X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fanalysis%2FCrossRef.java;h=8fd07067962c515b62ac728c59a1578950312fb4;hb=438af5436afff7bc1ce5ab380ec5813a7beb11b9;hp=72382393811ab4b8a7e97b6bcae75726958dec0c;hpb=3ef44bef1f825d26977dedd1608469712a87fe15;p=jalview.git
diff --git a/src/jalview/analysis/CrossRef.java b/src/jalview/analysis/CrossRef.java
index 7238239..8fd0706 100644
--- a/src/jalview/analysis/CrossRef.java
+++ b/src/jalview/analysis/CrossRef.java
@@ -20,21 +20,23 @@
*/
package jalview.analysis;
-import java.util.ArrayList;
-import java.util.List;
-import java.util.Vector;
-
import jalview.datamodel.AlignedCodonFrame;
import jalview.datamodel.Alignment;
import jalview.datamodel.AlignmentI;
import jalview.datamodel.DBRefEntry;
-import jalview.datamodel.DBRefSource;
+import jalview.datamodel.Mapping;
import jalview.datamodel.Sequence;
+import jalview.datamodel.SequenceFeature;
import jalview.datamodel.SequenceI;
+import jalview.util.Comparison;
import jalview.util.DBRefUtils;
-import jalview.ws.SequenceFetcher;
+import jalview.util.MapList;
+import jalview.ws.SequenceFetcherFactory;
import jalview.ws.seqfetcher.ASequenceFetcher;
+import java.util.ArrayList;
+import java.util.List;
+
/**
* Functions for cross-referencing sequence databases. user must first specify
* if cross-referencing from protein or dna (set dna==true)
@@ -44,219 +46,204 @@ import jalview.ws.seqfetcher.ASequenceFetcher;
*/
public class CrossRef
{
- /**
- * Select just the DNA or protein references for a protein or dna sequence
- *
- * @param fromDna
- * if true, select references from DNA (i.e. Protein databases), else
- * DNA database references
- * @param refs
- * a set of references to select from
- * @return
+ /*
+ * A sub-class that ignores Parent attribute when comparing sequence
+ * features. This avoids 'duplicate' CDS features that only
+ * differ in their parent Transcript ids.
*/
- public static DBRefEntry[] findXDbRefs(boolean fromDna, DBRefEntry[] refs)
+ class MySequenceFeature extends SequenceFeature
{
- return DBRefUtils.selectRefs(refs, fromDna ? DBRefSource.PROTEINDBS
- : DBRefSource.DNACODINGDBS);
- // could attempt to find other cross
- // refs here - ie PDB xrefs
- // (not dna, not protein seq)
- }
+ private SequenceFeature feat;
- /**
- * @param dna
- * true if seqs are DNA seqs
- * @param seqs
- * @return a list of sequence database cross reference source types
- */
- public static String[] findSequenceXrefTypes(boolean dna, SequenceI[] seqs)
- {
- return findSequenceXrefTypes(dna, seqs, null);
+ MySequenceFeature(SequenceFeature sf)
+ {
+ this.feat = sf;
+ }
+
+ @Override
+ public boolean equals(Object o)
+ {
+ return feat.equals(o, true);
+ }
}
/**
- * Indirect references are references from other sequences from the dataset to
- * any of the direct DBRefEntrys on the given sequences.
+
+ * Returns a list of distinct database sources for which sequences have either
+ *
+ * - a (dna-to-protein or protein-to-dna) cross-reference
+ * - an indirect cross-reference - a (dna-to-protein or protein-to-dna)
+ * reference from another sequence in the dataset which has a cross-reference
+ * to a direct DBRefEntry on the given sequence
+ *
*
* @param dna
- * true if seqs are DNA seqs
+ * true if seqs are nucleotide
* @param seqs
- * @return a list of sequence database cross reference source types
+ * sequences whose xrefs we are seeking
+ * @param dataset
+ * an alignment to search for indirect references
+ * @return
*/
- public static String[] findSequenceXrefTypes(boolean dna,
+ public static List findXrefSourcesForSequences(boolean dna,
SequenceI[] seqs, AlignmentI dataset)
{
- String[] dbrefs = null;
- List refs = new ArrayList();
- for (int s = 0; s < seqs.length; s++)
+ List sources = new ArrayList();
+ for (SequenceI seq : seqs)
{
- if (seqs[s] != null)
+ if (seq != null)
{
- SequenceI dss = seqs[s];
- while (dss.getDatasetSequence() != null)
- {
- dss = dss.getDatasetSequence();
- }
- DBRefEntry[] rfs = findXDbRefs(dna, dss.getDBRef());
- for (int r = 0; rfs != null && r < rfs.length; r++)
- {
- if (!refs.contains(rfs[r].getSource()))
- {
- refs.add(rfs[r].getSource());
- }
- }
- if (dataset != null)
- {
- // search for references to this sequence's direct references.
- DBRefEntry[] lrfs = CrossRef
- .findXDbRefs(!dna, seqs[s].getDBRef());
- List rseqs = new ArrayList();
- CrossRef.searchDatasetXrefs(seqs[s], !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()); // not used??
- for (int r = 0; rfs != null && r < rfs.length; r++)
- {
- if (!refs.contains(rfs[r].getSource()))
- {
- refs.add(rfs[r].getSource());
- }
- }
- }
- }
+ findXrefSourcesForSequence(seq, dna, dataset, sources);
}
}
- if (refs.size() > 0)
- {
- dbrefs = new String[refs.size()];
- refs.toArray(dbrefs);
- }
- return dbrefs;
+ return sources;
}
- /*
- * if (dna) { if (rfs[r].hasMap()) { // most likely this is a protein cross
- * reference if (!refs.contains(rfs[r].getSource())) {
- * refs.addElement(rfs[r].getSource()); } } }
+ /**
+ * Returns a list of distinct database sources for which a sequence has either
+ *
+ * - a (dna-to-protein or protein-to-dna) cross-reference
+ * - an indirect cross-reference - a (dna-to-protein or protein-to-dna)
+ * reference from another sequence in the dataset which has a cross-reference
+ * to a direct DBRefEntry on the given sequence
+ *
+ *
+ * @param seq
+ * the sequence whose dbrefs we are searching against
+ * @param dna
+ * true if the sequence is nucleotide
+ * @param dataset
+ * an alignment to search for indirect references
+ * @param sources
+ * a list of sources to add matches to
*/
- public static boolean hasCdnaMap(SequenceI[] seqs)
+ static void findXrefSourcesForSequence(SequenceI seq, boolean dna,
+ AlignmentI dataset, List sources)
{
- String[] reftypes = findSequenceXrefTypes(false, seqs);
- for (int s = 0; s < reftypes.length; s++)
+ /*
+ * first find seq's xrefs (dna-to-peptide or peptide-to-dna)
+ */
+ DBRefEntry[] rfs = DBRefUtils.selectDbRefs(!dna, seq.getDBRefs());
+ addXrefsToSources(rfs, sources);
+ if (dataset != null)
{
- if (reftypes.equals(DBRefSource.EMBLCDS))
+ /*
+ * find sequence's direct (dna-to-dna, peptide-to-peptide) xrefs
+ */
+ DBRefEntry[] lrfs = DBRefUtils.selectDbRefs(dna, seq.getDBRefs());
+ List rseqs = new ArrayList();
+
+ /*
+ * find sequences in the alignment which xref one of these DBRefs
+ * i.e. is xref-ed to a common sequence identifier
+ */
+ CrossRef.searchDatasetXrefs(seq, !dna, lrfs, dataset, rseqs, null);
+
+ /*
+ * add those sequences' (dna-to-peptide or peptide-to-dna) dbref sources
+ */
+ for (SequenceI rs : rseqs)
{
- return true;
- // no map
+ DBRefEntry[] xrs = DBRefUtils.selectDbRefs(!dna, rs.getDBRefs());
+ addXrefsToSources(xrs, sources);
}
}
- return false;
}
- public static SequenceI[] getCdnaMap(SequenceI[] seqs)
+ /**
+ * Helper method that adds the source identifiers of some cross-references to
+ * a (non-redundant) list of database sources
+ *
+ * @param xrefs
+ * @param sources
+ */
+ static void addXrefsToSources(DBRefEntry[] xrefs, List sources)
{
- Vector cseqs = new Vector();
- for (int s = 0; s < seqs.length; s++)
+ if (xrefs != null)
{
- DBRefEntry[] cdna = findXDbRefs(true, seqs[s].getDBRef());
- for (int c = 0; c < cdna.length; c++)
+ for (DBRefEntry ref : xrefs)
{
- if (cdna[c].getSource().equals(DBRefSource.EMBLCDS))
+ String source = ref.getSource();
+ if (!sources.contains(source))
{
- System.err
- .println("TODO: unimplemented sequence retrieval for coding region sequence.");
- // TODO: retrieve CDS dataset sequences
- // need global dataset sequence retriever/resolver to reuse refs
- // and construct Mapping entry.
- // insert gaps in CDS according to peptide gaps.
- // add gapped sequence to cseqs
+ sources.add(source);
}
}
}
- if (cseqs.size() > 0)
- {
- SequenceI[] rsqs = new SequenceI[cseqs.size()];
- cseqs.copyInto(rsqs);
- return rsqs;
- }
- return null;
-
- }
-
- /**
- *
- * @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)
* @return products (as dataset sequences)
*/
- public static Alignment findXrefSequences(SequenceI[] seqs, boolean dna,
- String source, AlignmentI dataset)
+ public static Alignment findXrefSequences(SequenceI[] seqs,
+ final boolean dna, final String source, AlignmentI al)
{
+ AlignmentI dataset = al.getDataset() == null ? al : al.getDataset();
List rseqs = new ArrayList();
- Alignment ral = null;
- AlignedCodonFrame cf = new AlignedCodonFrame(); // nominal width
- for (int s = 0; s < seqs.length; s++)
+ AlignedCodonFrame cf = new AlignedCodonFrame();
+ for (SequenceI seq : seqs)
{
- SequenceI dss = seqs[s];
+ SequenceI dss = seq;
while (dss.getDatasetSequence() != null)
{
dss = dss.getDatasetSequence();
}
boolean found = false;
- DBRefEntry[] xrfs = CrossRef.findXDbRefs(dna, dss.getDBRef());
+ DBRefEntry[] xrfs = DBRefUtils.selectDbRefs(!dna, dss.getDBRefs());
if ((xrfs == null || xrfs.length == 0) && dataset != null)
{
- System.out.println("Attempting to find ds Xrefs refs.");
- DBRefEntry[] lrfs = CrossRef.findXDbRefs(!dna, seqs[s].getDBRef());
- // less ambiguous would be a 'find primary dbRefEntry' method.
- // filter for desired source xref here
+ /*
+ * found no suitable dbrefs on sequence - look for sequences in the
+ * alignment which share a dbref with this one
+ */
+ DBRefEntry[] lrfs = DBRefUtils.selectDbRefs(dna, seq.getDBRefs());
+
+ /*
+ * find sequences (except this one!), of complementary type,
+ * which have a dbref to an accession id for this sequence,
+ * and add them to the results
+ */
found = CrossRef.searchDatasetXrefs(dss, !dna, lrfs, dataset,
rseqs, cf);
}
for (int r = 0; xrfs != null && r < xrfs.length; r++)
{
- if (source != null && !source.equals(xrfs[r].getSource()))
+ DBRefEntry xref = xrfs[r];
+ if (source != null && !source.equals(xref.getSource()))
{
continue;
}
- if (xrfs[r].hasMap())
+ if (xref.hasMap())
{
- if (xrfs[r].getMap().getTo() != null)
+ if (xref.getMap().getTo() != null)
{
- SequenceI rsq = new Sequence(xrfs[r].getMap().getTo());
+ found = true;
+ SequenceI rsq = new Sequence(xref.getMap().getTo());
rseqs.add(rsq);
- if (xrfs[r].getMap().getMap().getFromRatio() != xrfs[r]
+ if (xref.getMap().getMap().getFromRatio() != xref
.getMap().getMap().getToRatio())
{
// get sense of map correct for adding to product alignment.
if (dna)
{
// map is from dna seq to a protein product
- cf.addMap(dss, rsq, xrfs[r].getMap().getMap());
+ cf.addMap(dss, rsq, xref.getMap().getMap());
}
else
{
// map should be from protein seq to its coding dna
- cf.addMap(rsq, dss, xrfs[r].getMap().getMap().getInverse());
+ cf.addMap(rsq, dss, xref.getMap().getMap().getInverse());
}
}
- found = true;
}
}
if (!found)
@@ -265,9 +252,10 @@ public class CrossRef
// xrefs on this sequence.
if (dataset != null)
{
- found |= searchDataset(dss, xrfs[r], dataset, rseqs, cf); // ,false,!dna);
+ found = searchDataset(dss, xref, dataset, rseqs, cf, false,/*true?*/
+ !dna);
if (found)
- {
+ {
xrfs[r] = null; // we've recovered seqs for this one.
}
}
@@ -277,12 +265,8 @@ public class CrossRef
{
if (xrfs != null && xrfs.length > 0)
{
- // Try and get the sequence reference...
- /*
- * Ideal world - we ask for a sequence fetcher implementation here if
- * (jalview.io.RunTimeEnvironment.getSequenceFetcher()) (
- */
- ASequenceFetcher sftch = new SequenceFetcher();
+ ASequenceFetcher sftch = SequenceFetcherFactory
+ .getSequenceFetcher();
SequenceI[] retrieved = null;
int l = xrfs.length;
for (int r = 0; r < xrfs.length; r++)
@@ -299,8 +283,8 @@ public class CrossRef
}
if (l > 0)
{
- System.out
- .println("Attempting to retrieve cross referenced sequences.");
+ // System.out
+ // .println("Attempting to retrieve cross referenced sequences.");
DBRefEntry[] t = new DBRefEntry[l];
l = 0;
for (int r = 0; r < xrfs.length; r++)
@@ -313,37 +297,59 @@ public class CrossRef
xrfs = t;
try
{
- retrieved = sftch.getSequences(xrfs); // problem here is we don't
- // know which of xrfs
- // resulted in which
+ retrieved = sftch.getSequences(xrfs, !dna);
+ // problem here is we don't know which of xrfs resulted in which
// retrieved element
} catch (Exception e)
{
System.err
.println("Problem whilst retrieving cross references for Sequence : "
- + seqs[s].getName());
+ + seq.getName());
e.printStackTrace();
}
+
if (retrieved != null)
{
+ updateDbrefMappings(dna, seq, xrfs, retrieved, cf);
+
+ SequenceIdMatcher matcher = new SequenceIdMatcher(
+ dataset.getSequences());
+ List copiedFeatures = new ArrayList();
+ CrossRef me = new CrossRef();
for (int rs = 0; rs < retrieved.length; rs++)
{
// TODO: examine each sequence for 'redundancy'
- jalview.datamodel.DBRefEntry[] dbr = retrieved[rs]
- .getDBRef();
+ DBRefEntry[] dbr = retrieved[rs].getDBRefs();
if (dbr != null && dbr.length > 0)
{
for (int di = 0; di < dbr.length; di++)
{
// find any entry where we should put in the sequence being
// cross-referenced into the map
- jalview.datamodel.Mapping map = dbr[di].getMap();
+ Mapping map = dbr[di].getMap();
if (map != null)
{
if (map.getTo() != null && map.getMap() != null)
{
- // should search the local dataset to find any existing
- // candidates for To !
+ SequenceI matched = matcher
+ .findIdMatch(map.getTo());
+ if (matched != null)
+ {
+ /*
+ * already got an xref to this sequence; update this
+ * map to point to the same sequence, and add
+ * any new dbrefs to it
+ */
+ for (DBRefEntry ref : map.getTo().getDBRefs())
+ {
+ matched.addDBRef(ref); // add or update mapping
+ }
+ map.setTo(matched);
+ }
+ else
+ {
+ matcher.add(map.getTo());
+ }
try
{
// compare ms with dss and replace with dss in mapping
@@ -352,17 +358,55 @@ public class CrossRef
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. same variation 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());
+ }
+ // TODO remove this 'else' and the cf.addMap above?
+ else
+ {
+ cf.addMap(retrieved[rs].getDatasetSequence(),
+ map.getTo(), map.getMap());
}
} catch (Exception e)
{
@@ -382,12 +426,12 @@ public class CrossRef
}
}
}
+
+ Alignment ral = null;
if (rseqs.size() > 0)
{
- SequenceI[] rsqs = new SequenceI[rseqs.size()];
- rseqs.toArray(rsqs);
- ral = new Alignment(rsqs);
- if (cf != null && cf.getProtMappings() != null)
+ ral = new Alignment(rseqs.toArray(new SequenceI[rseqs.size()]));
+ if (cf != null && !cf.isEmpty())
{
ral.addCodonFrame(cf);
}
@@ -396,6 +440,69 @@ public class CrossRef
}
/**
+ * Updates any empty mappings in the cross-references with one to a compatible
+ * retrieved sequence if found, and adds any new mappings to the
+ * AlignedCodonFrame
+ *
+ * @param dna
+ * @param mapFrom
+ * @param xrefs
+ * @param retrieved
+ * @param acf
+ */
+ static void updateDbrefMappings(boolean dna, SequenceI mapFrom,
+ DBRefEntry[] xrefs, SequenceI[] retrieved, AlignedCodonFrame acf)
+ {
+ SequenceIdMatcher matcher = new SequenceIdMatcher(retrieved);
+ for (DBRefEntry xref : xrefs)
+ {
+ if (!xref.hasMap())
+ {
+ String targetSeqName = xref.getSource() + "|"
+ + xref.getAccessionId();
+ SequenceI[] matches = matcher.findAllIdMatches(targetSeqName);
+ if (matches == null)
+ {
+ return;
+ }
+ for (SequenceI seq : matches)
+ {
+ MapList mapping = null;
+ if (dna)
+ {
+ mapping = AlignmentUtils.mapCdnaToProtein(seq, mapFrom);
+ }
+ else
+ {
+ mapping = AlignmentUtils.mapCdnaToProtein(mapFrom, seq);
+ if (mapping != null)
+ {
+ mapping = mapping.getInverse();
+ }
+ }
+ if (mapping != null)
+ {
+ xref.setMap(new Mapping(seq, mapping));
+ if (dna)
+ {
+ AlignmentUtils.computeProteinFeatures(mapFrom, seq, mapping);
+ }
+ if (dna)
+ {
+ acf.addMap(mapFrom, seq, mapping);
+ }
+ else
+ {
+ acf.addMap(seq, mapFrom, mapping.getInverse());
+ }
+ continue;
+ }
+ }
+ }
+ }
+ }
+
+ /**
* find references to lrfs in the cross-reference set of each sequence in
* dataset (that is not equal to sequenceI) Identifies matching DBRefEntry
* based on source and accession string only - Map and Version are nulled.
@@ -408,8 +515,7 @@ public class CrossRef
*/
private static boolean searchDatasetXrefs(SequenceI sequenceI,
boolean dna, DBRefEntry[] lrfs, AlignmentI dataset,
- List rseqs,
- AlignedCodonFrame cf)
+ List rseqs, AlignedCodonFrame cf)
{
boolean found = false;
if (lrfs == null)
@@ -422,38 +528,26 @@ public class CrossRef
// add in wildcards
xref.setVersion(null);
xref.setMap(null);
- found = searchDataset(sequenceI, xref, dataset, rseqs, cf, false, dna);
+ found |= searchDataset(sequenceI, xref, dataset, rseqs, cf, false,
+ dna);
}
return found;
}
/**
- * search a given sequence dataset for references matching cross-references to
- * the given sequence
+ * Searches dataset for DBRefEntrys matching the given one (xrf) and adds the
+ * associated sequence to rseqs
*
* @param sequenceI
+ * a sequence to ignore (start point of search)
* @param xrf
+ * a cross-reference to try to match
* @param dataset
+ * sequences to search in
* @param rseqs
- * set of unique sequences
+ * result list to add to
* @param cf
- * @return true if one or more unique sequences were found and added
- */
- public static boolean searchDataset(SequenceI sequenceI, DBRefEntry xrf,
- AlignmentI dataset, List rseqs, AlignedCodonFrame cf)
- {
- return searchDataset(sequenceI, xrf, dataset, rseqs, cf, true, false);
- }
-
- /**
- * TODO: generalise to different protein classifications Search dataset for
- * DBRefEntrys matching the given one (xrf) and add the associated sequence to
- * rseq.
- *
- * @param sequenceI
- * @param xrf
- * @param dataset
- * @param rseqs
+ * a set of sequence mappings to add to
* @param direct
* - search all references or only subset
* @param dna
@@ -465,7 +559,6 @@ public class CrossRef
boolean direct, boolean dna)
{
boolean found = false;
- SequenceI[] typer = new SequenceI[1];
if (dataset == null)
{
return false;
@@ -487,105 +580,82 @@ public class CrossRef
System.err
.println("Implementation warning: getProducts passed a dataset alignment without dataset sequences in it!");
}
- if (nxt != sequenceI && nxt != sequenceI.getDatasetSequence())
+ if (nxt == sequenceI || nxt == sequenceI.getDatasetSequence())
{
- // check if this is the correct sequence type
+ continue;
+ }
+ // check if this is the correct sequence type
+ {
+ // TODO 'direct' is always set to false - remove?
+ // or should it be 'true' from findXrefSequences?
+ // also its Javadoc conflicts with its use:
+ // test below implies 'direct' means find complementary sequences,
+ // !direct means select same molecule type
+ boolean isDna = Comparison
+ .isNucleotide(new SequenceI[] { nxt });
+ if ((direct && isDna == dna) || (!direct && isDna != dna))
{
- typer[0] = nxt;
- boolean isDna = jalview.util.Comparison.isNucleotide(typer);
- if ((direct && isDna == dna) || (!direct && isDna != dna))
- {
- // skip this sequence because it is same molecule type
- continue;
- }
+ // skip this sequence because it is wrong molecule type
+ continue;
}
+ }
- // look for direct or indirect references in common
- DBRefEntry[] poss = nxt.getDBRef(), cands = null;
- if (direct)
- {
- cands = jalview.util.DBRefUtils.searchRefs(poss, xrf);
- }
- else
- {
- poss = CrossRef.findXDbRefs(dna, poss); //
- cands = jalview.util.DBRefUtils.searchRefs(poss, xrf);
- }
- if (cands != null)
+ // look for direct or indirect references in common
+ DBRefEntry[] poss = nxt.getDBRefs();
+ DBRefEntry[] cands = null;
+ /*
+ * TODO does this make any sense?
+ * if 'direct', search the dbrefs for xrf
+ * else, filter the dbrefs by type and then search for xrf
+ * - the result is the same isn't it?
+ */
+ if (direct)
+ {
+ cands = DBRefUtils.searchRefs(poss, xrf);
+ }
+ else
+ {
+ poss = DBRefUtils.selectDbRefs(!dna, poss);
+ cands = DBRefUtils.searchRefs(poss, xrf);
+ }
+ if (cands != null)
+ {
+ if (!rseqs.contains(nxt))
{
- if (!rseqs.contains(nxt))
+ found = true;
+ rseqs.add(nxt);
+ boolean foundmap = cf != null;
+ // don't search if we aren't given a codon map object
+ for (int r = 0; foundmap && r < cands.length; r++)
{
- rseqs.add(nxt);
- boolean foundmap = cf != null;
- // don't search if we aren't given a codon map object
- for (int r = 0; foundmap && r < cands.length; r++)
+ if (cands[r].hasMap())
{
- if (cands[r].hasMap())
+ Mapping mapping = cands[r].getMap();
+ MapList map = mapping.getMap();
+ if (mapping.getTo() != null
+ && map.getFromRatio() != map.getToRatio())
{
- if (cands[r].getMap().getTo() != null
- && cands[r].getMap().getMap().getFromRatio() != cands[r]
- .getMap().getMap().getToRatio())
+ // get sense of map correct for adding to product
+ // alignment.
+ if (dna)
{
- foundmap = true;
- // get sense of map correct for adding to product
- // alignment.
- if (dna)
- {
- // map is from dna seq to a protein product
- cf.addMap(sequenceI, nxt, cands[r].getMap()
- .getMap());
- }
- else
- {
- // map should be from protein seq to its coding dna
- cf.addMap(nxt, sequenceI, cands[r].getMap()
- .getMap().getInverse());
- }
+ // map is from dna seq to a protein product
+ cf.addMap(sequenceI, nxt, map);
+ }
+ else
+ {
+ // map should be from protein seq to its coding dna
+ cf.addMap(nxt, sequenceI, map.getInverse());
}
}
}
- // TODO: add mapping between sequences if necessary
- found = true;
}
+ // TODO: add mapping between sequences if necessary
}
-
}
}
}
}
return found;
}
-
- /**
- * precalculate different products that can be found for seqs in dataset and
- * return them.
- *
- * @param dna
- * @param seqs
- * @param dataset
- * @param fake
- * - don't actually build lists - just get types
- * @return public static Object[] buildXProductsList(boolean dna, SequenceI[]
- * seqs, AlignmentI dataset, boolean fake) { String types[] =
- * jalview.analysis.CrossRef.findSequenceXrefTypes( dna, seqs,
- * dataset); if (types != null) { System.out.println("Xref Types for:
- * "+(dna ? "dna" : "prot")); for (int t = 0; t < types.length; t++) {
- * System.out.println("Type: " + types[t]); SequenceI[] prod =
- * jalview.analysis.CrossRef.findXrefSequences(seqs, dna, types[t]);
- * System.out.println("Found " + ((prod == null) ? "no" : "" +
- * prod.length) + " products"); if (prod!=null) { for (int p=0;
- * p