*/
package jalview.datamodel;
-import jalview.util.MapList;
-import jalview.util.MappingUtils;
-
import java.util.AbstractList;
import java.util.ArrayList;
import java.util.List;
+import jalview.util.MapList;
+import jalview.util.MappingUtils;
+
/**
* Stores mapping between the columns of a protein alignment and a DNA alignment
* and a list of individual codon to amino acid mappings between sequences.
{
return mapping;
}
+
+ /**
+ * Returns true if the mapping covers the full length of the given sequence.
+ * This allows us to distinguish the CDS that codes for a protein from
+ * another overlapping CDS in the parent dna sequence.
+ *
+ * @param seq
+ * @return
+ */
+ public boolean covers(SequenceI seq)
+ {
+ List<int[]> mappedRanges = null;
+ MapList mapList = mapping.getMap();
+ if (fromSeq == seq || fromSeq == seq.getDatasetSequence())
+ {
+ mappedRanges = mapList.getFromRanges();
+ }
+ else if (mapping.to == seq || mapping.to == seq.getDatasetSequence())
+ {
+ mappedRanges = mapList.getToRanges();
+ }
+ else
+ {
+ return false;
+ }
+
+ /*
+ * check that each mapped range lies within the sequence range
+ * (necessary for circular CDS - example EMBL:J03321:AAA91567)
+ * and mapped length covers (at least) sequence length
+ */
+ int length = 0;
+ for (int[] range : mappedRanges)
+ {
+ int from = Math.min(range[0], range[1]);
+ int to = Math.max(range[0], range[1]);
+ if (from < seq.getStart() || to > seq.getEnd())
+ {
+ return false;
+ }
+ length += (to - from + 1);
+ }
+ // add 1 to mapped length to allow for a mapped stop codon
+ if (length + 1 < (seq.getEnd() - seq.getStart() + 1))
+ {
+ return false;
+ }
+ return true;
+ }
+
+ /**
+ * Adds any regions mapped to or from position {@code pos} in sequence
+ * {@code seq} to the given search results
+ *
+ * @param seq
+ * @param pos
+ * @param sr
+ */
+ public void markMappedRegion(SequenceI seq, int pos, SearchResultsI sr)
+ {
+ int[] codon = null;
+ SequenceI mappedSeq = null;
+ SequenceI ds = seq.getDatasetSequence();
+ if (ds == null)
+ {
+ ds = seq;
+ }
+
+ if (this.fromSeq == seq || this.fromSeq == ds)
+ {
+ codon = this.mapping.map.locateInTo(pos, pos);
+ mappedSeq = this.mapping.to;
+ }
+ else if (this.mapping.to == seq || this.mapping.to == ds)
+ {
+ codon = this.mapping.map.locateInFrom(pos, pos);
+ mappedSeq = this.fromSeq;
+ }
+
+ if (codon != null)
+ {
+ for (int i = 0; i < codon.length; i += 2)
+ {
+ sr.addResult(mappedSeq, codon[i], codon[i + 1]);
+ }
+ }
+ }
}
private List<SequenceToSequenceMapping> mappings;
/**
* Returns the first mapping found which is to or from the given sequence, or
- * null.
+ * null if none is found
*
* @param seq
* @return
*/
- public Mapping getMappingForSequence(SequenceI seq, boolean cdsOnly)
+ public Mapping getMappingForSequence(SequenceI seq)
{
SequenceI seqDs = seq.getDatasetSequence();
seqDs = seqDs != null ? seqDs : seq;
{
if (ssm.fromSeq == seqDs || ssm.mapping.to == seqDs)
{
- if (!cdsOnly || ssm.fromSeq.getName().startsWith("CDS")
- || ssm.mapping.to.getName().startsWith("CDS"))
- {
- return ssm.mapping;
- }
+ return ssm.mapping;
}
}
return null;
}
/**
+ * Return the corresponding aligned or dataset dna sequence for given amino
+ * acid sequence, or null if not found. returns the sequence from the first
+ * mapping found that involves the protein sequence.
*
- * @param sequenceRef
- * @return null or corresponding aaSeq entry for dnaSeq entry
+ * @param aaSeqRef
+ * @return
*/
public SequenceI getDnaForAaSeq(SequenceI aaSeqRef)
{
/**
* Add search results for regions in other sequences that translate or are
- * translated from a particular position in seq
+ * translated from a particular position in seq (which may be an aligned or
+ * dataset sequence)
*
* @param seq
* @param index
public void markMappedRegion(SequenceI seq, int index,
SearchResultsI results)
{
- int[] codon;
SequenceI ds = seq.getDatasetSequence();
- for (SequenceToSequenceMapping ssm : mappings)
+ if (ds == null)
{
- if (ssm.fromSeq == seq || ssm.fromSeq == ds)
- {
- codon = ssm.mapping.map.locateInTo(index, index);
- if (codon != null)
- {
- for (int i = 0; i < codon.length; i += 2)
- {
- results.addResult(ssm.mapping.to, codon[i], codon[i + 1]);
- }
- }
- }
- else if (ssm.mapping.to == seq || ssm.mapping.to == ds)
- {
- {
- codon = ssm.mapping.map.locateInFrom(index, index);
- if (codon != null)
- {
- for (int i = 0; i < codon.length; i += 2)
- {
- results.addResult(ssm.fromSeq, codon[i], codon[i + 1]);
- }
- }
- }
- }
+ ds = seq;
}
- }
-
- /**
- * 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++;
+ ssm.markMappedRegion(ds, index, results);
}
- return ml == null ? null : ml.locateInFrom(aaPos, aaPos);
}
/**
* Two AlignedCodonFrame objects are equal if they hold the same ordered list
* of mappings
*
- * @see SequenceToSequenceMapping#
+ * @see SequenceToSequenceMapping#equals
*/
@Override
public boolean equals(Object obj)
{
return mappings;
}
+
+ /**
+ * Returns the first mapping found which is between the two given sequences,
+ * and covers the full extent of both.
+ *
+ * @param seq1
+ * @param seq2
+ * @return
+ */
+ public SequenceToSequenceMapping getCoveringMapping(SequenceI seq1,
+ SequenceI seq2)
+ {
+ for (SequenceToSequenceMapping mapping : mappings)
+ {
+ if (mapping.covers(seq2) && mapping.covers(seq1))
+ {
+ return mapping;
+ }
+ }
+ return null;
+ }
+
+ /**
+ * Returns the first mapping found which is between the given dataset sequence
+ * and another, is a triplet mapping (3:1 or 1:3), and covers the full extent
+ * of both sequences involved
+ *
+ * @param seq
+ * @return
+ */
+ public SequenceToSequenceMapping getCoveringCodonMapping(SequenceI seq)
+ {
+ for (SequenceToSequenceMapping mapping : mappings)
+ {
+ if (mapping.getMapping().getMap().isTripletMap()
+ && mapping.covers(seq))
+ {
+ if (mapping.fromSeq == seq
+ && mapping.covers(mapping.getMapping().getTo()))
+ {
+ return mapping;
+ }
+ else if (mapping.getMapping().getTo() == seq
+ && mapping.covers(mapping.fromSeq))
+ {
+ return mapping;
+ }
+ }
+ }
+ return null;
+ }
}