X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fanalysis%2FCrossRef.java;h=7e77fc1e0903799660db67205966ddcf768a1897;hb=0ebcd487f08f9873987d9a32edce7df47aa27927;hp=bbf3f93c3e8501d1d533a358abec2697033887ca;hpb=a8f483d04205bb8273ee311c12968b7e86d205fa;p=jalview.git
diff --git a/src/jalview/analysis/CrossRef.java b/src/jalview/analysis/CrossRef.java
index bbf3f93..7e77fc1 100644
--- a/src/jalview/analysis/CrossRef.java
+++ b/src/jalview/analysis/CrossRef.java
@@ -1,38 +1,43 @@
/*
- * Jalview - A Sequence Alignment Editor and Viewer (Version 2.8.2)
- * Copyright (C) 2014 The Jalview Authors
+ * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
+ * Copyright (C) $$Year-Rel$$ The Jalview Authors
*
* This file is part of Jalview.
*
* Jalview is free software: you can redistribute it and/or
* modify it under the terms of the GNU General Public License
- * as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
+ * as published by the Free Software Foundation, either version 3
+ * of the License, or (at your option) any later version.
*
* Jalview is distributed in the hope that it will be useful, but
* WITHOUT ANY WARRANTY; without even the implied warranty
* of MERCHANTABILITY or FITNESS FOR A PARTICULAR
* PURPOSE. See the GNU General Public License for more details.
*
- * You should have received a copy of the GNU General Public License along with Jalview. If not, see .
+ * You should have received a copy of the GNU General Public License
+ * along with Jalview. If not, see .
* The Jalview Authors are detailed in the 'AUTHORS' file.
*/
package jalview.analysis;
-import java.util.Enumeration;
-import java.util.List;
-import java.util.Vector;
-import java.util.Hashtable;
-
import jalview.datamodel.AlignedCodonFrame;
import jalview.datamodel.Alignment;
import jalview.datamodel.AlignmentI;
-import jalview.datamodel.DBRefSource;
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.DBRefUtils;
+import jalview.util.MapList;
import jalview.ws.SequenceFetcher;
import jalview.ws.seqfetcher.ASequenceFetcher;
+import java.util.ArrayList;
+import java.util.List;
+import java.util.Vector;
+
/**
* Functions for cross-referencing sequence databases. user must first specify
* if cross-referencing from protein or dna (set dna==true)
@@ -42,40 +47,44 @@ import jalview.ws.seqfetcher.ASequenceFetcher;
*/
public class CrossRef
{
- /**
- * get the DNA or protein references for a protein or dna sequence
- *
- * @param dna
- * @param rfs
- * @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 dna, DBRefEntry[] rfs)
+ class MySequenceFeature extends SequenceFeature
{
- if (dna)
+ private SequenceFeature feat;
+
+ MySequenceFeature(SequenceFeature sf)
{
- rfs = jalview.util.DBRefUtils.selectRefs(rfs, DBRefSource.PROTEINDBS);
+ this.feat = sf;
}
- else
+
+ @Override
+ public boolean equals(Object o)
{
- rfs = jalview.util.DBRefUtils.selectRefs(rfs,
- DBRefSource.DNACODINGDBS); // could attempt to find other cross
- // refs and return here - ie PDB xrefs
- // (not dna, not protein seq)
+ return feat.equals(o, true);
}
- return rfs;
}
- public static Hashtable classifyDbRefs(DBRefEntry[] rfs)
+ /**
+ * 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
+ */
+ public static DBRefEntry[] findXDbRefs(boolean fromDna, DBRefEntry[] refs)
{
- Hashtable classes = new Hashtable();
- classes.put(DBRefSource.PROTEINDBS,
- jalview.util.DBRefUtils.selectRefs(rfs, DBRefSource.PROTEINDBS));
- classes.put(DBRefSource.DNACODINGDBS, jalview.util.DBRefUtils
- .selectRefs(rfs, DBRefSource.DNACODINGDBS));
- classes.put(DBRefSource.DOMAINDBS,
- jalview.util.DBRefUtils.selectRefs(rfs, DBRefSource.DOMAINDBS));
- // classes.put(OTHER, )
- return classes;
+ 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)
}
/**
@@ -102,45 +111,55 @@ public class CrossRef
SequenceI[] seqs, AlignmentI dataset)
{
String[] dbrefs = null;
- Vector refs = new Vector();
- for (int s = 0; s < seqs.length; s++)
+ List refs = new ArrayList();
+ for (SequenceI seq : seqs)
{
- if (seqs[s] != null)
+ if (seq != null)
{
-
- SequenceI dss = seqs[s];
+ SequenceI dss = seq;
while (dss.getDatasetSequence() != null)
{
dss = dss.getDatasetSequence();
}
- DBRefEntry[] rfs = findXDbRefs(dna, dss.getDBRef());
- for (int r = 0; rfs != null && r < rfs.length; r++)
+ DBRefEntry[] rfs = findXDbRefs(dna, dss.getDBRefs());
+ if (rfs != null)
{
- if (!refs.contains(rfs[r].getSource()))
+ for (DBRefEntry ref : rfs)
{
- refs.addElement(rfs[r].getSource());
+ if (!refs.contains(ref.getSource()))
+ {
+ refs.add(ref.getSource());
+ }
}
}
if (dataset != null)
{
// search for references to this sequence's direct references.
- DBRefEntry[] lrfs = CrossRef
- .findXDbRefs(!dna, seqs[s].getDBRef());
- Vector rseqs = new Vector();
- CrossRef.searchDatasetXrefs(seqs[s], !dna, lrfs, dataset, rseqs,
+ DBRefEntry[] lrfs = CrossRef.findXDbRefs(!dna, seq.getDBRefs());
+ List rseqs = new ArrayList();
+ CrossRef.searchDatasetXrefs(seq, !dna, lrfs, dataset, rseqs,
null); // don't need to specify codon frame for mapping here
- Enumeration lr = rseqs.elements();
- while (lr.hasMoreElements())
+ for (SequenceI rs : rseqs)
{
- SequenceI rs = (SequenceI) lr.nextElement();
- DBRefEntry[] xrs = findXDbRefs(dna, rs.getDBRef());
- for (int r = 0; rfs != null && r < rfs.length; r++)
+ DBRefEntry[] xrs = findXDbRefs(dna, rs.getDBRefs());
+ if (xrs != null)
{
- if (!refs.contains(rfs[r].getSource()))
+ for (DBRefEntry ref : xrs)
{
- refs.addElement(rfs[r].getSource());
+ if (!refs.contains(ref.getSource()))
+ {
+ refs.add(ref.getSource());
+ }
}
}
+ // looks like copy and paste - change rfs to xrs?
+ // for (int r = 0; rfs != null && r < rfs.length; r++)
+ // {
+ // if (!refs.contains(rfs[r].getSource()))
+ // {
+ // refs.add(rfs[r].getSource());
+ // }
+ // }
}
}
}
@@ -148,18 +167,14 @@ public class CrossRef
if (refs.size() > 0)
{
dbrefs = new String[refs.size()];
- refs.copyInto(dbrefs);
+ refs.toArray(dbrefs);
}
return dbrefs;
}
- /*
- * 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()); } } }
- */
public static boolean hasCdnaMap(SequenceI[] seqs)
{
+ // TODO unused - remove?
String[] reftypes = findSequenceXrefTypes(false, seqs);
for (int s = 0; s < reftypes.length; s++)
{
@@ -174,10 +189,11 @@ public class CrossRef
public static SequenceI[] getCdnaMap(SequenceI[] seqs)
{
+ // TODO unused - remove?
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))
@@ -204,78 +220,67 @@ public class CrossRef
/**
*
- * @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)
{
- Vector rseqs = new Vector();
- Alignment ral = null;
- AlignedCodonFrame cf = new AlignedCodonFrame(0); // nominal width
- for (int s = 0; s < seqs.length; s++)
+ AlignmentI dataset = al.getDataset() == null ? al : al.getDataset();
+ List rseqs = new ArrayList();
+ 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 = CrossRef.findXDbRefs(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.
+ // FIXME should be dss not seq here?
+ 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,
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)
{
- Sequence rsq = new Sequence(xrfs[r].getMap().getTo());
- rseqs.addElement(rsq);
- if (xrfs[r].getMap().getMap().getFromRatio() != xrfs[r]
+ SequenceI rsq = new Sequence(xref.getMap().getTo());
+ rseqs.add(rsq);
+ 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;
@@ -287,9 +292,12 @@ 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,
+ !dna);
if (found)
+ {
xrfs[r] = null; // we've recovered seqs for this one.
+ }
}
}
}
@@ -319,49 +327,73 @@ 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++)
{
if (xrfs[r] != null)
+ {
t[l++] = xrfs[r];
+ }
}
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
@@ -370,17 +402,54 @@ 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());
+ }
+ else
+ {
+ cf.addMap(retrieved[rs].getDatasetSequence(),
+ map.getTo(), map.getMap());
}
} catch (Exception e)
{
@@ -393,19 +462,19 @@ public class CrossRef
}
}
retrieved[rs].updatePDBIds();
- rseqs.addElement(retrieved[rs]);
+ rseqs.add(retrieved[rs]);
}
}
}
}
}
}
+
+ Alignment ral = null;
if (rseqs.size() > 0)
{
- SequenceI[] rsqs = new SequenceI[rseqs.size()];
- rseqs.copyInto(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);
}
@@ -414,6 +483,69 @@ 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.
@@ -425,12 +557,14 @@ public class CrossRef
* @return true if matches were found.
*/
private static boolean searchDatasetXrefs(SequenceI sequenceI,
- boolean dna, DBRefEntry[] lrfs, AlignmentI dataset, Vector rseqs,
- AlignedCodonFrame cf)
+ boolean dna, DBRefEntry[] lrfs, AlignmentI dataset,
+ List rseqs, AlignedCodonFrame cf)
{
boolean found = false;
if (lrfs == null)
+ {
return false;
+ }
for (int i = 0; i < lrfs.length; i++)
{
DBRefEntry xref = new DBRefEntry(lrfs[i]);
@@ -455,7 +589,7 @@ public class CrossRef
* @return true if one or more unique sequences were found and added
*/
public static boolean searchDataset(SequenceI sequenceI, DBRefEntry xrf,
- AlignmentI dataset, Vector rseqs, AlignedCodonFrame cf)
+ AlignmentI dataset, List rseqs, AlignedCodonFrame cf)
{
return searchDataset(sequenceI, xrf, dataset, rseqs, cf, true, false);
}
@@ -476,13 +610,15 @@ public class CrossRef
* @return true if relationship found and sequence added.
*/
public static boolean searchDataset(SequenceI sequenceI, DBRefEntry xrf,
- AlignmentI dataset, Vector rseqs, AlignedCodonFrame cf,
+ AlignmentI dataset, List rseqs, AlignedCodonFrame cf,
boolean direct, boolean dna)
{
boolean found = false;
SequenceI[] typer = new SequenceI[1];
if (dataset == null)
+ {
return false;
+ }
if (dataset.getSequences() == null)
{
System.err.println("Empty dataset sequence set - NO VECTOR");
@@ -492,6 +628,7 @@ public class CrossRef
synchronized (ds = dataset.getSequences())
{
for (SequenceI nxt : ds)
+ {
if (nxt != null)
{
if (nxt.getDatasetSequence() != null)
@@ -513,7 +650,7 @@ public class CrossRef
}
// 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);
@@ -527,10 +664,9 @@ public class CrossRef
{
if (!rseqs.contains(nxt))
{
- rseqs.addElement(nxt);
- boolean foundmap = cf != null; // don't search if we aren't
- // given
- // a codon map object
+ 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())
@@ -564,6 +700,7 @@ public class CrossRef
}
}
+ }
}
return found;
}