+ /**
+ * Reads the XML reply from file and unmarshals it to Java objects. Answers a
+ * (possibly empty) list of <code>EntryType</code> objects.
+ *
+ * is
+ *
+ * @return
+ */
+ List<EntryType> getEmblEntries(InputStream is)
+ {
+ List<EntryType> entries = new ArrayList<>();
+ try
+ {
+ JAXBContext jc = JAXBContext.newInstance("jalview.xml.binding.embl");
+ XMLStreamReader streamReader = XMLInputFactory.newInstance()
+ .createXMLStreamReader(is);
+ javax.xml.bind.Unmarshaller um = jc.createUnmarshaller();
+ JAXBElement<ROOT> rootElement = um.unmarshal(streamReader,
+ ROOT.class);
+ ROOT root = rootElement.getValue();
+
+ /*
+ * document root contains either "entry" or "entrySet"
+ */
+ if (root == null)
+ {
+ return entries;
+ }
+ if (root.getEntrySet() != null)
+ {
+ entries = root.getEntrySet().getEntry();
+ }
+ else if (root.getEntry() != null)
+ {
+ entries.add(root.getEntry());
+ }
+ } catch (JAXBException | XMLStreamException
+ | FactoryConfigurationError e)
+ {
+ e.printStackTrace();
+ }
+ return entries;
+ }
+
+ /**
+ * A helper method to parse XML data and construct a sequence, with any
+ * available database references and features
+ *
+ * @param emprefx
+ * @param entry
+ * @param peptides
+ * @return
+ */
+ SequenceI getSequence(String sourceDb, EntryType entry,
+ List<SequenceI> peptides)
+ {
+ String seqString = entry.getSequence();
+ if (seqString == null)
+ {
+ return null;
+ }
+ seqString = seqString.replace(" ", "").replace("\n", "").replace("\t",
+ "");
+ String accession = entry.getAccession();
+ SequenceI dna = new Sequence(sourceDb + "|" + accession, seqString);
+
+ dna.setDescription(entry.getDescription());
+ String sequenceVersion = String.valueOf(entry.getVersion().intValue());
+ DBRefEntry selfRref = new DBRefEntry(sourceDb, sequenceVersion,
+ accession);
+ dna.addDBRef(selfRref);
+ selfRref.setMap(
+ new Mapping(null, new int[]
+ { 1, dna.getLength() }, new int[] { 1, dna.getLength() }, 1,
+ 1));
+
+ /*
+ * add db references
+ */
+ List<XrefType> xrefs = entry.getXref();
+ if (xrefs != null)
+ {
+ for (XrefType xref : xrefs)
+ {
+ String acc = xref.getId();
+ String source = DBRefUtils.getCanonicalName(xref.getDb());
+ String version = xref.getSecondaryId();
+ if (version == null || "".equals(version))
+ {
+ version = "0";
+ }
+ dna.addDBRef(new DBRefEntry(source, version, acc));
+ }
+ }
+
+ SequenceIdMatcher matcher = new SequenceIdMatcher(peptides);
+ try
+ {
+ List<Feature> features = entry.getFeature();
+ if (features != null)
+ {
+ for (Feature feature : features)
+ {
+ if (FeatureProperties.isCodingFeature(sourceDb,
+ feature.getName()))
+ {
+ parseCodingFeature(entry, feature, sourceDb, dna, peptides,
+ matcher);
+ }
+ }
+ }
+ } catch (Exception e)
+ {
+ System.err.println("EMBL Record Features parsing error!");
+ System.err
+ .println("Please report the following to help@jalview.org :");
+ System.err.println("EMBL Record " + accession);
+ System.err.println("Resulted in exception: " + e.getMessage());
+ e.printStackTrace(System.err);
+ }
+
+ return dna;
+ }
+
+ /**
+ * Extracts coding region and product from a CDS feature and decorates it with
+ * annotations
+ *
+ * @param entry
+ * @param feature
+ * @param sourceDb
+ * @param dna
+ * @param peptides
+ * @param matcher
+ */
+ void parseCodingFeature(EntryType entry, Feature feature, String sourceDb,
+ SequenceI dna, List<SequenceI> peptides,
+ SequenceIdMatcher matcher)
+ {
+ final boolean isEmblCdna = sourceDb.equals(DBRefSource.EMBLCDS);
+ final String accession = entry.getAccession();
+ final String sequenceVersion = entry.getVersion().toString();
+
+ int[] exons = getCdsRanges(entry.getAccession(), feature);
+
+ String translation = null;
+ String proteinName = "";
+ String proteinId = null;
+ Map<String, String> vals = new Hashtable<>();
+
+ /*
+ * codon_start 1/2/3 in EMBL corresponds to phase 0/1/2 in CDS
+ * (phase is required for CDS features in GFF3 format)
+ */
+ int codonStart = 1;
+
+ /*
+ * parse qualifiers, saving protein translation, protein id,
+ * codon start position, product (name), and 'other values'
+ */
+ if (feature.getQualifier() != null)
+ {
+ for (Qualifier q : feature.getQualifier())
+ {
+ String qname = q.getName();
+ String value = q.getValue();
+ value = value == null ? ""
+ : value.trim().replace(" ", "").replace("\n", "")
+ .replace("\t", "");
+ if (qname.equals("translation"))
+ {
+ translation = value;
+ }
+ else if (qname.equals("protein_id"))
+ {
+ proteinId = value;
+ }
+ else if (qname.equals("codon_start"))
+ {
+ try
+ {
+ codonStart = Integer.parseInt(value.trim());
+ } catch (NumberFormatException e)
+ {
+ System.err.println("Invalid codon_start in XML for "
+ + entry.getAccession() + ": " + e.getMessage());
+ }
+ }
+ else if (qname.equals("product"))
+ {
+ // sometimes name is returned e.g. for V00488
+ proteinName = value;
+ }
+ else
+ {
+ // throw anything else into the additional properties hash
+ if (!"".equals(value))
+ {
+ vals.put(qname, value);
+ }
+ }
+ }
+ }
+
+ DBRefEntry proteinToEmblProteinRef = null;
+ exons = MappingUtils.removeStartPositions(codonStart - 1, exons);
+
+ SequenceI product = null;
+ Mapping dnaToProteinMapping = null;
+ if (translation != null && proteinName != null && proteinId != null)
+ {
+ int translationLength = translation.length();
+
+ /*
+ * look for product in peptides list, if not found, add it
+ */
+ product = matcher.findIdMatch(proteinId);
+ if (product == null)
+ {
+ product = new Sequence(proteinId, translation, 1,
+ translationLength);
+ product.setDescription(((proteinName.length() == 0)
+ ? "Protein Product from " + sourceDb
+ : proteinName));
+ peptides.add(product);
+ matcher.add(product);
+ }
+
+ // we have everything - create the mapping and perhaps the protein
+ // sequence
+ if (exons == null || exons.length == 0)
+ {
+ /*
+ * workaround until we handle dna location for CDS sequence
+ * e.g. location="X53828.1:60..1058" correctly
+ */
+ System.err.println(
+ "Implementation Notice: EMBLCDS records not properly supported yet - Making up the CDNA region of this sequence... may be incorrect ("
+ + sourceDb + ":" + entry.getAccession() + ")");
+ int dnaLength = dna.getLength();
+ if (translationLength * 3 == (1 - codonStart + dnaLength))
+ {
+ System.err.println(
+ "Not allowing for additional stop codon at end of cDNA fragment... !");
+ // this might occur for CDS sequences where no features are marked
+ exons = new int[] { dna.getStart() + (codonStart - 1),
+ dna.getEnd() };
+ dnaToProteinMapping = new Mapping(product, exons,
+ new int[]
+ { 1, translationLength }, 3, 1);
+ }
+ if ((translationLength + 1) * 3 == (1 - codonStart + dnaLength))
+ {
+ System.err.println(
+ "Allowing for additional stop codon at end of cDNA fragment... will probably cause an error in VAMSAs!");
+ exons = new int[] { dna.getStart() + (codonStart - 1),
+ dna.getEnd() - 3 };
+ dnaToProteinMapping = new Mapping(product, exons,
+ new int[]
+ { 1, translationLength }, 3, 1);
+ }
+ }
+ else
+ {
+ // Trim the exon mapping if necessary - the given product may only be a
+ // fragment of a larger protein. (EMBL:AY043181 is an example)
+
+ if (isEmblCdna)
+ {
+ // TODO: Add a DbRef back to the parent EMBL sequence with the exon
+ // map
+ // if given a dataset reference, search dataset for parent EMBL
+ // sequence if it exists and set its map
+ // make a new feature annotating the coding contig
+ }
+ else
+ {
+ // final product length truncation check
+ int[] cdsRanges = adjustForProteinLength(translationLength,
+ exons);
+ dnaToProteinMapping = new Mapping(product, cdsRanges,
+ new int[]
+ { 1, translationLength }, 3, 1);
+ if (product != null)
+ {
+ /*
+ * make xref with mapping from protein to EMBL dna
+ */
+ DBRefEntry proteinToEmblRef = new DBRefEntry(DBRefSource.EMBL,
+ sequenceVersion, proteinId,
+ new Mapping(dnaToProteinMapping.getMap().getInverse()));
+ product.addDBRef(proteinToEmblRef);
+
+ /*
+ * make xref from protein to EMBLCDS; we assume here that the
+ * CDS sequence version is same as dna sequence (?!)
+ */
+ MapList proteinToCdsMapList = new MapList(
+ new int[]
+ { 1, translationLength },
+ new int[]
+ { 1 + (codonStart - 1),
+ (codonStart - 1) + 3 * translationLength },
+ 1, 3);
+ DBRefEntry proteinToEmblCdsRef = new DBRefEntry(
+ DBRefSource.EMBLCDS, sequenceVersion, proteinId,
+ new Mapping(proteinToCdsMapList));
+ product.addDBRef(proteinToEmblCdsRef);
+
+ /*
+ * make 'direct' xref from protein to EMBLCDSPROTEIN
+ */
+ proteinToEmblProteinRef = new DBRefEntry(proteinToEmblCdsRef);
+ proteinToEmblProteinRef.setSource(DBRefSource.EMBLCDSProduct);
+ proteinToEmblProteinRef.setMap(null);
+ product.addDBRef(proteinToEmblProteinRef);
+ }
+ }
+ }
+
+ /*
+ * add cds features to dna sequence
+ */
+ String cds = feature.getName(); // "CDS"
+ for (int xint = 0; exons != null
+ && xint < exons.length - 1; xint += 2)
+ {
+ int exonStart = exons[xint];
+ int exonEnd = exons[xint + 1];
+ int begin = Math.min(exonStart, exonEnd);
+ int end = Math.max(exonStart, exonEnd);
+ int exonNumber = xint / 2 + 1;
+ String desc = String.format("Exon %d for protein '%s' EMBLCDS:%s",
+ exonNumber, proteinName, proteinId);
+
+ SequenceFeature sf = makeCdsFeature(cds, desc, begin, end, sourceDb,
+ vals);
+
+ sf.setEnaLocation(feature.getLocation());
+ boolean forwardStrand = exonStart <= exonEnd;
+ sf.setStrand(forwardStrand ? "+" : "-");
+ sf.setPhase(String.valueOf(codonStart - 1));
+ sf.setValue(FeatureProperties.EXONPOS, exonNumber);
+ sf.setValue(FeatureProperties.EXONPRODUCT, proteinName);
+
+ dna.addSequenceFeature(sf);
+ }
+ }
+
+ /*
+ * add feature dbRefs to sequence, and mappings for Uniprot xrefs
+ */
+ boolean hasUniprotDbref = false;
+ List<XrefType> xrefs = feature.getXref();
+ if (xrefs != null)
+ {
+ boolean mappingUsed = false;
+ for (XrefType xref : xrefs)
+ {
+ /*
+ * ensure UniProtKB/Swiss-Prot converted to UNIPROT
+ */
+ String source = DBRefUtils.getCanonicalName(xref.getDb());
+ String version = xref.getSecondaryId();
+ if (version == null || "".equals(version))
+ {
+ version = "0";
+ }
+ DBRefEntry dbref = new DBRefEntry(source, version, xref.getId());
+ DBRefEntry proteinDbRef = new DBRefEntry(source, version,
+ dbref.getAccessionId());
+ if (source.equals(DBRefSource.UNIPROT))
+ {
+ String proteinSeqName = DBRefSource.UNIPROT + "|"
+ + dbref.getAccessionId();
+ if (dnaToProteinMapping != null
+ && dnaToProteinMapping.getTo() != null)
+ {
+ if (mappingUsed)
+ {
+ /*
+ * two or more Uniprot xrefs for the same CDS -
+ * each needs a distinct Mapping (as to a different sequence)
+ */
+ dnaToProteinMapping = new Mapping(dnaToProteinMapping);
+ }
+ mappingUsed = true;
+
+ /*
+ * try to locate the protein mapped to (possibly by a
+ * previous CDS feature); if not found, construct it from
+ * the EMBL translation
+ */
+ SequenceI proteinSeq = matcher.findIdMatch(proteinSeqName);
+ if (proteinSeq == null)
+ {
+ proteinSeq = new Sequence(proteinSeqName,
+ product.getSequenceAsString());
+ matcher.add(proteinSeq);
+ proteinSeq.setDescription(product.getDescription());
+ peptides.add(proteinSeq);
+ }
+ dnaToProteinMapping.setTo(proteinSeq);
+ dnaToProteinMapping.setMappedFromId(proteinId);
+ proteinSeq.addDBRef(proteinDbRef);
+ dbref.setMap(dnaToProteinMapping);
+ }
+ hasUniprotDbref = true;
+ }
+ if (product != null)
+ {
+ /*
+ * copy feature dbref to our protein product
+ */
+ DBRefEntry pref = proteinDbRef;
+ pref.setMap(null); // reference is direct
+ product.addDBRef(pref);
+ // Add converse mapping reference
+ if (dnaToProteinMapping != null)
+ {
+ Mapping pmap = new Mapping(dna,
+ dnaToProteinMapping.getMap().getInverse());
+ pref = new DBRefEntry(sourceDb, sequenceVersion, accession);
+ pref.setMap(pmap);
+ if (dnaToProteinMapping.getTo() != null)
+ {
+ dnaToProteinMapping.getTo().addDBRef(pref);
+ }
+ }
+ }
+ dna.addDBRef(dbref);
+ }
+ }
+
+ /*
+ * if we have a product (translation) but no explicit Uniprot dbref
+ * (example: EMBL AAFI02000057 protein_id EAL65544.1)
+ * then construct mappings to an assumed EMBLCDSPROTEIN accession
+ */
+ if (!hasUniprotDbref && product != null)
+ {
+ if (proteinToEmblProteinRef == null)
+ {
+ // assuming CDSPROTEIN sequence version = dna version (?!)
+ proteinToEmblProteinRef = new DBRefEntry(DBRefSource.EMBLCDSProduct,
+ sequenceVersion, proteinId);
+ }
+ product.addDBRef(proteinToEmblProteinRef);
+
+ if (dnaToProteinMapping != null
+ && dnaToProteinMapping.getTo() != null)
+ {
+ DBRefEntry dnaToEmblProteinRef = new DBRefEntry(
+ DBRefSource.EMBLCDSProduct, sequenceVersion, proteinId);
+ dnaToEmblProteinRef.setMap(dnaToProteinMapping);
+ dnaToProteinMapping.setMappedFromId(proteinId);
+ dna.addDBRef(dnaToEmblProteinRef);
+ }
+ }
+ }
+