if (!rseqs.contains(matchInDataset))
{
rseqs.add(matchInDataset);
- // need to try harder to only add unique mappings
- if (xref.getMap().getMap().isTripletMap()
- && dataset.getMapping(seq, matchInDataset) == null
- && cf.getMappingBetween(seq, matchInDataset) == null)
+ }
+ // even if rseqs contained matchInDataset - check mappings between
+ // these seqs are added
+ // need to try harder to only add unique mappings
+ if (xref.getMap().getMap().isTripletMap()
+ && dataset.getMapping(seq, matchInDataset) == null
+ && cf.getMappingBetween(seq, matchInDataset) == null)
+ {
+ // materialise a mapping for highlighting between these
+ // sequences
+ if (fromDna)
{
- // materialise a mapping for highlighting between these sequences
- if (fromDna)
- {
- cf.addMap(dss, matchInDataset, xref.getMap().getMap(), xref.getMap().getMappedFromId());
- } else {
- cf.addMap(matchInDataset, dss, xref.getMap().getMap().getInverse(), xref.getMap().getMappedFromId());
- }
+ cf.addMap(dss, matchInDataset, xref.getMap().getMap(),
+ xref.getMap().getMappedFromId());
+ }
+ else
+ {
+ cf.addMap(matchInDataset, dss, xref.getMap().getMap()
+ .getInverse(), xref.getMap().getMappedFromId());
}
}
+
refIterator.remove();
continue;
}
SequenceI[] retrieved = null;
SequenceI dss = seq.getDatasetSequence() == null ? seq : seq
.getDatasetSequence();
+ // first filter in case we are retrieving crossrefs that have already been
+ // retrieved. this happens for cases where a database record doesn't yield
+ // protein products for CDS
+ removeAlreadyRetrievedSeqs(sourceRefs, fromDna);
+ if (sourceRefs.size() == 0)
+ {
+ // no more work to do! We already had all requested sequence records in
+ // the dataset.
+ return;
+ }
try
{
retrieved = sftch.getSequences(sourceRefs, !fromDna);
// try: Ensembl -> Nuc->Ensembl, Nuc->Uniprot-->Protein->EMBL->
SequenceI retrievedDss = retrievedSequence.getDatasetSequence() == null ? retrievedSequence
: retrievedSequence.getDatasetSequence();
- DBRefEntry[] dbr = retrievedSequence.getDBRefs();
- if (dbr != null)
+ importCrossRefSeq(cf, dss, retrievedDss);
+ rseqs.add(retrievedDss);
+ if (dataset.findIndex(retrievedDss) == -1)
+ {
+ dataset.addSequence(retrievedDss);
+ matcher.add(retrievedDss);
+ }
+ }
+ }
+ }
+
+ /**
+ * Search dataset for sequences with a primary reference contained in
+ * sourceRefs.
+ *
+ * @param sourceRefs
+ * - list of references to filter.
+ * @param fromDna
+ * - type of sequence to search for matching primary reference.
+ */
+ private void removeAlreadyRetrievedSeqs(List<DBRefEntry> sourceRefs,
+ boolean fromDna)
+ {
+ DBRefEntry[] dbrSourceSet = sourceRefs.toArray(new DBRefEntry[0]);
+ for (SequenceI sq : dataset.getSequences())
+ {
+ boolean dupeFound = false;
+ // !fromDna means we are looking only for nucleotide sequences, not
+ // protein
+ if (sq.isProtein() == fromDna)
+ {
+ for (DBRefEntry dbr : sq.getPrimaryDBRefs())
{
- for (DBRefEntry dbref : dbr)
+ for (DBRefEntry found : DBRefUtils.searchRefs(dbrSourceSet, dbr))
{
- // find any entry where we should put in the sequence being
- // cross-referenced into the map
- Mapping map = dbref.getMap();
- if (map != null)
+ sourceRefs.remove(found);
+ dupeFound = true;
+ }
+ }
+ }
+ if (dupeFound)
+ {
+ dbrSourceSet = sourceRefs.toArray(new DBRefEntry[0]);
+ }
+ }
+ }
+
+ /**
+ * process sequence retrieved via a dbref on source sequence to resolve and
+ * transfer data
+ *
+ * @param cf
+ * @param sourceSequence
+ * @param retrievedSequence
+ */
+ private void importCrossRefSeq(AlignedCodonFrame cf,
+ SequenceI sourceSequence, SequenceI retrievedSequence)
+ {
+ DBRefEntry[] dbr = retrievedSequence.getDBRefs();
+ if (dbr != null)
+ {
+ for (DBRefEntry dbref : dbr)
+ {
+ // find any entry where we should put in the sequence being
+ // cross-referenced into the map
+ Mapping map = dbref.getMap();
+ if (map != null)
+ {
+ if (map.getTo() != null && map.getMap() != null)
+ {
+ // TODO findInDataset requires exact sequence match but
+ // 'congruent' test is only for the mapped part
+ // maybe not a problem in practice since only ENA provide a
+ // mapping and it is to the full protein translation of CDS
+ SequenceI matched = findInDataset(dbref);
+ // matcher.findIdMatch(map.getTo());
+ if (matched != null)
{
- if (map.getTo() != null && map.getMap() != null)
+ /*
+ * already got an xref to this sequence; update this
+ * map to point to the same sequence, and add
+ * any new dbrefs to it
+ */
+ DBRefEntry[] toRefs = map.getTo().getDBRefs();
+ if (toRefs != null)
{
- // TODO findInDataset requires exact sequence match but
- // 'congruent' test is only for the mapped part
- // maybe not a problem in practice since only ENA provide a
- // mapping and it is to the full protein translation of CDS
- SequenceI matched = findInDataset(dbref);
- // matcher.findIdMatch(map.getTo());
- if (matched != null)
+ for (DBRefEntry ref : toRefs)
{
- /*
- * already got an xref to this sequence; update this
- * map to point to the same sequence, and add
- * any new dbrefs to it
- */
- DBRefEntry[] toRefs = map.getTo().getDBRefs();
- if (toRefs != null)
- {
- for (DBRefEntry ref : toRefs)
- {
- matched.addDBRef(ref); // add or update mapping
- }
- }
- map.setTo(matched);
+ matched.addDBRef(ref); // add or update mapping
}
- else
- {
- if (dataset.findIndex(map.getTo()) == -1)
- {
- dataset.addSequence(map.getTo());
- matcher.add(map.getTo());
- }
- }
- try
- {
- // compare ms with dss and replace with dss in mapping
- // if map is congruent
- SequenceI ms = map.getTo();
- int sf = map.getMap().getToLowest();
- int st = map.getMap().getToHighest();
- SequenceI mappedrg = ms.getSubSequence(sf, st);
- // SequenceI loc = dss.getSubSequence(sf, st);
- if (mappedrg.getLength() > 0
- && ms.getSequenceAsString().equals(
- dss.getSequenceAsString()))
- // && mappedrg.getSequenceAsString().equals(
- // loc.getSequenceAsString()))
- {
- String msg = "Mapping updated from " + ms.getName()
- + " to retrieved crossreference "
- + dss.getName();
- System.out.println(msg);
- map.setTo(dss);
+ }
+ map.setTo(matched);
+ }
+ else
+ {
+ if (dataset.findIndex(map.getTo()) == -1)
+ {
+ dataset.addSequence(map.getTo());
+ matcher.add(map.getTo());
+ }
+ }
- /*
- * give the reverse reference the inverse mapping
- * (if it doesn't have one already)
- */
- setReverseMapping(dss, dbref, cf);
+ try
+ {
+ // compare ms with dss and replace with dss in mapping
+ // if map is congruent
+ SequenceI ms = map.getTo();
+ int sf = map.getMap().getToLowest();
+ int st = map.getMap().getToHighest();
+ SequenceI mappedrg = ms.getSubSequence(sf, st);
+ if (mappedrg.getLength() > 0
+ && ms.getSequenceAsString().equals(
+ sourceSequence.getSequenceAsString()))
+ {
+ String msg = "Mapping updated from " + ms.getName()
+ + " to retrieved crossreference "
+ + sourceSequence.getName();
+ System.out.println(msg);
+ map.setTo(sourceSequence);
+
+ /*
+ * give the reverse reference the inverse mapping
+ * (if it doesn't have one already)
+ */
+ setReverseMapping(sourceSequence, dbref, cf);
+ /*
+ * copy sequence features as well, avoiding
+ * duplication (e.g. same variation from two
+ * transcripts)
+ */
+ SequenceFeature[] sfs = ms.getSequenceFeatures();
+ if (sfs != null)
+ {
+ for (SequenceFeature feat : sfs)
+ {
/*
- * copy sequence features as well, avoiding
- * duplication (e.g. same variation from two
- * transcripts)
+ * make a flyweight feature object which ignores Parent
+ * attribute in equality test; this avoids creating many
+ * otherwise duplicate exon features on genomic sequence
*/
- SequenceFeature[] sfs = ms.getSequenceFeatures();
- if (sfs != null)
+ SequenceFeature newFeature = new SequenceFeature(
+ feat)
{
- for (SequenceFeature feat : sfs)
+ @Override
+ public boolean equals(Object o)
{
- /*
- * make a flyweight feature object which ignores Parent
- * attribute in equality test; this avoids creating many
- * otherwise duplicate exon features on genomic sequence
- */
- SequenceFeature newFeature = new SequenceFeature(
- feat)
- {
- @Override
- public boolean equals(Object o)
- {
- return super.equals(o, true);
- }
- };
- dss.addSequenceFeature(newFeature);
+ return super.equals(o, true);
}
- }
+ };
+ sourceSequence.addSequenceFeature(newFeature);
}
- cf.addMap(retrievedDss, map.getTo(), map.getMap());
- } catch (Exception e)
- {
- System.err
- .println("Exception when consolidating Mapped sequence set...");
- e.printStackTrace(System.err);
}
}
+ cf.addMap(retrievedSequence, map.getTo(), map.getMap());
+ } catch (Exception e)
+ {
+ System.err
+ .println("Exception when consolidating Mapped sequence set...");
+ e.printStackTrace(System.err);
}
}
}
- retrievedSequence.updatePDBIds();
- rseqs.add(retrievedDss);
- if (dataset.findIndex(retrievedDss) == -1)
- {
- dataset.addSequence(retrievedDss);
- matcher.add(retrievedDss);
- }
}
}
+ retrievedSequence.updatePDBIds();
}
/**
* Sets the inverse sequence mapping in the corresponding dbref of the mapped
String name2 = xref.getSource() + "|" + name;
SequenceI dss = mapsTo.getDatasetSequence() == null ? mapsTo : mapsTo
.getDatasetSequence();
+ // first check ds if ds is directly referenced
+ if (dataset.findIndex(dss) > -1)
+ {
+ return dss;
+ }
+ ;
for (SequenceI seq : dataset.getSequences())
{
/*
* </ul>
* @return true if relationship found and sequence added.
*/
- boolean searchDataset(boolean fromDna, SequenceI fromSeq,
- DBRefEntry xrf, List<SequenceI> foundSeqs, AlignedCodonFrame mappings,
+ boolean searchDataset(boolean fromDna, SequenceI fromSeq, DBRefEntry xrf,
+ List<SequenceI> foundSeqs, AlignedCodonFrame mappings,
boolean direct)
{
boolean found = false;
// }
if (!cands.isEmpty())
{
- if (!foundSeqs.contains(nxt))
+ if (foundSeqs.contains(nxt))
{
- found = true;
- foundSeqs.add(nxt);
- if (mappings != null && !direct)
+ continue;
+ }
+ found = true;
+ foundSeqs.add(nxt);
+ if (mappings != null && !direct)
+ {
+ /*
+ * if the matched sequence has mapped dbrefs to
+ * protein product / cdna, add equivalent mappings to
+ * our source sequence
+ */
+ for (DBRefEntry candidate : cands)
{
- /*
- * if the matched sequence has mapped dbrefs to
- * protein product / cdna, add equivalent mappings to
- * our source sequence
- */
- for (DBRefEntry candidate : cands)
+ Mapping mapping = candidate.getMap();
+ if (mapping != null)
{
- Mapping mapping = candidate.getMap();
- if (mapping != null)
+ MapList map = mapping.getMap();
+ if (mapping.getTo() != null
+ && map.getFromRatio() != map.getToRatio())
{
- MapList map = mapping.getMap();
- if (mapping.getTo() != null
- && map.getFromRatio() != map.getToRatio())
+ /*
+ * add a mapping, as from dna to peptide sequence
+ */
+ if (map.getFromRatio() == 3)
{
- /*
- * add a mapping, as from dna to peptide sequence
- */
- if (map.getFromRatio() == 3)
- {
- mappings.addMap(nxt, fromSeq, map);
- }
- else
- {
- mappings.addMap(nxt, fromSeq, map.getInverse());
- }
+ mappings.addMap(nxt, fromSeq, map);
+ }
+ else
+ {
+ mappings.addMap(nxt, fromSeq, map.getInverse());
}
}
}
git://source.jalview.org / jalview.git / blobdiff