import jalview.datamodel.DBRefSource;
import jalview.datamodel.Mapping;
import jalview.datamodel.Sequence;
+import jalview.datamodel.SequenceFeature;
import jalview.datamodel.SequenceI;
import jalview.util.DBRefUtils;
import jalview.ws.SequenceFetcher;
*/
public class CrossRef
{
+ /*
+ * A sub-class that ignores Parent attribute when comparing sequence
+ * features. This avoids 'duplicate' CDS features that only
+ * differ in their parent Transcript ids.
+ */
+ class MySequenceFeature extends SequenceFeature
+ {
+ private SequenceFeature feat;
+
+ MySequenceFeature(SequenceFeature sf)
+ {
+ this.feat = sf;
+ }
+
+ @Override
+ public boolean equals(Object o)
+ {
+ return feat.equals(o, true);
+ }
+ }
+
/**
* Select just the DNA or protein references for a protein or dna sequence
*
{
dss = dss.getDatasetSequence();
}
- DBRefEntry[] rfs = findXDbRefs(dna, dss.getDBRef());
+ DBRefEntry[] rfs = findXDbRefs(dna, dss.getDBRefs());
if (rfs != null)
{
for (DBRefEntry ref : rfs)
if (dataset != null)
{
// search for references to this sequence's direct references.
- DBRefEntry[] lrfs = CrossRef.findXDbRefs(!dna, seq.getDBRef());
+ DBRefEntry[] lrfs = CrossRef.findXDbRefs(!dna, seq.getDBRefs());
List<SequenceI> rseqs = new ArrayList<SequenceI>();
CrossRef.searchDatasetXrefs(seq, !dna, lrfs, dataset, rseqs,
null); // don't need to specify codon frame for mapping here
for (SequenceI rs : rseqs)
{
- DBRefEntry[] xrs = findXDbRefs(dna, rs.getDBRef());
+ DBRefEntry[] xrs = findXDbRefs(dna, rs.getDBRefs());
if (xrs != null)
{
for (DBRefEntry ref : xrs)
Vector cseqs = new Vector();
for (int s = 0; s < seqs.length; s++)
{
- DBRefEntry[] cdna = findXDbRefs(true, seqs[s].getDBRef());
+ DBRefEntry[] cdna = findXDbRefs(true, seqs[s].getDBRefs());
for (int c = 0; c < cdna.length; c++)
{
if (cdna[c].getSource().equals(DBRefSource.EMBLCDS))
/**
*
- * @param dna
- * @param seqs
- * @return
- */
- public static Alignment findXrefSequences(SequenceI[] seqs, boolean dna,
- String source)
- {
- return findXrefSequences(seqs, dna, source, null);
- }
-
- /**
- *
* @param seqs
* sequences whose xrefs are being retrieved
* @param dna
* true if sequences are nucleotide
* @param source
- * @param dataset
- * alignment to search for product sequences.
+ * @param al
+ * alignment to search for cross-referenced sequences (and possibly
+ * add to)
+ * @param addedPeers
+ * a list of sequences to add to if 'peers' to the original sequences
+ * are found e.g. alternative protein products for a protein's gene
* @return products (as dataset sequences)
*/
public static Alignment findXrefSequences(SequenceI[] seqs, boolean dna,
- String source, AlignmentI dataset)
+ String source, AlignmentI al, List<SequenceI> addedPeers)
{
+ AlignmentI dataset = al.getDataset() == null ? al : al.getDataset();
List<SequenceI> rseqs = new ArrayList<SequenceI>();
AlignedCodonFrame cf = new AlignedCodonFrame();
for (SequenceI seq : seqs)
dss = dss.getDatasetSequence();
}
boolean found = false;
- DBRefEntry[] xrfs = CrossRef.findXDbRefs(dna, dss.getDBRef());
+ DBRefEntry[] xrfs = CrossRef.findXDbRefs(dna, dss.getDBRefs());
if ((xrfs == null || xrfs.length == 0) && dataset != null)
{
System.out.println("Attempting to find ds Xrefs refs.");
// FIXME should be dss not seq here?
- DBRefEntry[] lrfs = CrossRef.findXDbRefs(!dna, seq.getDBRef());
+ DBRefEntry[] lrfs = CrossRef.findXDbRefs(!dna, seq.getDBRefs());
// less ambiguous would be a 'find primary dbRefEntry' method.
// filter for desired source xref here
found = CrossRef.searchDatasetXrefs(dss, !dna, lrfs, dataset,
+ seq.getName());
e.printStackTrace();
}
+
if (retrieved != null)
{
+ List<SequenceFeature> copiedFeatures = new ArrayList<SequenceFeature>();
+ CrossRef me = new CrossRef();
for (int rs = 0; rs < retrieved.length; rs++)
{
// TODO: examine each sequence for 'redundancy'
- DBRefEntry[] dbr = retrieved[rs].getDBRef();
+ DBRefEntry[] dbr = retrieved[rs].getDBRefs();
if (dbr != null && dbr.length > 0)
{
for (int di = 0; di < dbr.length; di++)
int sf = map.getMap().getToLowest();
int st = map.getMap().getToHighest();
SequenceI mappedrg = ms.getSubSequence(sf, st);
- SequenceI loc = dss.getSubSequence(sf, st);
+ // SequenceI loc = dss.getSubSequence(sf, st);
if (mappedrg.getLength() > 0
- && mappedrg.getSequenceAsString().equals(
- loc.getSequenceAsString()))
+ && ms.getSequenceAsString().equals(
+ dss.getSequenceAsString()))
+ // && mappedrg.getSequenceAsString().equals(
+ // loc.getSequenceAsString()))
{
- System.err
- .println("Mapping updated for retrieved crossreference");
+ String msg = "Mapping updated from "
+ + ms.getName()
+ + " to retrieved crossreference "
+ + dss.getName();
+ System.out.println(msg);
// method to update all refs of existing To on
// retrieved sequence with dss and merge any props
// on To onto dss.
map.setTo(dss);
+ /*
+ * copy sequence features as well, avoiding
+ * duplication (e.g. from 2 transcripts)
+ */
+ SequenceFeature[] sfs = ms
+ .getSequenceFeatures();
+ if (sfs != null)
+ {
+ for (SequenceFeature feat : sfs)
+ {
+ /*
+ * we override SequenceFeature.equals here (but
+ * not elsewhere) to ignore Parent attribute
+ * TODO not quite working yet!
+ */
+ if (!copiedFeatures
+ .contains(me.new MySequenceFeature(
+ feat)))
+ {
+ dss.addSequenceFeature(feat);
+ copiedFeatures.add(feat);
+ }
+ }
+ }
+ cf.addMap(retrieved[rs].getDatasetSequence(),
+ dss, map.getMap());
+ }
+ else
+ {
+ addedPeers.add(map.getTo());
+ cf.addMap(retrieved[rs].getDatasetSequence(),
+ map.getTo(), map.getMap());
}
} catch (Exception e)
{
Alignment ral = null;
if (rseqs.size() > 0)
{
- SequenceI[] rsqs = new SequenceI[rseqs.size()];
- rseqs.toArray(rsqs);
- ral = new Alignment(rsqs);
+ ral = new Alignment(rseqs.toArray(new SequenceI[rseqs.size()]));
if (cf != null && !cf.isEmpty())
{
ral.addCodonFrame(cf);
}
// look for direct or indirect references in common
- DBRefEntry[] poss = nxt.getDBRef(), cands = null;
+ DBRefEntry[] poss = nxt.getDBRefs(), cands = null;
if (direct)
{
cands = jalview.util.DBRefUtils.searchRefs(poss, xrf);
git://source.jalview.org / jalview.git / blobdiff