+
+ /**
+ * attempt to extract coding region and product from a feature and properly decorate it with annotations.
+ * @param feature coding feature
+ * @param sourceDb source database for the EMBLXML
+ * @param seqs place where sequences go
+ * @param dna parent dna sequence for this record
+ * @param noPeptide flag for generation of Peptide sequence objects
+ */
+ private void parseCodingFeature(EmblFeature feature, String sourceDb, Vector seqs, Sequence dna, boolean noPeptide)
+ {
+ boolean isEmblCdna = sourceDb.equals(DBRefSource.EMBLCDS);
+ // extract coding region(s)
+ jalview.datamodel.Mapping map = null;
+ int[] exon = null;
+ if (feature.locations != null && feature.locations.size() > 0)
+ {
+ for (Enumeration locs = feature.locations.elements(); locs
+ .hasMoreElements();)
+ {
+ EmblFeatureLocations loc = (EmblFeatureLocations) locs
+ .nextElement();
+ int[] se = loc.getElementRanges(accession);
+ if (exon == null)
+ {
+ exon = se;
+ }
+ else
+ {
+ int[] t = new int[exon.length + se.length];
+ System.arraycopy(exon, 0, t, 0, exon.length);
+ System.arraycopy(se, 0, t, exon.length, se.length);
+ exon = t;
+ }
+ }
+ }
+ String prseq = null;
+ String prname = new String();
+ String prid = null;
+ Hashtable vals = new Hashtable();
+ int prstart = 1;
+ // get qualifiers
+ if (feature.getQualifiers() != null
+ && feature.getQualifiers().size() > 0)
+ {
+ for (Iterator quals = feature.getQualifiers().iterator(); quals
+ .hasNext();)
+ {
+ Qualifier q = (Qualifier) quals.next();
+ if (q.getName().equals("translation"))
+ {
+ StringBuffer prsq = new StringBuffer(q.getValues()[0]);
+ int p = prsq.indexOf(" ");
+ while (p > -1)
+ {
+ prsq.deleteCharAt(p);
+ p = prsq.indexOf(" ", p);
+ }
+ prseq = prsq.toString();
+ prsq = null;
+
+ }
+ else if (q.getName().equals("protein_id"))
+ {
+ prid = q.getValues()[0];
+ }
+ else if (q.getName().equals("codon_start"))
+ {
+ prstart = Integer.parseInt(q.getValues()[0]);
+ }
+ else if (q.getName().equals("product"))
+ {
+ prname = q.getValues()[0];
+ }
+ else
+ {
+ // throw anything else into the additional properties hash
+ vals.put(q.getName(), q.getValues().toString());
+ }
+ }
+ }
+ Sequence product = null;
+ if (prseq != null && prname != null && prid != null)
+ {
+ // extract proteins.
+ product = new Sequence(sourceDb + "|" + "EMBLCDS|" + prid
+ +((prname.length()==0) ? "" : " " + prname), prseq, prstart, prstart
+ + prseq.length() - 1);
+ product.setDescription("Protein Product from " + sourceDb);
+
+ if (!noPeptide)
+ {
+ // Protein is also added to vector of sequences returned
+ seqs.add(product);
+ }
+ // we have everything - create the mapping and perhaps the protein
+ // sequence
+ if (exon == null || exon.length == 0)
+ {
+ System.err
+ .println("Implementation Notice: EMBLCDS records not properly supported yet - Making up the CDNA region of this sequence... may be incorrect ("
+ + sourceDb + ":" + getAccession() + ")");
+ if (prseq.length() * 3 == dna.getSequence().length)
+ {
+ // this might occur for CDS sequences where no features are
+ // marked.
+ exon = new int[]
+ { dna.getStart(), dna.getEnd() };
+ map = new jalview.datamodel.Mapping(product, exon,
+ new int[]
+ { prstart, prstart + prseq.length() - 1 }, 3, 1);
+ }
+ if ((prseq.length() + 1) * 3 == dna.getSequence().length)
+ {
+ exon = new int[]
+ { dna.getStart(), dna.getEnd() - 3 };
+ map = new jalview.datamodel.Mapping(product, exon,
+ new int[]
+ { prstart, prstart + prseq.length() - 1 }, 3, 1);
+ }
+ }
+ else
+ {
+ if (isEmblCdna)
+ {
+ // TODO: Add a DbRef back to the parent EMBL sequence with the exon
+ // map
+
+ // make a new feature annotating the coding contig
+ }
+ else
+ {
+ map = new jalview.datamodel.Mapping(product, exon,
+ new int[]
+ { prstart, prstart + prseq.length() - 1 }, 3, 1);
+ // reconstruct the EMBLCDS entry
+ DBRefEntry pcdnaref = new DBRefEntry();
+ pcdnaref.setAccessionId(prid);
+ pcdnaref.setSource(DBRefSource.EMBLCDS);
+ pcdnaref.setVersion(getVersion()); // same as parent EMBL version.
+ jalview.util.MapList mp = new jalview.util.MapList(new int[]
+ { 1+(prstart-1)*3, 1+(prstart-1)*3 + (prseq.length()-1)*3 }, new int[] { prstart, prstart+prseq.length() - 1 }, 3, 1);
+ pcdnaref.setMap(new Mapping(mp));
+ if (product!=null)
+ product.addDBRef(pcdnaref);
+
+ }
+ }
+ // add cds feature to dna seq - this may include the stop codon
+ for (int xint = 0; exon != null && xint < exon.length; xint += 2)
+ {
+ SequenceFeature sf = new SequenceFeature();
+ sf.setBegin(exon[xint]);
+ sf.setEnd(exon[xint + 1]);
+ sf.setType(feature.getName());
+ sf.setFeatureGroup(sourceDb);
+ sf.setDescription("Exon " + (1 + xint) + " for protein '"
+ + prname + "' EMBLCDS:" + prid);
+ sf.setValue(FeatureProperties.EXONPOS, new Integer(1 + xint));
+ sf.setValue(FeatureProperties.EXONPRODUCT, prname);
+ if (vals != null && vals.size() > 0)
+ {
+ Enumeration kv = vals.elements();
+ while (kv.hasMoreElements())
+ {
+ Object key = kv.nextElement();
+ if (key != null)
+ sf.setValue(key.toString(), vals.get(key));
+ }
+ }
+ dna.addSequenceFeature(sf);
+ }
+ }
+ // add dbRefs to sequence
+ if (feature.dbRefs != null && feature.dbRefs.size() > 0)
+ {
+ for (Iterator dbr = feature.dbRefs.iterator(); dbr.hasNext();)
+ {
+ DBRefEntry ref = (DBRefEntry) dbr.next();
+ ref.setSource(jalview.util.DBRefUtils.getCanonicalName(ref
+ .getSource()));
+ // Hard code the kind of protein product accessions that EMBL cite
+ if (ref.getSource().equals(
+ jalview.datamodel.DBRefSource.UNIPROT))
+ {
+ ref.setMap(map);
+ /*if (map.getTo()!=null)
+ {
+ map.getTo().setName(map.getTo().getName()+"|"+ref.getSource()+"|"+ref.getAccessionId());
+ }*/
+ }
+ if (product != null)
+ {
+ DBRefEntry pref = new DBRefEntry(ref.getSource(), ref
+ .getVersion(), ref.getAccessionId());
+ pref.setMap(null); // reference is direct
+ product.addDBRef(pref);
+ // Add converse mapping reference
+ if (map != null)
+ {
+ Mapping pmap = new Mapping(dna, map.getMap().getInverse());
+ pref = new DBRefEntry(sourceDb, getVersion(), this
+ .getAccession());
+ pref.setMap(pmap);
+ if (map.getTo()!=null)
+ {
+ map.getTo().addDBRef(pref);
+ }
+ }
+ }
+ dna.addDBRef(ref);
+ }
+ }
+ }