mappedExons = makeExonSequences(ds, acf,
+ newMapping);
+ if (!mappedExons.isEmpty())
+ {
+ exonSequences.addAll(mappedExons);
+ newMappings.add(newMapping);
+ }
+ }
+ }
+ AlignmentI al = new Alignment(
+ exonSequences.toArray(new SequenceI[exonSequences.size()]));
+ al.setDataset(null);
+
+ /*
+ * Replace the old mappings with the new ones
+ */
+ mappings.clear();
+ mappings.addAll(newMappings);
+
+ return al;
+ }
+
+ /**
+ * Helper method to make exon-only sequences and populate their mappings to
+ * protein products
+ *
+ * For example, if ggCCaTTcGAg has mappings [3, 4, 6, 7, 9, 10] to protein
+ * then generate a sequence CCTTGA with mapping [1, 6] to the same protein
+ * residues
+ *
+ * Typically eukaryotic dna will include exons encoding for a single peptide
+ * sequence i.e. return a single result. Bacterial dna may have overlapping
+ * exon mappings coding for multiple peptides so return multiple results
+ * (example EMBL KF591215).
+ *
+ * @param dnaSeq
+ * a dna dataset sequence
+ * @param mapping
+ * containing one or more mappings of the sequence to protein
+ * @param newMapping
+ * the new mapping to populate, from the exon-only sequences to their
+ * mapped protein sequences
+ * @return
+ */
+ protected static List makeExonSequences(SequenceI dnaSeq,
+ AlignedCodonFrame mapping, AlignedCodonFrame newMapping)
+ {
+ List exonSequences = new ArrayList();
+ List seqMappings = mapping.getMappingsForSequence(dnaSeq);
+ final char[] dna = dnaSeq.getSequence();
+ for (Mapping seqMapping : seqMappings)
+ {
+ StringBuilder newSequence = new StringBuilder(dnaSeq.getLength());
+
+ /*
+ * Get the codon regions as { [2, 5], [7, 12], [14, 14] etc }
+ */
+ final List dnaExonRanges = seqMapping.getMap().getFromRanges();
+ for (int[] range : dnaExonRanges)
+ {
+ for (int pos = range[0]; pos <= range[1]; pos++)
+ {
+ newSequence.append(dna[pos - 1]);
+ }
+ }
+
+ SequenceI exon = new Sequence(dnaSeq.getName(),
+ newSequence.toString());
+
+ /*
+ * Locate any xrefs to CDS database on the protein product and attach to
+ * the CDS sequence. Also add as a sub-token of the sequence name.
+ */
+ // default to "CDS" if we can't locate an actual gene id
+ String cdsAccId = FeatureProperties
+ .getCodingFeature(DBRefSource.EMBL);
+ DBRefEntry[] cdsRefs = DBRefUtils.selectRefs(seqMapping.getTo()
+ .getDBRef(), DBRefSource.CODINGDBS);
+ if (cdsRefs != null)
+ {
+ for (DBRefEntry cdsRef : cdsRefs)
+ {
+ exon.addDBRef(new DBRefEntry(cdsRef));
+ cdsAccId = cdsRef.getAccessionId();
+ }
+ }
+ exon.setName(exon.getName() + "|" + cdsAccId);
+ exon.createDatasetSequence();
+
+ /*
+ * Build new mappings - from the same protein regions, but now to
+ * contiguous exons
+ */
+ List exonRange = new ArrayList();
+ exonRange.add(new int[] { 1, newSequence.length() });
+ MapList map = new MapList(exonRange, seqMapping.getMap()
+ .getToRanges(), 3, 1);
+ newMapping.addMap(exon.getDatasetSequence(), seqMapping.getTo(), map);
+ MapList cdsToDnaMap = new MapList(dnaExonRanges, exonRange, 1, 1);
+ newMapping.addMap(dnaSeq, exon.getDatasetSequence(), cdsToDnaMap);
+
+ exonSequences.add(exon);
+ }
+ return exonSequences;
+ }
}