*/
public boolean covers(SequenceI seq)
{
- List<int[]> mappedRanges = null;
+ return covers(seq,false,false);
+ }
+ /**
+ *
+ * @param seq
+ * @param localCover - when true - compare extent of seq's dataset sequence rather than the local extent
+ * @param either - when true coverage is required for either seq or the mapped sequence
+ * @return true if mapping covers full length of given sequence (or the other if either==true)
+ */
+ public boolean covers(SequenceI seq, boolean localCover,boolean either)
+ {
+ List<int[]> mappedRanges = null,otherRanges=null;
MapList mapList = mapping.getMap();
+ int mstart=seq.getStart(),mend=seq.getEnd(),ostart,oend;
+ ;
if (fromSeq == seq || fromSeq == seq.getDatasetSequence())
{
+ if (localCover && fromSeq !=seq)
+ {
+ mstart=fromSeq.getStart();
+ mend=fromSeq.getEnd();
+ }
mappedRanges = mapList.getFromRanges();
+ otherRanges=mapList.getToRanges();
+ ostart=mapping.to.getStart();
+ oend=mapping.to.getEnd();
}
else if (mapping.to == seq || mapping.to == seq.getDatasetSequence())
{
+ if (localCover && mapping.to !=seq)
+ {
+ mstart=mapping.to.getStart();
+ mend=mapping.to.getEnd();
+ }
mappedRanges = mapList.getToRanges();
+ otherRanges=mapList.getFromRanges();
+ ostart=fromSeq.getStart();
+ oend=fromSeq.getEnd();
}
else
{
* (necessary for circular CDS - example EMBL:J03321:AAA91567)
* and mapped length covers (at least) sequence length
*/
- int length = 0;
+ int length = countRange(mappedRanges,mstart,mend);
+
+ if (length != -1)
+ {
+ // add 3 to mapped length to allow for a mapped stop codon
+ if (length + 3 >= (mend - mstart + 1))
+ {
+ return true;
+ }
+ }
+ if (either)
+ {
+ // also check coverage of the other range
+ length = countRange(otherRanges, ostart, oend);
+ if (length != -1)
+ {
+ if (length + 1 >= (oend - ostart + 1))
+ {
+ return true;
+ }
+ }
+ }
+ return false;
+ }
+ private int countRange(List<int[]> mappedRanges,int mstart,int mend)
+ {
+ 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())
+ if (from < mstart || to > mend)
{
- return false;
+ return -1;
}
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;
+ return length;
}
/**
* Adds any regions mapped to or from position {@code pos} in sequence
* {@code seq} to the given search results
- *
+ * Note: recommend first using the .covers(,true,true) to ensure mapping covers both sequences
* @param seq
* @param pos
* @param sr
/**
* 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
SearchResultsI results)
{
SequenceI ds = seq.getDatasetSequence();
- for (SequenceToSequenceMapping ssm : mappings)
+ if (ds == null)
{
- ssm.markMappedRegion(ds, index, results);
+ 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;
+ if (ssm.covers(seq,true,true)) {
+ ssm.markMappedRegion(ds, index, results);
}
- i++;
}
- return ml == null ? null : ml.locateInFrom(aaPos, aaPos);
}
/**
}
/**
- * Returns the first mapping found which is between the given sequence and
- * another, is a triplet mapping (3:1 or 1:3), and covers the full extent of
- * both sequences involved.
+ * 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