+
+ /**
+ * Returns the DNA codon positions (base 1) for the given position (base 1) in
+ * a mapped protein sequence, or null if no mapping is found.
+ *
+ * Intended for use in aligning cDNA to match aligned protein. Only the first
+ * mapping found is returned, so not suitable for use if multiple protein
+ * sequences are mapped to the same cDNA (but aligning cDNA as protein is
+ * ill-defined for this case anyway).
+ *
+ * @param seq
+ * the DNA dataset sequence
+ * @param aaPos
+ * residue position (base 1) in a protein sequence
+ * @return
+ */
+ public int[] getDnaPosition(SequenceI seq, int aaPos)
+ {
+ /*
+ * Adapted from markMappedRegion().
+ */
+ MapList ml = null;
+ int i = 0;
+ for (SequenceToSequenceMapping ssm : mappings)
+ {
+ if (ssm.fromSeq == seq)
+ {
+ ml = getdnaToProt()[i];
+ break;
+ }
+ i++;
+ }
+ return ml == null ? null : ml.locateInFrom(aaPos, aaPos);
+ }
+
+ /**
+ * Convenience method to return the first aligned sequence in the given
+ * alignment whose dataset has a mapping with the given (aligned or dataset)
+ * sequence.
+ *
+ * @param seq
+ *
+ * @param al
+ * @return
+ */
+ public SequenceI findAlignedSequence(SequenceI seq, AlignmentI al)
+ {
+ /*
+ * Search mapped protein ('to') sequences first.
+ */
+ for (SequenceToSequenceMapping ssm : mappings)
+ {
+ if (ssm.fromSeq == seq || ssm.fromSeq == seq.getDatasetSequence())
+ {
+ for (SequenceI sourceAligned : al.getSequences())
+ {
+ if (ssm.mapping.to == sourceAligned.getDatasetSequence()
+ || ssm.mapping.to == sourceAligned)
+ {
+ return sourceAligned;
+ }
+ }
+ }
+ }
+
+ /*
+ * Then try mapped dna sequences.
+ */
+ for (SequenceToSequenceMapping ssm : mappings)
+ {
+ if (ssm.mapping.to == seq
+ || ssm.mapping.to == seq.getDatasetSequence())
+ {
+ for (SequenceI sourceAligned : al.getSequences())
+ {
+ if (ssm.fromSeq == sourceAligned.getDatasetSequence())
+ {
+ return sourceAligned;
+ }
+ }
+ }
+ }
+
+ return null;
+ }
+
+ /**
+ * Returns the region in the target sequence's dataset that is mapped to the
+ * given position (base 1) in the query sequence's dataset. The region is a
+ * set of start/end position pairs.
+ *
+ * @param target
+ * @param query
+ * @param queryPos
+ * @return
+ */
+ public int[] getMappedRegion(SequenceI target, SequenceI query,
+ int queryPos)
+ {
+ SequenceI targetDs = target.getDatasetSequence() == null ? target
+ : target.getDatasetSequence();
+ SequenceI queryDs = query.getDatasetSequence() == null ? query : query
+ .getDatasetSequence();
+ if (targetDs == null || queryDs == null /*|| dnaToProt == null*/)
+ {
+ return null;
+ }
+ for (SequenceToSequenceMapping ssm : mappings)
+ {
+ /*
+ * try mapping from target to query
+ */
+ if (ssm.fromSeq == targetDs && ssm.mapping.to == queryDs)
+ {
+ int[] codon = ssm.mapping.map.locateInFrom(queryPos, queryPos);
+ if (codon != null)
+ {
+ return codon;
+ }
+ }
+ /*
+ * else try mapping from query to target
+ */
+ else if (ssm.fromSeq == queryDs && ssm.mapping.to == targetDs)
+ {
+ int[] codon = ssm.mapping.map.locateInTo(queryPos, queryPos);
+ if (codon != null)
+ {
+ return codon;
+ }
+ }
+ }
+ return null;
+ }
+
+ /**
+ * Returns the mapped DNA codons for the given position in a protein sequence,
+ * or null if no mapping is found. Returns a list of (e.g.) ['g', 'c', 't']
+ * codons. There may be more than one codon mapped to the protein if (for
+ * example), there are mappings to cDNA variants.
+ *
+ * @param protein
+ * the peptide dataset sequence
+ * @param aaPos
+ * residue position (base 1) in the peptide sequence
+ * @return
+ */
+ public List<char[]> getMappedCodons(SequenceI protein, int aaPos)
+ {
+ MapList ml = null;
+ SequenceI dnaSeq = null;
+ List<char[]> result = new ArrayList<char[]>();
+
+ for (SequenceToSequenceMapping ssm : mappings)
+ {
+ if (ssm.mapping.to == protein
+ && ssm.mapping.getMap().getFromRatio() == 3)
+ {
+ ml = ssm.mapping.map;
+ dnaSeq = ssm.fromSeq;
+
+ int[] codonPos = ml.locateInFrom(aaPos, aaPos);
+ if (codonPos == null)
+ {
+ return null;
+ }
+
+ /*
+ * Read off the mapped nucleotides (converting to position base 0)
+ */
+ codonPos = MappingUtils.flattenRanges(codonPos);
+ char[] dna = dnaSeq.getSequence();
+ int start = dnaSeq.getStart();
+ result.add(new char[] { dna[codonPos[0] - start],
+ dna[codonPos[1] - start], dna[codonPos[2] - start] });
+ }
+ }
+ return result.isEmpty() ? null : result;
+ }
+
+ /**
+ * Returns any mappings found which are from the given sequence, and to
+ * distinct sequences.
+ *
+ * @param seq
+ * @return
+ */
+ public List<Mapping> getMappingsFromSequence(SequenceI seq)
+ {
+ List<Mapping> result = new ArrayList<Mapping>();
+ List<SequenceI> related = new ArrayList<SequenceI>();
+ SequenceI seqDs = seq.getDatasetSequence();
+ seqDs = seqDs != null ? seqDs : seq;
+
+ for (SequenceToSequenceMapping ssm : mappings)
+ {
+ final Mapping mapping = ssm.mapping;
+ if (ssm.fromSeq == seqDs)
+ {
+ if (!related.contains(mapping.to))
+ {
+ result.add(mapping);
+ related.add(mapping.to);
+ }
+ }
+ }
+ return result;
+ }
+
+ /**
+ * Test whether the given sequence is substitutable for one or more dummy
+ * sequences in this mapping
+ *
+ * @param map
+ * @param seq
+ * @return
+ */
+ public boolean isRealisableWith(SequenceI seq)
+ {
+ return realiseWith(seq, false) > 0;
+ }
+
+ /**
+ * Replace any matchable mapped dummy sequences with the given real one.
+ * Returns the count of sequence mappings instantiated.
+ *
+ * @param seq
+ * @return
+ */
+ public int realiseWith(SequenceI seq)
+ {
+ return realiseWith(seq, true);
+ }
+
+ /**
+ * Returns the number of mapped dummy sequences that could be replaced with
+ * the given real sequence.
+ *
+ * @param seq
+ * a dataset sequence
+ * @param doUpdate
+ * if true, performs replacements, else only counts
+ * @return
+ */
+ protected int realiseWith(SequenceI seq, boolean doUpdate)
+ {
+ SequenceI ds = seq.getDatasetSequence() != null ? seq
+ .getDatasetSequence() : seq;
+ int count = 0;
+
+ /*
+ * check for replaceable DNA ('map from') sequences
+ */
+ for (SequenceToSequenceMapping ssm : mappings)
+ {
+ SequenceI dna = ssm.fromSeq;
+ if (dna instanceof SequenceDummy
+ && dna.getName().equals(ds.getName()))
+ {
+ Mapping mapping = ssm.mapping;
+ int mapStart = mapping.getMap().getFromLowest();
+ int mapEnd = mapping.getMap().getFromHighest();
+ boolean mappable = couldRealiseSequence(dna, ds, mapStart, mapEnd);
+ if (mappable)
+ {
+ count++;
+ if (doUpdate)
+ {
+ // TODO: new method ? ds.realise(dna);
+ // might want to copy database refs as well
+ ds.setSequenceFeatures(dna.getSequenceFeatures());
+ // dnaSeqs[i] = ds;
+ ssm.fromSeq = ds;
+ System.out.println("Realised mapped sequence " + ds.getName());
+ }
+ }
+ }
+
+ /*
+ * check for replaceable protein ('map to') sequences
+ */
+ Mapping mapping = ssm.mapping;
+ SequenceI prot = mapping.getTo();
+ int mapStart = mapping.getMap().getToLowest();
+ int mapEnd = mapping.getMap().getToHighest();
+ boolean mappable = couldRealiseSequence(prot, ds, mapStart, mapEnd);
+ if (mappable)
+ {
+ count++;
+ if (doUpdate)
+ {
+ // TODO: new method ? ds.realise(dna);
+ // might want to copy database refs as well
+ ds.setSequenceFeatures(dna.getSequenceFeatures());
+ ssm.mapping.setTo(ds);
+ }
+ }
+ }
+ return count;
+ }
+
+ /**
+ * Helper method to test whether a 'real' sequence could replace a 'dummy'
+ * sequence in the map. The criteria are that they have the same name, and
+ * that the mapped region overlaps the candidate sequence.
+ *
+ * @param existing
+ * @param replacement
+ * @param mapStart
+ * @param mapEnd
+ * @return
+ */
+ protected static boolean couldRealiseSequence(SequenceI existing,
+ SequenceI replacement, int mapStart, int mapEnd)
+ {
+ if (existing instanceof SequenceDummy
+ && !(replacement instanceof SequenceDummy)
+ && existing.getName().equals(replacement.getName()))
+ {
+ int start = replacement.getStart();
+ int end = replacement.getEnd();
+ boolean mappingOverlapsSequence = (mapStart >= start && mapStart <= end)
+ || (mapEnd >= start && mapEnd <= end);
+ if (mappingOverlapsSequence)
+ {
+ return true;
+ }
+ }
+ return false;
+ }
+
+ /**
+ * Change any mapping to the given sequence to be to its dataset sequence
+ * instead. For use when mappings are created before their referenced
+ * sequences are instantiated, for example when parsing GFF data.
+ *
+ * @param seq
+ */
+ public void updateToDataset(SequenceI seq)
+ {
+ if (seq == null || seq.getDatasetSequence() == null)
+ {
+ return;
+ }
+ SequenceI ds = seq.getDatasetSequence();
+
+ for (SequenceToSequenceMapping ssm : mappings)
+ /*
+ * 'from' sequences
+ */
+ {
+ if (ssm.fromSeq == seq)
+ {
+ ssm.fromSeq = ds;
+ }
+
+ /*
+ * 'to' sequences
+ */
+ if (ssm.mapping.to == seq)
+ {
+ ssm.mapping.to = ds;
+ }
+ }
+ }
+
+ /**
+ * Answers true if this object contains no mappings
+ *
+ * @return
+ */
+ public boolean isEmpty()
+ {
+ return mappings.isEmpty();
+ }
+
+ /**
+ * Method for debug / inspection purposes only, may change in future
+ */
+ @Override
+ public String toString()
+ {
+ return mappings == null ? "null" : mappings.toString();
+ }
+
+ /**
+ * Returns the first mapping found that is between 'fromSeq' and 'toSeq', or
+ * null if none found
+ *
+ * @param fromSeq
+ * aligned or dataset sequence
+ * @param toSeq
+ * aligned or dataset sequence
+ * @return
+ */
+ public Mapping getMappingBetween(SequenceI fromSeq, SequenceI toSeq)
+ {
+ SequenceI dssFrom = fromSeq.getDatasetSequence() == null ? fromSeq
+ : fromSeq.getDatasetSequence();
+ SequenceI dssTo = toSeq.getDatasetSequence() == null ? toSeq : toSeq
+ .getDatasetSequence();
+
+ for (SequenceToSequenceMapping mapping : mappings)
+ {
+ SequenceI from = mapping.fromSeq;
+ SequenceI to = mapping.mapping.to;
+ if ((from == dssFrom && to == dssTo)
+ || (from == dssTo && to == dssFrom))
+ {
+ return mapping.mapping;
+ }
+ }
+ return null;
+ }
+
+ /**
+ * Returns a hashcode derived from the list of sequence mappings
+ *
+ * @see SequenceToSequenceMapping#hashCode()
+ * @see AbstractList#hashCode()
+ */
+ @Override
+ public int hashCode()
+ {
+ return this.mappings.hashCode();
+ }
+
+ /**
+ * Two AlignedCodonFrame objects are equal if they hold the same ordered list
+ * of mappings
+ *
+ * @see SequenceToSequenceMapping#
+ */
+ @Override
+ public boolean equals(Object obj)
+ {
+ if (!(obj instanceof AlignedCodonFrame))
+ {
+ return false;
+ }
+ return this.mappings.equals(((AlignedCodonFrame) obj).mappings);
+ }
+
+ public List<SequenceToSequenceMapping> getMappings()
+ {
+ return mappings;
+ }