e.g. [3, 4, 5] or [4, 7, 10] - */ - int[] codon = peptidePosition == lastPeptidePostion ? lastCodon - : MappingUtils.flattenRanges(dnaToProtein.locateInFrom( - peptidePosition, peptidePosition)); - lastPeptidePostion = peptidePosition; - lastCodon = codon; - - /* - * save nucleotide (and this variant) for each codon position - */ - for (int codonPos = 0; codonPos < 3; codonPos++) - { - String nucleotide = String.valueOf(dnaSeq - .getCharAt(codon[codonPos] - dnaStart)); - if (codon[codonPos] == dnaCol) - { - /* - * record current dna base and its alleles - */ - String[] dnaVariants = new String[alleles.length + 1]; - dnaVariants[0] = nucleotide; - System.arraycopy(alleles, 0, dnaVariants, 1, alleles.length); - codonVariants[codonPos] = dnaVariants; - } - else if (codonVariants[codonPos] == null) - { - /* - * record current dna base only - * (at least until we find any variation and overwrite it) - */ - codonVariants[codonPos] = new String[] { nucleotide }; - } - } - } - } - return variants; - } - - /** - * Returns a sorted, non-redundant list of all peptide translations generated - * by the given dna variants, excluding the current residue value - * - * @param codonVariants - * an array of base values (acgtACGT) for codon positions 1, 2, 3 - * @param residue - * the current residue translation - * @return - */ - static List computePeptideVariants( - String[][] codonVariants, String residue) - { - List result = new ArrayList(); - for (String base1 : codonVariants[0]) - { - for (String base2 : codonVariants[1]) - { - for (String base3 : codonVariants[2]) - { - String codon = base1 + base2 + base3; - // TODO: report frameshift/insertion/deletion - // and multiple-base variants?! - String peptide = codon.contains("-") ? "-" : ResidueProperties - .codonTranslate(codon); - if (peptide != null && !result.contains(peptide) - && !peptide.equalsIgnoreCase(residue)) - { - result.add(peptide); - } - } - } - } - - /* - * sort alphabetically with STOP at the end - */ - Collections.sort(result, new Comparator() - { - - @Override - public int compare(String o1, String o2) - { - if ("STOP".equals(o1)) - { - return 1; - } - else if ("STOP".equals(o2)) - { - return -1; - } - else - { - return o1.compareTo(o2); - } - } - }); - return result; - } - - /** * Answers true if the feature type is either 'NMD_transcript_variant' or * 'transcript' or one of its sub-types in the Sequence Ontology. This is * needed because NMD_transcript_variant behaves like 'transcript' in Ensembl @@ -1095,7 +928,8 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient */ public static boolean isTranscript(String featureType) { - return NMD_VARIANT.equals(featureType) - || SequenceOntology.getInstance().isA(featureType, SequenceOntology.TRANSCRIPT); + return SequenceOntologyI.NMD_TRANSCRIPT_VARIANT.equals(featureType) + || SequenceOntologyFactory.getInstance().isA(featureType, + SequenceOntologyI.TRANSCRIPT); } }