import jalview.datamodel.AlignedCodon;
import jalview.datamodel.AlignedCodonFrame;
+import jalview.datamodel.AlignedCodonFrame.SequenceToSequenceMapping;
import jalview.datamodel.Alignment;
import jalview.datamodel.AlignmentAnnotation;
import jalview.datamodel.AlignmentI;
*
* @param dna
* aligned dna sequences
- * @param mappings
- * from dna to protein
- * @param al
+ * @param dataset
* @return an alignment whose sequences are the cds-only parts of the dna
* sequences (or null if no mappings are found)
*/
public static AlignmentI makeCdsAlignment(SequenceI[] dna,
- List<AlignedCodonFrame> mappings, AlignmentI al)
+ AlignmentI dataset)
{
List<SequenceI> cdsSeqs = new ArrayList<SequenceI>();
+ List<AlignedCodonFrame> mappings = dataset.getCodonFrames();
+ /*
+ * construct CDS sequences from the (cds-to-protein) mappings made earlier;
+ * this makes it possible to model multiple products from dna (e.g. EMBL);
+ * however it does mean we don't have the EMBL protein_id (a property on
+ * the CDS features) in order to make the CDS sequence name :-(
+ */
for (SequenceI seq : dna)
{
- AlignedCodonFrame cdsMappings = new AlignedCodonFrame();
+ SequenceI seqDss = seq.getDatasetSequence() == null ? seq : seq
+ .getDatasetSequence();
List<AlignedCodonFrame> seqMappings = MappingUtils
.findMappingsForSequence(seq, mappings);
- List<AlignedCodonFrame> alignmentMappings = al.getCodonFrames();
for (AlignedCodonFrame mapping : seqMappings)
{
- for (Mapping aMapping : mapping.getMappingsFromSequence(seq))
+ List<Mapping> mappingsFromSequence = mapping.getMappingsFromSequence(seq);
+
+ for (Mapping aMapping : mappingsFromSequence)
{
- SequenceI cdsSeq = makeCdsSequence(seq.getDatasetSequence(),
- aMapping);
+ if (aMapping.getMap().getFromRatio() == 1)
+ {
+ /*
+ * not a dna-to-protein mapping (likely dna-to-cds)
+ */
+ continue;
+ }
+
+ /*
+ * check for an existing CDS sequence i.e. a 3:1 mapping to
+ * the dna mapping's product
+ */
+ SequenceI cdsSeq = null;
+ // TODO better mappings collection data model so we can do
+ // a table lookup instead of double loops to find mappings
+ SequenceI proteinProduct = aMapping.getTo();
+ for (AlignedCodonFrame acf : MappingUtils
+ .findMappingsForSequence(proteinProduct, mappings))
+ {
+ for (SequenceToSequenceMapping map : acf.getMappings())
+ {
+ if (map.getMapping().getMap().getFromRatio() == 3
+ && proteinProduct == map.getMapping().getTo()
+ && seqDss != map.getFromSeq())
+ {
+ /*
+ * found a 3:1 mapping to the protein product which is not
+ * from the dna sequence...assume it is from the CDS sequence
+ * TODO mappings data model that brings together related
+ * dna-cds-protein mappings in one object
+ */
+ cdsSeq = map.getFromSeq();
+ }
+ }
+ }
+ if (cdsSeq != null)
+ {
+ /*
+ * mappings are always to dataset sequences so create an aligned
+ * sequence to own it; add the dataset sequence to the dataset
+ */
+ SequenceI derivedSequence = cdsSeq.deriveSequence();
+ cdsSeqs.add(derivedSequence);
+ if (!dataset.getSequences().contains(cdsSeq))
+ {
+ dataset.addSequence(cdsSeq);
+ }
+ continue;
+ }
+
+ /*
+ * didn't find mapped CDS sequence - construct it and add
+ * its dataset sequence to the dataset
+ */
+ cdsSeq = makeCdsSequence(seq.getDatasetSequence(), aMapping);
+ SequenceI cdsSeqDss = cdsSeq.createDatasetSequence();
cdsSeqs.add(cdsSeq);
-
+ if (!dataset.getSequences().contains(cdsSeqDss))
+ {
+ dataset.addSequence(cdsSeqDss);
+ }
+
/*
* add a mapping from CDS to the (unchanged) mapped to range
*/
MapList map = new MapList(cdsRange, aMapping.getMap()
.getToRanges(), aMapping.getMap().getFromRatio(),
aMapping.getMap().getToRatio());
- cdsMappings.addMap(cdsSeq, aMapping.getTo(), map);
+ AlignedCodonFrame cdsToProteinMapping = new AlignedCodonFrame();
+ cdsToProteinMapping.addMap(cdsSeq, proteinProduct, map);
+
+ /*
+ * guard against duplicating the mapping if repeating this action
+ */
+ if (!mappings.contains(cdsToProteinMapping))
+ {
+ mappings.add(cdsToProteinMapping);
+ }
/*
* add another mapping from original 'from' range to CDS
*/
+ AlignedCodonFrame dnaToProteinMapping = new AlignedCodonFrame();
map = new MapList(aMapping.getMap().getFromRanges(), cdsRange, 1,
1);
- cdsMappings.addMap(seq.getDatasetSequence(), cdsSeq, map);
+ dnaToProteinMapping.addMap(seq.getDatasetSequence(), cdsSeq, map);
+ if (!mappings.contains(dnaToProteinMapping))
+ {
+ mappings.add(dnaToProteinMapping);
+ }
- alignmentMappings.add(cdsMappings);
/*
* transfer any features on dna that overlap the CDS
}
}
- /*
- * add CDS seqs to shared dataset
- */
- Alignment dataset = al.getDataset();
- for (SequenceI seq : cdsSeqs)
- {
- if (!dataset.getSequences().contains(seq.getDatasetSequence()))
- {
- dataset.addSequence(seq.getDatasetSequence());
- }
- }
AlignmentI cds = new Alignment(cdsSeqs.toArray(new SequenceI[cdsSeqs
.size()]));
- cds.setDataset(dataset);
+ cds.setDataset((Alignment) dataset);
return cds;
}
*
* @param seq
* @param mapping
- * @return
+ * @return CDS sequence (as a dataset sequence)
*/
static SequenceI makeCdsSequence(SequenceI seq, Mapping mapping)
{
SequenceI newSeq = new Sequence(seq.getName() + "|"
+ mapping.getTo().getName(), newSeqChars, 1, newPos);
- newSeq.createDatasetSequence();
return newSeq;
}
* sort to get sequence features in start position order
* - would be better to store in Sequence as a TreeSet or NCList?
*/
- Arrays.sort(peptide.getSequenceFeatures(),
- new Comparator<SequenceFeature>()
- {
- @Override
- public int compare(SequenceFeature o1, SequenceFeature o2)
+ if (peptide.getSequenceFeatures() != null)
+ {
+ Arrays.sort(peptide.getSequenceFeatures(),
+ new Comparator<SequenceFeature>()
{
- int c = Integer.compare(o1.getBegin(), o2.getBegin());
- return c == 0 ? Integer.compare(o1.getEnd(), o2.getEnd())
- : c;
- }
- });
+ @Override
+ public int compare(SequenceFeature o1, SequenceFeature o2)
+ {
+ int c = Integer.compare(o1.getBegin(), o2.getBegin());
+ return c == 0 ? Integer.compare(o1.getEnd(), o2.getEnd())
+ : c;
+ }
+ });
+ }
return count;
}
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.util.MapList;
-import jalview.ws.SequenceFetcher;
+import jalview.ws.SequenceFetcherFactory;
import jalview.ws.seqfetcher.ASequenceFetcher;
import java.util.ArrayList;
+import java.util.Iterator;
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)
+ * Functions for cross-referencing sequence databases.
*
* @author JimP
*
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.
+ * the dataset of the alignment for which we are searching for
+ * cross-references; in some cases we may resolve xrefs by
+ * searching in the dataset
*/
- class MySequenceFeature extends SequenceFeature
- {
- private SequenceFeature feat;
-
- MySequenceFeature(SequenceFeature sf)
- {
- this.feat = sf;
- }
+ private AlignmentI dataset;
- @Override
- public boolean equals(Object o)
- {
- return feat.equals(o, true);
- }
- }
+ /*
+ * true if we are searching for cross-references from nucleotide,
+ * i.e. for protein sequences, false if the reverse
+ */
+ private boolean fromDna;
- /**
- * 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
+ /*
+ * the sequences for which we are seeking cross-references
*/
- public static DBRefEntry[] findXDbRefs(boolean fromDna, DBRefEntry[] refs)
- {
- 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 SequenceI[] fromSeqs;
/**
- * @param dna
- * true if seqs are DNA seqs
+ * Constructor
+ *
* @param seqs
- * @return a list of sequence database cross reference source types
+ * the sequences for which we are seeking cross-references
+ * @param ds
+ * the containing alignment dataset (may be searched to resolve
+ * cross-references)
*/
- public static String[] findSequenceXrefTypes(boolean dna, SequenceI[] seqs)
+ public CrossRef(SequenceI[] seqs, AlignmentI ds)
{
- return findSequenceXrefTypes(dna, seqs, null);
+ fromSeqs = seqs;
+ fromDna = ds.isNucleotide();
+ dataset = ds.getDataset() == null ? ds : ds.getDataset();
}
/**
- * Indirect references are references from other sequences from the dataset to
- * any of the direct DBRefEntrys on the given sequences.
- *
- * @param dna
- * true if seqs are DNA seqs
- * @param seqs
- * @return a list of sequence database cross reference source types
+ * Returns a list of distinct database sources for which sequences have either
+ * <ul>
+ * <li>a (dna-to-protein or protein-to-dna) cross-reference</li>
+ * <li>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</li>
+ * </ul>
+ * @return
*/
- public static String[] findSequenceXrefTypes(boolean dna,
- SequenceI[] seqs, AlignmentI dataset)
+ public List<String> findXrefSourcesForSequences()
{
- String[] dbrefs = null;
- List<String> refs = new ArrayList<String>();
- for (SequenceI seq : seqs)
+ List<String> sources = new ArrayList<String>();
+ for (SequenceI seq : fromSeqs)
{
if (seq != null)
{
- SequenceI dss = seq;
- while (dss.getDatasetSequence() != null)
- {
- dss = dss.getDatasetSequence();
- }
- DBRefEntry[] rfs = findXDbRefs(dna, dss.getDBRefs());
- if (rfs != null)
- {
- for (DBRefEntry ref : rfs)
- {
- 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, 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.getDBRefs());
- if (xrs != null)
- {
- for (DBRefEntry ref : xrs)
- {
- 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());
- // }
- // }
- }
- }
+ findXrefSourcesForSequence(seq, sources);
}
}
- if (refs.size() > 0)
- {
- dbrefs = new String[refs.size()];
- refs.toArray(dbrefs);
- }
- return dbrefs;
+ return sources;
}
- public static boolean hasCdnaMap(SequenceI[] seqs)
+ /**
+ * Returns a list of distinct database sources for which a sequence has either
+ * <ul>
+ * <li>a (dna-to-protein or protein-to-dna) cross-reference</li>
+ * <li>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</li>
+ * </ul>
+ *
+ * @param seq
+ * the sequence whose dbrefs we are searching against
+ * @param sources
+ * a list of sources to add matches to
+ */
+ void findXrefSourcesForSequence(SequenceI seq, List<String> sources)
{
- // TODO unused - remove?
- 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(!fromDna, 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(fromDna, seq.getDBRefs());
+ List<SequenceI> rseqs = new ArrayList<SequenceI>();
+
+ /*
+ * find sequences in the alignment which xref one of these DBRefs
+ * i.e. is xref-ed to a common sequence identifier
+ */
+ searchDatasetXrefs(seq, lrfs, 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(!fromDna, 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
+ */
+ void addXrefsToSources(DBRefEntry[] xrefs, List<String> sources)
{
- // TODO unused - remove?
- Vector cseqs = new Vector();
- for (int s = 0; s < seqs.length; s++)
+ if (xrefs != null)
{
- DBRefEntry[] cdna = findXDbRefs(true, seqs[s].getDBRefs());
- for (int c = 0; c < cdna.length; c++)
+ for (DBRefEntry ref : xrefs)
{
- if (cdna[c].getSource().equals(DBRefSource.EMBLCDS))
+ /*
+ * avoid duplication e.g. ENSEMBL and Ensembl
+ */
+ String source = DBRefUtils.getCanonicalName(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;
-
}
/**
+ * Attempts to find cross-references from the sequences provided in the
+ * constructor to the given source database. Cross-references may be found
+ * <ul>
+ * <li>in dbrefs on the sequence which hold a mapping to a sequence
+ * <ul>
+ * <li>provided with a fetched sequence (e.g. ENA translation), or</li>
+ * <li>populated previously after getting cross-references</li>
+ * </ul>
+ * <li>as other sequences in the alignment which share a dbref identifier with
+ * the sequence</li>
+ * <li>by fetching from the remote database</li>
+ * </ul>
+ * The cross-referenced sequences, and mappings to them, are added to the
+ * alignment dataset.
*
- * @param seqs
- * sequences whose xrefs are being retrieved
- * @param dna
- * true if sequences are nucleotide
* @param source
- * @param al
- * alignment to search for cross-referenced sequences (and possibly
- * add to)
- * @return products (as dataset sequences)
+ * @return cross-referenced sequences (as dataset sequences)
*/
- public static Alignment findXrefSequences(SequenceI[] seqs,
- final boolean dna, final String source, AlignmentI al)
+ public Alignment findXrefSequences(String source)
{
- AlignmentI dataset = al.getDataset() == null ? al : al.getDataset();
+
List<SequenceI> rseqs = new ArrayList<SequenceI>();
AlignedCodonFrame cf = new AlignedCodonFrame();
- for (SequenceI seq : seqs)
+ SequenceIdMatcher matcher = new SequenceIdMatcher(
+ dataset.getSequences());
+
+ for (SequenceI seq : fromSeqs)
{
SequenceI dss = seq;
while (dss.getDatasetSequence() != null)
dss = dss.getDatasetSequence();
}
boolean found = false;
- DBRefEntry[] xrfs = CrossRef.findXDbRefs(dna, dss.getDBRefs());
+ DBRefEntry[] xrfs = DBRefUtils
+ .selectDbRefs(!fromDna, 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.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);
+ /*
+ * found no suitable dbrefs on sequence - look for sequences in the
+ * alignment which share a dbref with this one
+ */
+ DBRefEntry[] lrfs = DBRefUtils.selectDbRefs(fromDna,
+ 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 = searchDatasetXrefs(dss, lrfs, rseqs, cf);
}
- for (int r = 0; xrfs != null && r < xrfs.length; r++)
+ if (xrfs == null && !found)
{
- DBRefEntry xref = xrfs[r];
- if (source != null && !source.equals(xref.getSource()))
- {
- continue;
- }
+ /*
+ * no dbref to source on this sequence or matched
+ * complementary sequence in the dataset
+ */
+ continue;
+ }
+ List<DBRefEntry> sourceRefs = DBRefUtils.searchRefsForSource(xrfs,
+ source);
+ Iterator<DBRefEntry> refIterator = sourceRefs.iterator();
+ while (refIterator.hasNext())
+ {
+ DBRefEntry xref = refIterator.next();
+ found = false;
if (xref.hasMap())
{
- if (xref.getMap().getTo() != null)
+ SequenceI mappedTo = xref.getMap().getTo();
+ if (mappedTo != null)
{
- SequenceI rsq = new Sequence(xref.getMap().getTo());
+ /*
+ * dbref contains the sequence it maps to; add it to the
+ * results unless we have done so already (could happen if
+ * fetching xrefs for sequences which have xrefs in common)
+ * for example: UNIPROT {P0CE19, P0CE20} -> EMBL {J03321, X06707}
+ */
+ found = true;
+ /*
+ * problem: matcher.findIdMatch() is lenient - returns a sequence
+ * with a dbref to the search arg e.g. ENST for ENSP - wrong
+ * but findInDataset() matches ENSP when looking for Uniprot...
+ */
+ SequenceI matchInDataset = findInDataset(xref);
+ /*matcher.findIdMatch(mappedTo);*/
+ if (matchInDataset != null)
+ {
+ if (!rseqs.contains(matchInDataset))
+ {
+ rseqs.add(matchInDataset);
+ }
+ refIterator.remove();
+ continue;
+ }
+ SequenceI rsq = new Sequence(mappedTo);
rseqs.add(rsq);
- if (xref.getMap().getMap().getFromRatio() != xref
- .getMap().getMap().getToRatio())
+ if (xref.getMap().getMap().getFromRatio() != xref.getMap()
+ .getMap().getToRatio())
{
// get sense of map correct for adding to product alignment.
- if (dna)
+ if (fromDna)
{
// map is from dna seq to a protein product
cf.addMap(dss, rsq, xref.getMap().getMap());
cf.addMap(rsq, dss, xref.getMap().getMap().getInverse());
}
}
- found = true;
}
}
+
if (!found)
{
- // do a bit more work - search for sequences with references matching
- // xrefs on this sequence.
- if (dataset != null)
+ SequenceI matchedSeq = matcher.findIdMatch(xref.getSource() + "|"
+ + xref.getAccessionId());
+ if (matchedSeq != null)
{
- found |= searchDataset(dss, xref, dataset, rseqs, cf, false,
- !dna);
- if (found)
+ if (constructMapping(seq, matchedSeq, xref, cf))
{
- xrfs[r] = null; // we've recovered seqs for this one.
+ found = true;
}
}
}
+
+ if (!found)
+ {
+ // do a bit more work - search for sequences with references matching
+ // xrefs on this sequence.
+ found = searchDataset(dss, xref, rseqs, cf, false);
+ }
+ if (found)
+ {
+ refIterator.remove();
+ }
}
- if (!found)
+
+ /*
+ * fetch from source database any dbrefs we haven't resolved up to here
+ */
+ if (!sourceRefs.isEmpty())
{
- if (xrfs != null && xrfs.length > 0)
+ ASequenceFetcher sftch = SequenceFetcherFactory
+ .getSequenceFetcher();
+ SequenceI[] retrieved = null;
+ try
{
- // 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();
- SequenceI[] retrieved = null;
- int l = xrfs.length;
- for (int r = 0; r < xrfs.length; r++)
- {
- // filter out any irrelevant or irretrievable references
- if (xrfs[r] == null
- || ((source != null && !source.equals(xrfs[r]
- .getSource())) || !sftch.isFetchable(xrfs[r]
- .getSource())))
- {
- l--;
- xrfs[r] = null;
- }
- }
- if (l > 0)
- {
- // 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, !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 : "
- + seq.getName());
- e.printStackTrace();
- }
+ retrieved = sftch.getSequences(sourceRefs, !fromDna);
+ } catch (Exception e)
+ {
+ System.err
+ .println("Problem whilst retrieving cross references for Sequence : "
+ + seq.getName());
+ e.printStackTrace();
+ }
- if (retrieved != null)
+ if (retrieved != null)
+ {
+ updateDbrefMappings(seq, xrfs, retrieved, cf);
+ for (SequenceI retrievedSequence : retrieved)
+ {
+ SequenceI retrievedDss = retrievedSequence.getDatasetSequence() == null ? retrievedSequence
+ : retrievedSequence.getDatasetSequence();
+ DBRefEntry[] dbr = retrievedSequence.getDBRefs();
+ if (dbr != null)
{
- updateDbrefMappings(dna, seq, xrfs, retrieved, cf);
-
- SequenceIdMatcher matcher = new SequenceIdMatcher(
- dataset.getSequences());
- List<SequenceFeature> copiedFeatures = new ArrayList<SequenceFeature>();
- CrossRef me = new CrossRef();
- for (int rs = 0; rs < retrieved.length; rs++)
+ for (DBRefEntry dbref : dbr)
{
- // TODO: examine each sequence for 'redundancy'
- DBRefEntry[] dbr = retrieved[rs].getDBRefs();
- if (dbr != null && dbr.length > 0)
+ // find any entry where we should put in the sequence being
+ // cross-referenced into the map
+ Mapping map = dbref.getMap();
+ if (map != null)
{
- for (int di = 0; di < dbr.length; di++)
+ if (map.getTo() != null && map.getMap() != null)
{
- // find any entry where we should put in the sequence being
- // cross-referenced into the map
- Mapping map = dbr[di].getMap();
- if (map != 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)
{
- 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
+ for (DBRefEntry ref : toRefs)
{
- matcher.add(map.getTo());
+ matched.addDBRef(ref); // add or update mapping
}
- try
+ }
+ map.setTo(matched);
+ }
+ else
+ {
+ 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);
+
+ /*
+ * give the reverse reference the inverse mapping
+ * (if it doesn't have one already)
+ */
+ setReverseMapping(dss, dbref, cf);
+
+ /*
+ * copy sequence features as well, avoiding
+ * duplication (e.g. same variation from two
+ * transcripts)
+ */
+ SequenceFeature[] sfs = ms.getSequenceFeatures();
+ if (sfs != null)
{
- // 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()))
+ for (SequenceFeature feat : sfs)
{
- 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)
+ * 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)
{
- /*
- * 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);
- }
+ return super.equals(o, true);
}
- }
- cf.addMap(retrieved[rs].getDatasetSequence(),
- dss, map.getMap());
- }
- else
- {
- cf.addMap(retrieved[rs].getDatasetSequence(),
- map.getTo(), map.getMap());
+ };
+ dss.addSequenceFeature(newFeature);
}
- } catch (Exception e)
- {
- System.err
- .println("Exception when consolidating Mapped sequence set...");
- e.printStackTrace(System.err);
}
}
+ cf.addMap(retrievedDss, map.getTo(), map.getMap());
+ } catch (Exception e)
+ {
+ System.err
+ .println("Exception when consolidating Mapped sequence set...");
+ e.printStackTrace(System.err);
}
}
}
- retrieved[rs].updatePDBIds();
- rseqs.add(retrieved[rs]);
}
}
+ retrievedSequence.updatePDBIds();
+ rseqs.add(retrievedDss);
+ dataset.addSequence(retrievedDss);
+ matcher.add(retrievedDss);
}
}
}
if (rseqs.size() > 0)
{
ral = new Alignment(rseqs.toArray(new SequenceI[rseqs.size()]));
- if (cf != null && !cf.isEmpty())
+ if (!cf.isEmpty())
{
- ral.addCodonFrame(cf);
+ dataset.addCodonFrame(cf);
}
}
return ral;
}
/**
+ * Sets the inverse sequence mapping in the corresponding dbref of the mapped
+ * to sequence (if any). This is used after fetching a cross-referenced
+ * sequence, if the fetched sequence has a mapping to the original sequence,
+ * to set the mapping in the original sequence's dbref.
+ *
+ * @param mapFrom
+ * the sequence mapped from
+ * @param dbref
+ * @param mappings
+ */
+ void setReverseMapping(SequenceI mapFrom, DBRefEntry dbref,
+ AlignedCodonFrame mappings)
+ {
+ SequenceI mapTo = dbref.getMap().getTo();
+ if (mapTo == null)
+ {
+ return;
+ }
+ DBRefEntry[] dbrefs = mapTo.getDBRefs();
+ if (dbrefs == null)
+ {
+ return;
+ }
+ for (DBRefEntry toRef : dbrefs)
+ {
+ if (toRef.hasMap() && mapFrom == toRef.getMap().getTo())
+ {
+ /*
+ * found the reverse dbref; update its mapping if null
+ */
+ if (toRef.getMap().getMap() == null)
+ {
+ MapList inverse = dbref.getMap().getMap().getInverse();
+ toRef.getMap().setMap(inverse);
+ mappings.addMap(mapTo, mapFrom, inverse);
+ }
+ }
+ }
+ }
+
+ /**
+ * Returns the first identical sequence in the dataset if any, else null
+ *
+ * @param xref
+ * @return
+ */
+ SequenceI findInDataset(DBRefEntry xref)
+ {
+ if (xref == null || !xref.hasMap() || xref.getMap().getTo() == null)
+ {
+ return null;
+ }
+ SequenceI mapsTo = xref.getMap().getTo();
+ String name = xref.getAccessionId();
+ String name2 = xref.getSource() + "|" + name;
+ SequenceI dss = mapsTo.getDatasetSequence() == null ? mapsTo : mapsTo
+ .getDatasetSequence();
+ for (SequenceI seq : dataset.getSequences())
+ {
+ /*
+ * clumsy alternative to using SequenceIdMatcher which currently
+ * returns sequences with a dbref to the matched accession id
+ * which we don't want
+ */
+ if (name.equals(seq.getName()) || seq.getName().startsWith(name2))
+ {
+ if (sameSequence(seq, dss))
+ {
+ return seq;
+ }
+ }
+ }
+ return null;
+ }
+
+ /**
+ * Answers true if seq1 and seq2 contain exactly the same characters (ignoring
+ * case), else false. This method compares the lengths, then each character in
+ * turn, in order to 'fail fast'. For case-sensitive comparison, it would be
+ * possible to use Arrays.equals(seq1.getSequence(), seq2.getSequence()).
+ *
+ * @param seq1
+ * @param seq2
+ * @return
+ */
+ // TODO move to Sequence / SequenceI
+ static boolean sameSequence(SequenceI seq1, SequenceI seq2)
+ {
+ if (seq1 == seq2)
+ {
+ return true;
+ }
+ if (seq1 == null || seq2 == null)
+ {
+ return false;
+ }
+ char[] c1 = seq1.getSequence();
+ char[] c2 = seq2.getSequence();
+ if (c1.length != c2.length)
+ {
+ return false;
+ }
+ for (int i = 0; i < c1.length; i++)
+ {
+ int diff = c1[i] - c2[i];
+ /*
+ * same char or differ in case only ('a'-'A' == 32)
+ */
+ if (diff != 0 && diff != 32 && diff != -32)
+ {
+ return false;
+ }
+ }
+ return true;
+ }
+
+ /**
* 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,
+ void updateDbrefMappings(SequenceI mapFrom,
DBRefEntry[] xrefs, SequenceI[] retrieved, AlignedCodonFrame acf)
{
SequenceIdMatcher matcher = new SequenceIdMatcher(retrieved);
}
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;
- }
+ constructMapping(mapFrom, seq, xref, acf);
}
}
}
}
/**
+ * Tries to make a mapping between sequences. If successful, adds the mapping
+ * to the dbref and the mappings collection and answers true, otherwise
+ * answers false. The following methods of making are mapping are tried in
+ * turn:
+ * <ul>
+ * <li>if 'mapTo' holds a mapping to 'mapFrom', take the inverse; this is, for
+ * example, the case after fetching EMBL cross-references for a Uniprot
+ * sequence</li>
+ * <li>else check if the dna translates exactly to the protein (give or take
+ * start and stop codons></li>
+ * <li>else try to map based on CDS features on the dna sequence</li>
+ * </ul>
+ *
+ * @param mapFrom
+ * @param mapTo
+ * @param xref
+ * @param mappings
+ * @return
+ */
+ boolean constructMapping(SequenceI mapFrom, SequenceI mapTo,
+ DBRefEntry xref, AlignedCodonFrame mappings)
+ {
+ MapList mapping = null;
+
+ /*
+ * look for a reverse mapping, if found make its inverse
+ */
+ if (mapTo.getDBRefs() != null)
+ {
+ for (DBRefEntry dbref : mapTo.getDBRefs())
+ {
+ String name = dbref.getSource() + "|" + dbref.getAccessionId();
+ if (dbref.hasMap() && mapFrom.getName().startsWith(name))
+ {
+ /*
+ * looks like we've found a map from 'mapTo' to 'mapFrom'
+ * - invert it to make the mapping the other way
+ */
+ MapList reverse = dbref.getMap().getMap().getInverse();
+ xref.setMap(new Mapping(mapTo, reverse));
+ mappings.addMap(mapFrom, mapTo, reverse);
+ return true;
+ }
+ }
+ }
+
+ if (fromDna)
+ {
+ mapping = AlignmentUtils.mapCdnaToProtein(mapTo, mapFrom);
+ }
+ else
+ {
+ mapping = AlignmentUtils.mapCdnaToProtein(mapFrom, mapTo);
+ if (mapping != null)
+ {
+ mapping = mapping.getInverse();
+ }
+ }
+ if (mapping == null)
+ {
+ return false;
+ }
+ xref.setMap(new Mapping(mapTo, mapping));
+ if (fromDna)
+ {
+ AlignmentUtils.computeProteinFeatures(mapFrom, mapTo, mapping);
+ mappings.addMap(mapFrom, mapTo, mapping);
+ }
+ else
+ {
+ mappings.addMap(mapTo, mapFrom, mapping.getInverse());
+ }
+
+ return true;
+ }
+
+ /**
* 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.
*
* @param sequenceI
* @param lrfs
- * @param dataset
* @param rseqs
* @return true if matches were found.
*/
- private static boolean searchDatasetXrefs(SequenceI sequenceI,
- boolean dna, DBRefEntry[] lrfs, AlignmentI dataset,
- List<SequenceI> rseqs, AlignedCodonFrame cf)
+ private boolean searchDatasetXrefs(SequenceI sequenceI,
+ DBRefEntry[] lrfs, List<SequenceI> rseqs, AlignedCodonFrame cf)
{
boolean found = false;
if (lrfs == null)
// add in wildcards
xref.setVersion(null);
xref.setMap(null);
- found = searchDataset(sequenceI, xref, dataset, rseqs, cf, false, dna);
+ found |= searchDataset(sequenceI, xref, rseqs, cf, false);
}
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
- * @param dataset
+ * a cross-reference to try to match
* @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<SequenceI> 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
- * search dna or protein xrefs (if direct=false)
* @return true if relationship found and sequence added.
*/
- public static boolean searchDataset(SequenceI sequenceI, DBRefEntry xrf,
- AlignmentI dataset, List<SequenceI> rseqs, AlignedCodonFrame cf,
- boolean direct, boolean dna)
+ boolean searchDataset(SequenceI sequenceI, DBRefEntry xrf,
+ List<SequenceI> rseqs, AlignedCodonFrame cf, boolean direct)
{
boolean found = false;
- SequenceI[] typer = new SequenceI[1];
if (dataset == null)
{
return false;
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())
+ {
+ continue;
+ }
+ /*
+ * only look at same molecule type if 'direct', or
+ * complementary type if !direct
+ */
{
- // check if this is the correct sequence type
+ boolean isDna = Comparison
+ .isNucleotide(new SequenceI[] { nxt });
+ if (direct ? (isDna != fromDna) : (isDna == fromDna))
{
- 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.getDBRefs(), 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();
+ List<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(!fromDna, poss);
+ cands = DBRefUtils.searchRefs(poss, xrf);
+ }
+ if (!cands.isEmpty())
+ {
+ if (!rseqs.contains(nxt))
{
- if (!rseqs.contains(nxt))
+ found = true;
+ rseqs.add(nxt);
+ if (cf != null)
{
- 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++)
+ for (DBRefEntry candidate : cands)
{
- if (cands[r].hasMap())
+ Mapping mapping = candidate.getMap();
+ if (mapping != null)
{
- if (cands[r].getMap().getTo() != null
- && cands[r].getMap().getMap().getFromRatio() != cands[r]
- .getMap().getMap().getToRatio())
+ MapList map = mapping.getMap();
+ if (mapping.getTo() != null
+ && map.getFromRatio() != map.getToRatio())
{
- foundmap = true;
// get sense of map correct for adding to product
// alignment.
- if (dna)
+ if (fromDna)
{
// map is from dna seq to a protein product
- cf.addMap(sequenceI, nxt, cands[r].getMap()
- .getMap());
+ cf.addMap(sequenceI, nxt, map);
}
else
{
// map should be from protein seq to its coding dna
- cf.addMap(nxt, sequenceI, cands[r].getMap()
- .getMap().getInverse());
+ 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<prod.length; p++) { System.out.println("Prod "+p+":
- * "+prod[p].getDisplayId(true)); } } } } else {
- * System.out.println("Trying getProducts for
- * "+al.getSequenceAt(0).getDisplayId(true));
- * System.out.println("Search DS Xref for: "+(dna ? "dna" : "prot"));
- * // have a bash at finding the products amongst all the retrieved
- * sequences. SequenceI[] prod =
- * jalview.analysis.CrossRef.findXrefSequences(al
- * .getSequencesArray(), dna, null, ds); System.out.println("Found " +
- * ((prod == null) ? "no" : "" + prod.length) + " products"); if
- * (prod!=null) { // select non-equivalent sequences from dataset list
- * for (int p=0; p<prod.length; p++) { System.out.println("Prod "+p+":
- * "+prod[p].getDisplayId(true)); } } } }
- */
}
{
if (s != null)
{
- id = new String(s.toLowerCase());
+ id = s.toLowerCase();
}
else
{
.indexOf(s.charAt(id.length())) > -1)) : false;
}
}
+
+ /**
+ * toString method returns the wrapped sequence id. For debugging purposes
+ * only, behaviour not guaranteed not to change.
+ */
+ @Override
+ public String toString()
+ {
+ return id;
+ }
}
}
import jalview.util.MapList;
import jalview.util.MappingUtils;
+import java.util.AbstractList;
import java.util.ArrayList;
import java.util.List;
/*
* Data bean to hold mappings from one sequence to another
*/
- private class SequenceToSequenceMapping
+ public class SequenceToSequenceMapping
{
private SequenceI fromSeq;
return String.format("From %s %s", fromSeq.getName(),
mapping.toString());
}
+
+ /**
+ * Returns a hashCode derived from the hashcodes of the mappings
+ *
+ * @see SequenceToSequenceMapping#hashCode()
+ */
+ @Override
+ public int hashCode()
+ {
+ return mappings.hashCode();
+ }
+
+ /**
+ * Answers true if the objects hold the same mapping between the same two
+ * sequences
+ *
+ * @see Mapping#equals
+ */
+ @Override
+ public boolean equals(Object obj)
+ {
+ if (!(obj instanceof SequenceToSequenceMapping))
+ {
+ return false;
+ }
+ SequenceToSequenceMapping that = (SequenceToSequenceMapping) obj;
+ if (this.mapping == null)
+ {
+ return that.mapping == null;
+ }
+ return this.mapping.equals(that.mapping);
+ }
+
+ public SequenceI getFromSeq()
+ {
+ return fromSeq;
+ }
+
+ public Mapping getMapping()
+ {
+ return mapping;
+ }
}
private List<SequenceToSequenceMapping> mappings;
/*
* if we already hold a mapping between these sequences, just add to it
+ * note that 'adding' a duplicate map does nothing; this protects against
+ * creating duplicate mappings in AlignedCodonFrame
*/
for (SequenceToSequenceMapping ssm : mappings)
{
}
return null;
}
+
+ /**
+ * Returns a hashcode derived from the list of sequence mappings
+ *
+ * @see SequenceToSequenceMapping#hashCode()
+ * @see AbstractList#hashCode()
+ */
+ @Override
+ public int hashCode()
+ {
+ return this.mappings.hashCode();
+ }
+
+ /**
+ * Two AlignedCodonFrame objects are equal if they hold the same ordered list
+ * of mappings
+ *
+ * @see SequenceToSequenceMapping#
+ */
+ @Override
+ public boolean equals(Object obj)
+ {
+ if (!(obj instanceof AlignedCodonFrame))
+ {
+ return false;
+ }
+ return this.mappings.equals(((AlignedCodonFrame) obj).mappings);
+ }
+
+ public List<SequenceToSequenceMapping> getMappings()
+ {
+ return mappings;
+ }
}
/**
* Equals that compares both the to references and MapList mappings.
*
- * @param other
+ * @param o
* @return
+ * @see MapList#equals
*/
@Override
public boolean equals(Object o)
{
- // TODO should override Object.hashCode() to ensure that equal objects have
- // equal hashcodes
if (o == null || !(o instanceof Mapping))
{
return false;
}
/**
+ * Returns a hashCode made from the sequence and maplist
+ */
+ @Override
+ public int hashCode()
+ {
+ int hashCode = (this.to == null ? 1 : this.to.hashCode());
+ if (this.map != null)
+ {
+ hashCode = hashCode * 31 + this.map.hashCode();
+ }
+
+ return hashCode;
+ }
+
+ /**
* get the 'initial' position in the associated sequence for a position in the
* mapped reference frame
*
{
for (DBRefEntry dbref : dbRefs)
{
+ dbref.setSource(DBRefUtils.getCanonicalName(dbref.getSource()));
dna.addDBRef(dbref);
}
}
{
for (DBRefEntry dbref : feature.dbRefs)
{
+ /*
+ * convert UniProtKB/Swiss-Prot to UNIPROT
+ */
+ dbref.setSource(DBRefUtils.getCanonicalName(dbref.getSource()));
dna.addDBRef(dbref);
}
}
}
/*
- * add dbRefs to sequence, and mappings for Uniprot xrefs
+ * add mappings for Uniprot xrefs
*/
if (feature.dbRefs != null)
{
boolean mappingUsed = false;
for (DBRefEntry ref : feature.dbRefs)
{
- ref.setSource(DBRefUtils.getCanonicalName(ref.getSource()));
if (ref.getSource().equals(DBRefSource.UNIPROT))
{
String proteinSeqName = DBRefSource.UNIPROT + "|"
}
}
}
- dna.addDBRef(ref);
}
if (noProteinDbref && product != null)
{
}
/**
- * Searches selected sequences for xRef products and builds the Show
- * Cross-References menu (formerly called Show Products)
+ * Searches the alignment sequences for xRefs and builds the Show
+ * Cross-References menu (formerly called Show Products), with database
+ * sources for which cross-references are found (protein sources for a
+ * nucleotide alignment and vice versa)
*
- * @return true if Show Cross-references menu should be enabled.
+ * @return true if Show Cross-references menu should be enabled
*/
public boolean canShowProducts()
{
- SequenceI[] selection = viewport.getSequenceSelection();
+ SequenceI[] seqs = viewport.getAlignment().getSequencesArray();
AlignmentI dataset = viewport.getAlignment().getDataset();
boolean showp = false;
try
{
showProducts.removeAll();
final boolean dna = viewport.getAlignment().isNucleotide();
- String[] ptypes = (selection == null || selection.length == 0) ? null
- : CrossRef.findSequenceXrefTypes(dna, selection, dataset);
+ List<String> ptypes = (seqs == null || seqs.length == 0) ? null
+ : new CrossRef(seqs, dataset)
+ .findXrefSourcesForSequences();
- for (int t = 0; ptypes != null && t < ptypes.length; t++)
+ for (final String source : ptypes)
{
showp = true;
final AlignFrame af = this;
- final String source = ptypes[t];
- JMenuItem xtype = new JMenuItem(ptypes[t]);
+ JMenuItem xtype = new JMenuItem(source);
xtype.addActionListener(new ActionListener()
{
-
@Override
public void actionPerformed(ActionEvent e)
{
showProductsFor(af.viewport.getSequenceSelection(), dna, source);
}
-
});
showProducts.add(xtype);
}
showProducts.setEnabled(showp);
} catch (Exception e)
{
- jalview.bin.Cache.log
+ Cache.log
.warn("canShowProducts threw an exception - please report to help@jalview.org",
e);
return false;
{
AlignmentI alignment = AlignFrame.this.getViewport()
.getAlignment();
- AlignmentI xrefs = CrossRef.findXrefSequences(sel, dna, source,
- alignment);
- if (xrefs != null)
+ AlignmentI dataset = alignment.getDataset() == null ? alignment
+ : alignment.getDataset();
+ AlignmentI xrefs = new CrossRef(sel, alignment)
+ .findXrefSequences(source);
+ if (xrefs == null)
{
- /*
- * get display scheme (if any) to apply to features
- */
- FeatureSettingsModelI featureColourScheme = new SequenceFetcher()
- .getFeatureColourScheme(source);
+ return;
+ }
+ /*
+ * get display scheme (if any) to apply to features
+ */
+ FeatureSettingsModelI featureColourScheme = new SequenceFetcher()
+ .getFeatureColourScheme(source);
+
+ AlignmentI xrefsAlignment = makeCrossReferencesAlignment(dataset,
+ xrefs);
- AlignmentI al = makeCrossReferencesAlignment(
- alignment.getDataset(), xrefs);
+ AlignFrame newFrame = new AlignFrame(xrefsAlignment, DEFAULT_WIDTH,
+ DEFAULT_HEIGHT);
+ if (Cache.getDefault("HIDE_INTRONS", true))
+ {
+ newFrame.hideFeatureColumns(SequenceOntologyI.EXON, false);
+ }
+ String newtitle = String.format("%s %s %s", MessageManager
+ .getString(dna ? "label.proteins" : "label.nucleotides"),
+ MessageManager.getString("label.for"), getTitle());
+ newFrame.setTitle(newtitle);
- AlignFrame newFrame = new AlignFrame(al, DEFAULT_WIDTH,
+ if (!Cache.getDefault(Preferences.ENABLE_SPLIT_FRAME, true))
+ {
+ /*
+ * split frame display is turned off in preferences file
+ */
+ Desktop.addInternalFrame(newFrame, newtitle, DEFAULT_WIDTH,
DEFAULT_HEIGHT);
- if (Cache.getDefault("HIDE_INTRONS", true))
- {
- newFrame.hideFeatureColumns(SequenceOntologyI.EXON, false);
- }
- String newtitle = String.format("%s %s %s",
- MessageManager.getString(dna ? "label.proteins"
- : "label.nucleotides"), MessageManager
- .getString("label.for"), getTitle());
- newFrame.setTitle(newtitle);
+ return; // via finally clause
+ }
- if (!Cache.getDefault(Preferences.ENABLE_SPLIT_FRAME, true))
+ /*
+ * Make a copy of this alignment (sharing the same dataset
+ * sequences). If we are DNA, drop introns and update mappings
+ */
+ AlignmentI copyAlignment = null;
+ final SequenceI[] sequenceSelection = AlignFrame.this.viewport
+ .getSequenceSelection();
+ // List<AlignedCodonFrame> cf = xrefs.getCodonFrames();
+ boolean copyAlignmentIsAligned = false;
+ if (dna)
+ {
+ copyAlignment = AlignmentUtils.makeCdsAlignment(
+ sequenceSelection, dataset);
+ if (copyAlignment.getHeight() == 0)
{
- /*
- * split frame display is turned off in preferences file
- */
- Desktop.addInternalFrame(newFrame, newtitle, DEFAULT_WIDTH,
- DEFAULT_HEIGHT);
- return; // via finally clause
+ System.err.println("Failed to make CDS alignment");
}
/*
- * Make a copy of this alignment (sharing the same dataset
- * sequences). If we are DNA, drop introns and update mappings
+ * pending getting Embl transcripts to 'align',
+ * we are only doing this for Ensembl
*/
- AlignmentI copyAlignment = null;
- final SequenceI[] sequenceSelection = AlignFrame.this.viewport
- .getSequenceSelection();
- List<AlignedCodonFrame> cf = xrefs.getCodonFrames();
- boolean copyAlignmentIsAligned = false;
- if (dna)
- {
- copyAlignment = AlignmentUtils.makeCdsAlignment(
- sequenceSelection, cf, alignment);
- if (copyAlignment.getHeight() == 0)
- {
- System.err.println("Failed to make CDS alignment");
- }
- al.getCodonFrames().clear();
- al.addCodonFrames(copyAlignment.getCodonFrames());
- al.addCodonFrames(cf);
-
- /*
- * pending getting Embl transcripts to 'align',
- * we are only doing this for Ensembl
- */
- // TODO proper criteria for 'can align as cdna'
- if (DBRefSource.ENSEMBL.equalsIgnoreCase(source)
- || AlignmentUtils.looksLikeEnsembl(alignment))
- {
- copyAlignment.alignAs(alignment);
- copyAlignmentIsAligned = true;
- }
- }
- else
+ // TODO proper criteria for 'can align as cdna'
+ if (DBRefSource.ENSEMBL.equalsIgnoreCase(source)
+ || AlignmentUtils.looksLikeEnsembl(alignment))
{
- copyAlignment = AlignmentUtils.makeCopyAlignment(
- sequenceSelection, xrefs.getSequencesArray());
- copyAlignment.addCodonFrames(cf);
- al.addCodonFrames(copyAlignment.getCodonFrames());
- al.addCodonFrames(cf);
+ copyAlignment.alignAs(alignment);
+ copyAlignmentIsAligned = true;
}
- copyAlignment.setGapCharacter(AlignFrame.this.viewport
- .getGapCharacter());
+ }
+ else
+ {
+ copyAlignment = AlignmentUtils.makeCopyAlignment(
+ sequenceSelection, xrefs.getSequencesArray());
+ }
+ copyAlignment.setGapCharacter(AlignFrame.this.viewport
+ .getGapCharacter());
- StructureSelectionManager ssm = StructureSelectionManager
- .getStructureSelectionManager(Desktop.instance);
- ssm.registerMappings(cf);
+ StructureSelectionManager ssm = StructureSelectionManager
+ .getStructureSelectionManager(Desktop.instance);
- if (copyAlignment.getHeight() <= 0)
- {
- System.err.println("No Sequences generated for xRef type "
- + source);
- return;
- }
+ /*
+ * register any new mappings for sequence mouseover etc
+ * (will not duplicate any previously registered mappings)
+ */
+ ssm.registerMappings(dataset.getCodonFrames());
+
+ if (copyAlignment.getHeight() <= 0)
+ {
+ System.err.println("No Sequences generated for xRef type "
+ + source);
+ return;
+ }
+ /*
+ * align protein to dna
+ */
+ if (dna && copyAlignmentIsAligned)
+ {
+ xrefsAlignment.alignAs(copyAlignment);
+ }
+ else
+ {
/*
- * align protein to dna
+ * align cdna to protein - currently only if
+ * fetching and aligning Ensembl transcripts!
*/
- if (dna && copyAlignmentIsAligned)
+ if (DBRefSource.ENSEMBL.equalsIgnoreCase(source))
{
- al.alignAs(copyAlignment);
- }
- else
- {
- /*
- * align cdna to protein - currently only if
- * fetching and aligning Ensembl transcripts!
- */
- if (DBRefSource.ENSEMBL.equalsIgnoreCase(source))
- {
- copyAlignment.alignAs(al);
- }
+ copyAlignment.alignAs(xrefsAlignment);
}
+ }
- AlignFrame copyThis = new AlignFrame(copyAlignment,
- AlignFrame.DEFAULT_WIDTH, AlignFrame.DEFAULT_HEIGHT);
- copyThis.setTitle(AlignFrame.this.getTitle());
+ AlignFrame copyThis = new AlignFrame(copyAlignment,
+ AlignFrame.DEFAULT_WIDTH, AlignFrame.DEFAULT_HEIGHT);
+ copyThis.setTitle(AlignFrame.this.getTitle());
- boolean showSequenceFeatures = viewport
- .isShowSequenceFeatures();
- newFrame.setShowSeqFeatures(showSequenceFeatures);
- copyThis.setShowSeqFeatures(showSequenceFeatures);
- FeatureRenderer myFeatureStyling = alignPanel.getSeqPanel().seqCanvas
- .getFeatureRenderer();
+ boolean showSequenceFeatures = viewport.isShowSequenceFeatures();
+ newFrame.setShowSeqFeatures(showSequenceFeatures);
+ copyThis.setShowSeqFeatures(showSequenceFeatures);
+ FeatureRenderer myFeatureStyling = alignPanel.getSeqPanel().seqCanvas
+ .getFeatureRenderer();
- /*
- * copy feature rendering settings to split frame
- */
- newFrame.alignPanel.getSeqPanel().seqCanvas
- .getFeatureRenderer()
- .transferSettings(myFeatureStyling);
- copyThis.alignPanel.getSeqPanel().seqCanvas
- .getFeatureRenderer()
- .transferSettings(myFeatureStyling);
+ /*
+ * copy feature rendering settings to split frame
+ */
+ newFrame.alignPanel.getSeqPanel().seqCanvas.getFeatureRenderer()
+ .transferSettings(myFeatureStyling);
+ copyThis.alignPanel.getSeqPanel().seqCanvas.getFeatureRenderer()
+ .transferSettings(myFeatureStyling);
- /*
- * apply 'database source' feature configuration
- * if any was found
- */
- // TODO is this the feature colouring for the original
- // alignment or the fetched xrefs? either could be Ensembl
- newFrame.getViewport().applyFeaturesStyle(featureColourScheme);
- copyThis.getViewport().applyFeaturesStyle(featureColourScheme);
-
- SplitFrame sf = new SplitFrame(dna ? copyThis : newFrame,
- dna ? newFrame : copyThis);
- newFrame.setVisible(true);
- copyThis.setVisible(true);
- String linkedTitle = MessageManager
- .getString("label.linked_view_title");
- Desktop.addInternalFrame(sf, linkedTitle, -1, -1);
- sf.adjustDivider();
- }
- } catch (Exception e)
- {
- Cache.log.error("Exception when finding crossreferences", e);
+ /*
+ * apply 'database source' feature configuration
+ * if any was found
+ */
+ // TODO is this the feature colouring for the original
+ // alignment or the fetched xrefs? either could be Ensembl
+ newFrame.getViewport().applyFeaturesStyle(featureColourScheme);
+ copyThis.getViewport().applyFeaturesStyle(featureColourScheme);
+
+ SplitFrame sf = new SplitFrame(dna ? copyThis : newFrame,
+ dna ? newFrame : copyThis);
+ newFrame.setVisible(true);
+ copyThis.setVisible(true);
+ String linkedTitle = MessageManager
+ .getString("label.linked_view_title");
+ Desktop.addInternalFrame(sf, linkedTitle, -1, -1);
+ sf.adjustDivider();
} catch (OutOfMemoryError e)
{
new OOMWarning("whilst fetching crossreferences", e);
}
/**
- * Makes an alignment containing the given sequences. If this is of the
- * same type as the given dataset (nucleotide/protein), then the new
- * alignment shares the same dataset, and its dataset sequences are added
- * to it. Otherwise a new dataset sequence is created for the
- * cross-references.
+ * Makes an alignment containing the given sequences, and adds them to the
+ * given dataset, which is also set as the dataset for the new alignment
*
* @param dataset
* @param seqs
protected AlignmentI makeCrossReferencesAlignment(AlignmentI dataset,
AlignmentI seqs)
{
- boolean sameType = dataset.isNucleotide() == seqs.isNucleotide();
-
SequenceI[] sprods = new SequenceI[seqs.getHeight()];
for (int s = 0; s < sprods.length; s++)
{
sprods[s] = (seqs.getSequenceAt(s)).deriveSequence();
- if (sameType)
+ if (dataset.getSequences() == null
+ || !dataset.getSequences().contains(
+ sprods[s].getDatasetSequence()))
{
- if (dataset.getSequences() == null
- || !dataset.getSequences().contains(
- sprods[s].getDatasetSequence()))
- {
- dataset.addSequence(sprods[s].getDatasetSequence());
- }
+ dataset.addSequence(sprods[s].getDatasetSequence());
}
sprods[s].updatePDBIds();
}
Alignment al = new Alignment(sprods);
- if (sameType)
- {
- al.setDataset((Alignment) dataset);
- }
- else
- {
- al.createDatasetAlignment();
- }
+ al.setDataset((Alignment) dataset);
return al;
}
Cache.log.info(
"Error retrieving " + accession
+ " from " + proxy.getDbName(), e);
- } finally
- {
- return success;
}
+ return success;
}
/**
for (String q : queries)
{
- DBRefEntry[] found = null;
DBRefEntry dbr = new DBRefEntry();
dbr.setSource(proxy.getDbSource());
dbr.setVersion(null);
{
if (rs[r] != null)
{
- found = DBRefUtils.searchRefs(rs[r].getDBRefs(), accId);
- if (found != null && found.length > 0)
+ List<DBRefEntry> found = DBRefUtils.searchRefs(rs[r].getDBRefs(),
+ accId);
+ if (!found.isEmpty())
{
rfound = true;
break;
}
/**
+ * Overloaded method signature to test whether a single sequence is nucleotide
+ * (that is, more than 85% CGTA)
+ *
+ * @param seq
+ * @return
+ */
+ public static final boolean isNucleotide(SequenceI seq)
+ {
+ return isNucleotide(new SequenceI[] { seq });
+ }
+
+ /**
* Answers true if more than 85% of the sequence residues (ignoring gaps) are
* A, G, C, T or U, else false. This is just a heuristic guess and may give a
* wrong answer (as AGCT are also amino acid codes).
}
/**
+ * Returns those DBRefEntry objects whose source identifier (once converted to
+ * Jalview's canonical form) is in the list of sources to search for. Returns
+ * null if no matches found.
*
* @param dbrefs
- * array of DBRef objects to search
+ * DBRefEntry objects to search
* @param sources
- * String[] array of source DBRef IDs to retrieve
+ * array of sources to select
* @return
*/
public static DBRefEntry[] selectRefs(DBRefEntry[] dbrefs,
}
/**
- * Returns an array of those references that match the given entry, or null if
- * no matches. Currently uses a comparator which matches if
+ * Returns a (possibly empty) list of those references that match the given
+ * entry. Currently uses a comparator which matches if
* <ul>
* <li>database sources are the same</li>
* <li>accession ids are the same</li>
* pattern to match
* @return
*/
- public static DBRefEntry[] searchRefs(DBRefEntry[] ref, DBRefEntry entry)
+ public static List<DBRefEntry> searchRefs(DBRefEntry[] ref,
+ DBRefEntry entry)
{
return searchRefs(ref, entry,
matchDbAndIdAndEitherMapOrEquivalentMapList);
}
/**
- * Returns an array of those references that match the given accession id
+ * Returns a list of those references that match the given accession id
* <ul>
* <li>database sources are the same</li>
* <li>accession ids are the same</li>
* <li>both have no mapping, or the mappings are the same</li>
* </ul>
*
- * @param ref
+ * @param refs
* Set of references to search
- * @param entry
- * pattern to match
+ * @param accId
+ * accession id to match
* @return
*/
- public static DBRefEntry[] searchRefs(DBRefEntry[] ref, String accId)
+ public static List<DBRefEntry> searchRefs(DBRefEntry[] refs, String accId)
{
- return searchRefs(ref, new DBRefEntry("", "", accId), matchId);
+ return searchRefs(refs, new DBRefEntry("", "", accId), matchId);
}
/**
- * Returns an array of those references that match the given entry, according
- * to the given comparator. Returns null if no matches.
+ * Returns a (possibly empty) list of those references that match the given
+ * entry, according to the given comparator.
*
* @param refs
* an array of database references to search
* @param comparator
* @return
*/
- static DBRefEntry[] searchRefs(DBRefEntry[] refs, DBRefEntry entry,
+ static List<DBRefEntry> searchRefs(DBRefEntry[] refs, DBRefEntry entry,
DbRefComp comparator)
{
+ List<DBRefEntry> rfs = new ArrayList<DBRefEntry>();
if (refs == null || entry == null)
{
- return null;
+ return rfs;
}
- List<DBRefEntry> rfs = new ArrayList<DBRefEntry>();
for (int i = 0; i < refs.length; i++)
{
if (comparator.matches(entry, refs[i]))
rfs.add(refs[i]);
}
}
- return rfs.size() == 0 ? null : rfs.toArray(new DBRefEntry[rfs.size()]);
+ return rfs;
}
interface DbRefComp
};
/**
- * accession ID and DB must be identical. Version is ignored. No map on either
- * or map but no maplist on either or maplist of map on a is equivalent to the
- * maplist of map on b.
+ * accession ID and DB must be identical, or null on a. Version is ignored. No
+ * map on either or map but no maplist on either or maplist of map on a is
+ * equivalent to the maplist of map on b.
*/
public static DbRefComp matchDbAndIdAndEitherMapOrEquivalentMapList = new DbRefComp()
{
&& refb.getSource().equals(refa.getSource()))
{
// We dont care about version
- if (refa.getAccessionId() != null && refb.getAccessionId() != null
- && refb.getAccessionId().equals(refa.getAccessionId()))
+
+ if (refa.getAccessionId() == null
+ || refa.getAccessionId().equals(refb.getAccessionId()))
{
if (refa.getMap() == null || refb.getMap() == null)
{
|| (refb.getMap().getMap() != null
&& refa.getMap().getMap() != null && (refb
.getMap().getMap().equals(refa.getMap().getMap()))))
- { // getMap().getMap().containsEither(false,refa.getMap().getMap())
+ {
return true;
}
}
return (o1 == null ? o2.equals(o1) : o1.equals(o2));
}
+ /**
+ * Selects just the DNA or protein references from a set of references
+ *
+ * @param selectDna
+ * if true, select references to 'standard' DNA databases, else to
+ * 'standard' peptide databases
+ * @param refs
+ * a set of references to select from
+ * @return
+ */
+ public static DBRefEntry[] selectDbRefs(boolean selectDna,
+ DBRefEntry[] refs)
+ {
+ return selectRefs(refs, selectDna ? DBRefSource.DNACODINGDBS
+ : DBRefSource.PROTEINDBS);
+ // could attempt to find other cross
+ // refs here - ie PDB xrefs
+ // (not dna, not protein seq)
+ }
+
+ /**
+ * Returns the (possibly empty) list of those supplied dbrefs which have the
+ * specified source databse
+ *
+ * @param dbRefs
+ * @param source
+ * @return
+ */
+ public static List<DBRefEntry> searchRefsForSource(DBRefEntry[] dbRefs,
+ String source)
+ {
+ List<DBRefEntry> matches = new ArrayList<DBRefEntry>();
+ if (dbRefs != null && source != null)
+ {
+ for (DBRefEntry dbref : dbRefs)
+ {
+ if (source.equals(dbref.getSource()))
+ {
+ matches.add(dbref);
+ }
+ }
+ }
+ return matches;
+ }
+
}
@Override
public boolean equals(Object o)
{
- // TODO should also override hashCode to ensure equal objects have equal
- // hashcodes
if (o == null || !(o instanceof MapList))
{
return false;
}
/**
+ * Returns a hashcode made from the fromRatio, toRatio, and from/to ranges
+ */
+ @Override
+ public int hashCode()
+ {
+ int hashCode = 31 * fromRatio;
+ hashCode = 31 * hashCode + toRatio;
+ hashCode = 31 * hashCode + fromShifts.toArray().hashCode();
+ hashCode = 31 * hashCode + toShifts.toArray().hashCode();
+ return hashCode;
+ }
+
+ /**
* Returns the 'from' ranges as {[start1, end1], [start2, end2], ...}
*
* @return
{
/*
* note lowest and highest values - bearing in mind the
- * direction may be revesed
+ * direction may be reversed
*/
fromLowest = Math.min(fromLowest, Math.min(from[i], from[i + 1]));
fromHighest = Math.max(fromHighest, Math.max(from[i], from[i + 1]));
*/
public void addMapList(MapList map)
{
+ if (this.equals(map))
+ {
+ return;
+ }
this.fromLowest = Math.min(fromLowest, map.fromLowest);
this.toLowest = Math.min(toLowest, map.toLowest);
this.fromHighest = Math.max(fromHighest, map.fromHighest);
}
return forwardStrand;
}
+
}
--- /dev/null
+package jalview.ws;
+
+import jalview.ws.seqfetcher.ASequenceFetcher;
+
+public class SequenceFetcherFactory
+{
+
+ private static SequenceFetcher instance;
+
+ /**
+ * Returns a new SequenceFetcher object, or a mock object if one has been set
+ *
+ * @return
+ */
+ public static ASequenceFetcher getSequenceFetcher()
+ {
+ return instance == null ? new SequenceFetcher() : instance;
+ }
+
+ /**
+ * Set the instance object to use (intended for unit testing with mock
+ * objects).
+ *
+ * Be sure to reset to null in the tearDown method of any tests!
+ *
+ * @param sf
+ */
+ public static void setSequenceFetcher(SequenceFetcher sf)
+ {
+ instance = sf;
+ }
+}
/**
* Constructor
*/
- public ASequenceFetcher()
+ protected ASequenceFetcher()
{
super();
* if true, only fetch from nucleotide data sources, else peptide
* @return
*/
- public SequenceI[] getSequences(DBRefEntry[] refs, boolean dna)
+ public SequenceI[] getSequences(List<DBRefEntry> refs, boolean dna)
{
Vector<SequenceI> rseqs = new Vector<SequenceI>();
Hashtable<String, List<String>> queries = new Hashtable<String, List<String>>();
- for (int r = 0; r < refs.length; r++)
+ for (DBRefEntry ref : refs)
{
- if (!queries.containsKey(refs[r].getSource()))
+ if (!queries.containsKey(ref.getSource()))
{
- queries.put(refs[r].getSource(), new ArrayList<String>());
+ queries.put(ref.getSource(), new ArrayList<String>());
}
- List<String> qset = queries.get(refs[r].getSource());
- if (!qset.contains(refs[r].getAccessionId()))
+ List<String> qset = queries.get(ref.getSource());
+ if (!qset.contains(ref.getAccessionId()))
{
- qset.add(refs[r].getAccessionId());
+ qset.add(ref.getAccessionId());
}
}
Enumeration<String> e = queries.keys();
for (int is = 0; is < seqs.length; is++)
{
rseqs.addElement(seqs[is]);
- DBRefEntry[] frefs = DBRefUtils.searchRefs(seqs[is]
+ List<DBRefEntry> frefs = DBRefUtils.searchRefs(seqs[is]
.getDBRefs(), new DBRefEntry(db, null, null));
- if (frefs != null)
+ for (DBRefEntry dbr : frefs)
{
- for (DBRefEntry dbr : frefs)
- {
- queriesFound.add(dbr.getAccessionId());
- queriesMade.remove(dbr.getAccessionId());
- }
+ queriesFound.add(dbr.getAccessionId());
+ queriesMade.remove(dbr.getAccessionId());
}
seqs[is] = null;
}
AlignmentI dna = new Alignment(new SequenceI[] { dna1, dna2 });
dna.setDataset(null);
- List<AlignedCodonFrame> mappings = new ArrayList<AlignedCodonFrame>();
MapList map = new MapList(new int[] { 4, 6, 10, 12 },
new int[] { 1, 2 }, 3, 1);
AlignedCodonFrame acf = new AlignedCodonFrame();
acf.addMap(dna1.getDatasetSequence(), pep1.getDatasetSequence(), map);
- mappings.add(acf);
+ dna.addCodonFrame(acf);
map = new MapList(new int[] { 1, 3, 7, 9, 13, 15 }, new int[] { 1, 3 },
3, 1);
acf = new AlignedCodonFrame();
acf.addMap(dna2.getDatasetSequence(), pep2.getDatasetSequence(), map);
- mappings.add(acf);
+ dna.addCodonFrame(acf);
/*
* execute method under test:
*/
AlignmentI cds = AlignmentUtils.makeCdsAlignment(new SequenceI[] {
- dna1, dna2 }, mappings, dna);
+ dna1, dna2 }, dna);
assertEquals(2, cds.getSequences().size());
assertEquals("GGGTTT", cds.getSequenceAt(0)
new DBRefEntry("EMBLCDS", "4", "A12347"));
/*
+ * Create the CDS alignment
+ */
+ AlignmentI dna = new Alignment(new SequenceI[] { dna1 });
+ dna.setDataset(null);
+
+ /*
* Make the mappings from dna to protein
*/
- List<AlignedCodonFrame> mappings = new ArrayList<AlignedCodonFrame>();
// map ...GGG...TTT to GF
MapList map = new MapList(new int[] { 4, 6, 10, 12 },
new int[] { 1, 2 }, 3, 1);
AlignedCodonFrame acf = new AlignedCodonFrame();
acf.addMap(dna1.getDatasetSequence(), pep1.getDatasetSequence(), map);
- mappings.add(acf);
+ dna.addCodonFrame(acf);
// map aaa...ccc to KP
map = new MapList(new int[] { 1, 3, 7, 9 }, new int[] { 1, 2 }, 3, 1);
acf = new AlignedCodonFrame();
acf.addMap(dna1.getDatasetSequence(), pep2.getDatasetSequence(), map);
- mappings.add(acf);
+ dna.addCodonFrame(acf);
// map aaa......TTT to KF
map = new MapList(new int[] { 1, 3, 10, 12 }, new int[] { 1, 2 }, 3, 1);
acf = new AlignedCodonFrame();
acf.addMap(dna1.getDatasetSequence(), pep3.getDatasetSequence(), map);
- mappings.add(acf);
-
- /*
- * Create the CDS alignment; also augments the dna-to-protein mappings with
- * exon-to-protein and exon-to-dna mappings
- */
- AlignmentI dna = new Alignment(new SequenceI[] { dna1 });
- dna.setDataset(null);
+ dna.addCodonFrame(acf);
/*
* execute method under test
*/
AlignmentI cdsal = AlignmentUtils.makeCdsAlignment(
- new SequenceI[] { dna1 }, mappings, dna);
+ new SequenceI[] { dna1 }, dna);
/*
* Verify we have 3 cds sequences, mapped to pep1/2/3 respectively
null));
dna2.addSequenceFeature(new SequenceFeature("CDS", "cds", 16, 18, 0f,
null));
+
+ AlignmentI dna = new Alignment(new SequenceI[] { dna1, dna2, dna3 });
+ dna.setDataset(null);
- List<AlignedCodonFrame> mappings = new ArrayList<AlignedCodonFrame>();
MapList map = new MapList(new int[] { 4, 12, 16, 18 },
new int[] { 1, 4 }, 3, 1);
AlignedCodonFrame acf = new AlignedCodonFrame();
acf.addMap(dna1.getDatasetSequence(), pep1.getDatasetSequence(), map);
- mappings.add(acf);
+ dna.addCodonFrame(acf);
map = new MapList(new int[] { 4, 8, 12, 12, 16, 18 },
new int[] { 1, 3 },
3, 1);
acf = new AlignedCodonFrame();
acf.addMap(dna2.getDatasetSequence(), pep2.getDatasetSequence(), map);
- mappings.add(acf);
+ dna.addCodonFrame(acf);
- AlignmentI dna = new Alignment(new SequenceI[] { dna1, dna2, dna3 });
- dna.setDataset(null);
AlignmentI cds = AlignmentUtils.makeCdsAlignment(new SequenceI[] {
- dna1, dna2, dna3 }, mappings, dna);
+ dna1, dna2, dna3 }, dna);
List<SequenceI> cdsSeqs = cds.getSequences();
assertEquals(2, cdsSeqs.size());
assertEquals("GGGCCCTTTGGG", cdsSeqs.get(0).getSequenceAsString());
package jalview.analysis;
import static org.testng.AssertJUnit.assertEquals;
+import static org.testng.AssertJUnit.assertFalse;
+import static org.testng.AssertJUnit.assertNotNull;
+import static org.testng.AssertJUnit.assertNotSame;
+import static org.testng.AssertJUnit.assertNull;
import static org.testng.AssertJUnit.assertSame;
+import static org.testng.AssertJUnit.assertTrue;
+import static org.testng.internal.junit.ArrayAsserts.assertArrayEquals;
+import jalview.datamodel.Alignment;
+import jalview.datamodel.AlignmentI;
import jalview.datamodel.DBRefEntry;
+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.SequenceFetcherFactory;
+import java.util.ArrayList;
+import java.util.List;
+
+import org.testng.annotations.AfterClass;
import org.testng.annotations.Test;
public class CrossRefTest
DBRefEntry ref6 = new DBRefEntry("emblCDS", "1", "A123");
DBRefEntry ref7 = new DBRefEntry("GeneDB", "1", "A123");
DBRefEntry ref8 = new DBRefEntry("PFAM", "1", "A123");
+ // ENSEMBL is a source of either dna or protein sequence data
+ DBRefEntry ref9 = new DBRefEntry("ENSEMBL", "1", "A123");
DBRefEntry[] refs = new DBRefEntry[] { ref1, ref2, ref3, ref4, ref5,
- ref6, ref7, ref8 };
+ ref6, ref7, ref8, ref9 };
/*
* Just the DNA refs:
*/
- DBRefEntry[] found = CrossRef.findXDbRefs(false, refs);
- assertEquals(3, found.length);
+ DBRefEntry[] found = DBRefUtils.selectDbRefs(true, refs);
+ assertEquals(4, found.length);
assertSame(ref5, found[0]);
assertSame(ref6, found[1]);
assertSame(ref7, found[2]);
+ assertSame(ref9, found[3]);
/*
* Just the protein refs:
*/
- found = CrossRef.findXDbRefs(true, refs);
- assertEquals(4, found.length);
+ found = DBRefUtils.selectDbRefs(false, refs);
+ assertEquals(5, found.length);
assertSame(ref1, found[0]);
assertSame(ref2, found[1]);
assertSame(ref3, found[2]);
assertSame(ref4, found[3]);
+ assertSame(ref9, found[4]);
+ }
+
+ /**
+ * Test the method that finds a sequence's "product" xref source databases,
+ * which may be direct (dbrefs on the sequence), or indirect (dbrefs on
+ * sequences which share a dbref with the sequence
+ */
+ @Test(groups = { "Functional" }, enabled = false)
+ public void testFindXrefSourcesForSequence_proteinToDna()
+ {
+ SequenceI seq = new Sequence("Seq1", "MGKYQARLSS");
+ List<String> sources = new ArrayList<String>();
+ AlignmentI al = new Alignment(new SequenceI[] {});
+
+ /*
+ * first with no dbrefs to search
+ */
+ sources = new CrossRef(new SequenceI[] { seq }, al)
+ .findXrefSourcesForSequences();
+ assertTrue(sources.isEmpty());
+
+ /*
+ * add some dbrefs to sequence
+ */
+ // protein db is not a candidate for findXrefSources
+ seq.addDBRef(new DBRefEntry("UNIPROT", "0", "A1234"));
+ // dna coding databatases are
+ seq.addDBRef(new DBRefEntry("EMBL", "0", "E2345"));
+ // a second EMBL xref should not result in a duplicate
+ seq.addDBRef(new DBRefEntry("EMBL", "0", "E2346"));
+ seq.addDBRef(new DBRefEntry("EMBLCDS", "0", "E2347"));
+ seq.addDBRef(new DBRefEntry("GENEDB", "0", "E2348"));
+ seq.addDBRef(new DBRefEntry("ENSEMBL", "0", "E2349"));
+ seq.addDBRef(new DBRefEntry("ENSEMBLGENOMES", "0", "E2350"));
+ sources = new CrossRef(new SequenceI[] { seq }, al)
+ .findXrefSourcesForSequences();
+ assertEquals(4, sources.size());
+ assertEquals("[EMBL, EMBLCDS, GENEDB, ENSEMBL]",
+ sources.toString());
+
+ /*
+ * add a sequence to the alignment which has a dbref to UNIPROT|A1234
+ * and others to dna coding databases
+ */
+ sources.clear();
+ seq.setDBRefs(null);
+ seq.addDBRef(new DBRefEntry("UNIPROT", "0", "A1234"));
+ seq.addDBRef(new DBRefEntry("EMBLCDS", "0", "E2347"));
+ SequenceI seq2 = new Sequence("Seq2", "MGKYQARLSS");
+ seq2.addDBRef(new DBRefEntry("UNIPROT", "0", "A1234"));
+ seq2.addDBRef(new DBRefEntry("EMBL", "0", "E2345"));
+ seq2.addDBRef(new DBRefEntry("GENEDB", "0", "E2348"));
+ // TODO include ENSEMBLGENOMES in DBRefSource.DNACODINGDBS ?
+ al.addSequence(seq2);
+ sources = new CrossRef(new SequenceI[] { seq }, al)
+ .findXrefSourcesForSequences();
+ assertEquals(3, sources.size());
+ assertEquals("[EMBLCDS, EMBL, GENEDB]", sources.toString());
+ }
+
+ /**
+ * Test for finding 'product' sequences for the case where only an indirect
+ * xref is found - not on the nucleotide sequence but on a peptide sequence in
+ * the alignment which which it shares a nucleotide dbref
+ */
+ @Test(groups = { "Functional" }, enabled = false)
+ public void testFindXrefSequences_indirectDbrefToProtein()
+ {
+ /*
+ * Alignment setup:
+ * - nucleotide dbref EMBL|AF039662
+ * - peptide dbrefs EMBL|AF039662, UNIPROT|Q9ZTS2
+ */
+ SequenceI emblSeq = new Sequence("AF039662", "GGGGCAGCACAAGAAC");
+ emblSeq.addDBRef(new DBRefEntry("EMBL", "0", "AF039662"));
+ SequenceI uniprotSeq = new Sequence("Q9ZTS2", "MASVSATMISTS");
+ uniprotSeq.addDBRef(new DBRefEntry("EMBL", "0", "AF039662"));
+ uniprotSeq.addDBRef(new DBRefEntry("UNIPROT", "0", "Q9ZTS2"));
+
+ /*
+ * Find UNIPROT xrefs for nucleotide
+ * - it has no UNIPROT dbref of its own
+ * - but peptide with matching nucleotide dbref does, so is returned
+ */
+ AlignmentI al = new Alignment(new SequenceI[] { emblSeq, uniprotSeq });
+ Alignment xrefs = new CrossRef(new SequenceI[] { emblSeq }, al)
+ .findXrefSequences("UNIPROT");
+ assertEquals(1, xrefs.getHeight());
+ assertSame(uniprotSeq, xrefs.getSequenceAt(0));
+ }
+
+ /**
+ * Test for finding 'product' sequences for the case where only an indirect
+ * xref is found - not on the peptide sequence but on a nucleotide sequence in
+ * the alignment which which it shares a protein dbref
+ */
+ @Test(groups = { "Functional" }, enabled = false)
+ public void testFindXrefSequences_indirectDbrefToNucleotide()
+ {
+ /*
+ * Alignment setup:
+ * - peptide dbref UNIPROT|Q9ZTS2
+ * - nucleotide dbref EMBL|AF039662, UNIPROT|Q9ZTS2
+ */
+ SequenceI uniprotSeq = new Sequence("Q9ZTS2", "MASVSATMISTS");
+ uniprotSeq.addDBRef(new DBRefEntry("UNIPROT", "0", "Q9ZTS2"));
+ SequenceI emblSeq = new Sequence("AF039662", "GGGGCAGCACAAGAAC");
+ emblSeq.addDBRef(new DBRefEntry("EMBL", "0", "AF039662"));
+ emblSeq.addDBRef(new DBRefEntry("UNIPROT", "0", "Q9ZTS2"));
+
+ /*
+ * find EMBL xrefs for peptide sequence - it has no direct
+ * dbrefs, but the 'corresponding' nucleotide sequence does, so is returned
+ */
+ /*
+ * Find EMBL xrefs for peptide
+ * - it has no EMBL dbref of its own
+ * - but nucleotide with matching peptide dbref does, so is returned
+ */
+ AlignmentI al = new Alignment(new SequenceI[] { emblSeq, uniprotSeq });
+ Alignment xrefs = new CrossRef(new SequenceI[] { uniprotSeq },
+ al).findXrefSequences("EMBL");
+ assertEquals(1, xrefs.getHeight());
+ assertSame(emblSeq, xrefs.getSequenceAt(0));
+ }
+
+ /**
+ * Test for finding 'product' sequences for the case where the selected
+ * sequence has no dbref to the desired source, and there are no indirect
+ * references via another sequence in the alignment
+ */
+ @Test(groups = { "Functional" })
+ public void testFindXrefSequences_noDbrefs()
+ {
+ /*
+ * two nucleotide sequences, one with UNIPROT dbref
+ */
+ SequenceI dna1 = new Sequence("AF039662", "GGGGCAGCACAAGAAC");
+ dna1.addDBRef(new DBRefEntry("UNIPROT", "0", "Q9ZTS2"));
+ SequenceI dna2 = new Sequence("AJ307031", "AAACCCTTT");
+
+ /*
+ * find UNIPROT xrefs for peptide sequence - it has no direct
+ * dbrefs, and the other sequence (which has a UNIPROT dbref) is not
+ * equatable to it, so no results found
+ */
+ AlignmentI al = new Alignment(new SequenceI[] { dna1, dna2 });
+ Alignment xrefs = new CrossRef(new SequenceI[] { dna2 }, al)
+ .findXrefSequences("UNIPROT");
+ assertNull(xrefs);
+ }
+
+ /**
+ * Tests for the method that searches an alignment (with one sequence
+ * excluded) for protein/nucleotide sequences with a given cross-reference
+ */
+ @Test(groups = { "Functional" }, enabled = false)
+ public void testSearchDataset()
+ {
+ /*
+ * nucleotide sequence with UNIPROT AND EMBL dbref
+ * peptide sequence with UNIPROT dbref
+ */
+ SequenceI dna1 = new Sequence("AF039662", "GGGGCAGCACAAGAAC");
+ dna1.addDBRef(new DBRefEntry("UNIPROT", "0", "Q9ZTS2"));
+ dna1.addDBRef(new DBRefEntry("EMBL", "0", "AF039662"));
+ SequenceI pep1 = new Sequence("Q9ZTS2", "MLAVSRGQ");
+ pep1.addDBRef(new DBRefEntry("UNIPROT", "0", "Q9ZTS2"));
+ AlignmentI al = new Alignment(new SequenceI[] { dna1, pep1 });
+
+ List<SequenceI> result = new ArrayList<SequenceI>();
+
+ /*
+ * first search for a dbref nowhere on the alignment:
+ */
+ DBRefEntry dbref = new DBRefEntry("UNIPROT", "0", "P30419");
+ CrossRef testee = new CrossRef(al.getSequencesArray(), al);
+ boolean found = testee.searchDataset(dna1, dbref, result, null, true);
+ assertFalse(found);
+ assertTrue(result.isEmpty());
+
+ // TODO we are setting direct=true here but it is set to
+ // false in Jalview code...
+
+ /*
+ * search for a protein sequence with dbref UNIPROT:Q9ZTS2
+ */
+ dbref = new DBRefEntry("UNIPROT", "0", "Q9ZTS2");
+ found = testee.searchDataset(dna1, dbref, result, null, true);
+ assertTrue(found);
+ assertEquals(1, result.size());
+ assertSame(pep1, result.get(0));
+
+ /*
+ * search for a nucleotide sequence with dbref UNIPROT:Q9ZTS2
+ */
+ result.clear();
+ dbref = new DBRefEntry("UNIPROT", "0", "Q9ZTS2");
+ found = testee.searchDataset(pep1, dbref, result, null, false);
+ assertTrue(found);
+ assertEquals(1, result.size());
+ assertSame(dna1, result.get(0));
+ }
+
+ /**
+ * Test for finding 'product' sequences for the case where the selected
+ * sequence has a dbref with a mapping to a sequence
+ */
+ @Test(groups = { "Functional" })
+ public void testFindXrefSequences_fromDbRefMap()
+ {
+ /*
+ * two peptide sequences each with a DBRef and SequenceFeature
+ */
+ SequenceI pep1 = new Sequence("Q9ZTS2", "MALFQRSV");
+ pep1.addDBRef(new DBRefEntry("Pfam", "0", "PF00111"));
+ pep1.addSequenceFeature(new SequenceFeature("type", "desc", 12, 14, 1f,
+ "group"));
+ SequenceI pep2 = new Sequence("P30419", "MTRRSQIF");
+ pep2.addDBRef(new DBRefEntry("PDB", "0", "3JTK"));
+ pep2.addSequenceFeature(new SequenceFeature("type2", "desc2", 13, 15,
+ 12f, "group2"));
+
+ /*
+ * nucleotide sequence (to go in the alignment)
+ */
+ SequenceI dna1 = new Sequence("AF039662", "GGGGCAGCACAAGAAC");
+
+ /*
+ * add DBRefEntry's to dna1 with mappings from dna to both peptides
+ */
+ MapList mapList = new MapList(new int[] { 1, 24 }, new int[] { 1, 3 },
+ 3, 1);
+ Mapping map = new Mapping(pep1, mapList);
+ DBRefEntry dbRef1 = new DBRefEntry("UNIPROT", "0", "Q9ZTS2", map);
+ dna1.addDBRef(dbRef1);
+ mapList = new MapList(new int[] { 1, 24 }, new int[] { 1, 3 }, 3, 1);
+ map = new Mapping(pep2, mapList);
+ DBRefEntry dbRef2 = new DBRefEntry("UNIPROT", "0", "P30419", map);
+ dna1.addDBRef(dbRef2);
+
+ /*
+ * find UNIPROT xrefs for nucleotide sequence - it should pick up
+ * mapped sequences
+ */
+ AlignmentI al = new Alignment(new SequenceI[] { dna1 });
+ Alignment xrefs = new CrossRef(new SequenceI[] { dna1 }, al)
+ .findXrefSequences("UNIPROT");
+ assertEquals(2, xrefs.getHeight());
+
+ /*
+ * cross-refs alignment holds copies of the mapped sequences
+ * including copies of their dbrefs and features
+ */
+ checkCopySequence(pep1, xrefs.getSequenceAt(0));
+ checkCopySequence(pep2, xrefs.getSequenceAt(1));
+ }
+
+ /**
+ * Helper method to assert seq1 looks like a copy of seq2
+ *
+ * @param seq1
+ * @param seq2
+ */
+ private void checkCopySequence(SequenceI seq1, SequenceI seq2)
+ {
+ assertNotSame(seq1, seq2);
+ assertEquals(seq1.getName(), seq2.getName());
+ assertEquals(seq1.getStart(), seq2.getStart());
+ assertEquals(seq1.getEnd(), seq2.getEnd());
+ assertEquals(seq1.getSequenceAsString(), seq2.getSequenceAsString());
+
+ /*
+ * compare dbrefs
+ */
+ assertArrayEquals(seq1.getDBRefs(), seq2.getDBRefs());
+ // check one to verify a copy, not the same object
+ if (seq1.getDBRefs().length > 0)
+ {
+ assertNotSame(seq1.getDBRefs()[0], seq2.getDBRefs()[0]);
+ }
+
+ /*
+ * compare features
+ */
+ assertArrayEquals(seq1.getSequenceFeatures(),
+ seq2.getSequenceFeatures());
+ if (seq1.getSequenceFeatures().length > 0)
+ {
+ assertNotSame(seq1.getSequenceFeatures()[0],
+ seq2.getSequenceFeatures()[0]);
+ }
+ }
+
+ /**
+ * Test for finding 'product' sequences for the case where the selected
+ * sequence has a dbref with no mapping, triggering a fetch from database
+ */
+ @Test(groups = { "Functional" })
+ public void testFindXrefSequences_withFetch()
+ {
+ SequenceI dna1 = new Sequence("AF039662", "GGGGCAGCACAAGAAC");
+ dna1.addDBRef(new DBRefEntry("UNIPROT", "0", "Q9ZTS2"));
+ dna1.addDBRef(new DBRefEntry("UNIPROT", "0", "P30419"));
+ dna1.addDBRef(new DBRefEntry("UNIPROT", "0", "P00314"));
+ final SequenceI pep1 = new Sequence("Q9ZTS2", "MYQLIRSSW");
+ final SequenceI pep2 = new Sequence("P00314", "MRKLLAASG");
+
+ /*
+ * argument false suppresses adding DAS sources
+ * todo: define an interface type SequenceFetcherI and mock that
+ */
+ SequenceFetcher mockFetcher = new SequenceFetcher(false)
+ {
+ @Override
+ public boolean isFetchable(String source)
+ {
+ return true;
+ }
+
+ @Override
+ public SequenceI[] getSequences(List<DBRefEntry> refs, boolean dna)
+ {
+ return new SequenceI[] { pep1, pep2 };
+ }
+ };
+ SequenceFetcherFactory.setSequenceFetcher(mockFetcher);
+
+ /*
+ * find UNIPROT xrefs for nucleotide sequence
+ */
+ AlignmentI al = new Alignment(new SequenceI[] { dna1 });
+ Alignment xrefs = new CrossRef(new SequenceI[] { dna1 }, al)
+ .findXrefSequences("UNIPROT");
+ assertEquals(2, xrefs.getHeight());
+ assertSame(pep1, xrefs.getSequenceAt(0));
+ assertSame(pep2, xrefs.getSequenceAt(1));
+ }
+
+ @AfterClass
+ public void tearDown()
+ {
+ SequenceFetcherFactory.setSequenceFetcher(null);
}
+ /**
+ * Test for finding 'product' sequences for the case where both gene and
+ * transcript sequences have dbrefs to Uniprot.
+ */
+ @Test(groups = { "Functional" })
+ public void testFindXrefSequences_forGeneAndTranscripts()
+ {
+ /*
+ * 'gene' sequence
+ */
+ SequenceI gene = new Sequence("ENSG00000157764", "CGCCTCCCTTCCCC");
+ gene.addDBRef(new DBRefEntry("UNIPROT", "0", "P15056"));
+ gene.addDBRef(new DBRefEntry("UNIPROT", "0", "H7C5K3"));
+
+ /*
+ * 'transcript' with CDS feature (supports mapping to protein)
+ */
+ SequenceI braf001 = new Sequence("ENST00000288602", "taagATGGCGGCGCTGa");
+ braf001.addDBRef(new DBRefEntry("UNIPROT", "0", "P15056"));
+ braf001.addSequenceFeature(new SequenceFeature("CDS", "", 5, 16, 0f,
+ null));
+
+ /*
+ * 'spliced transcript' with CDS ranges
+ */
+ SequenceI braf002 = new Sequence("ENST00000497784", "gCAGGCtaTCTGTTCaa");
+ braf002.addDBRef(new DBRefEntry("UNIPROT", "0", "H7C5K3"));
+ braf002.addSequenceFeature(new SequenceFeature("CDS", "", 2, 6, 0f,
+ null));
+ braf002.addSequenceFeature(new SequenceFeature("CDS", "", 9, 15, 0f,
+ null));
+
+ /*
+ * TODO code is fragile - use of SequenceIdMatcher depends on fetched
+ * sequences having a name starting Source|Accession
+ * which happens to be true for Uniprot,PDB,EMBL but not Pfam,Rfam,Ensembl
+ */
+ final SequenceI pep1 = new Sequence("UNIPROT|P15056", "MAAL");
+ final SequenceI pep2 = new Sequence("UNIPROT|H7C5K3", "QALF");
+
+ /*
+ * argument false suppresses adding DAS sources
+ * todo: define an interface type SequenceFetcherI and mock that
+ */
+ SequenceFetcher mockFetcher = new SequenceFetcher(false)
+ {
+ @Override
+ public boolean isFetchable(String source)
+ {
+ return true;
+ }
+
+ @Override
+ public SequenceI[] getSequences(List<DBRefEntry> refs, boolean dna)
+ {
+ return new SequenceI[] { pep1, pep2 };
+ }
+ };
+ SequenceFetcherFactory.setSequenceFetcher(mockFetcher);
+
+ /*
+ * find UNIPROT xrefs for gene and transcripts
+ * verify that
+ * - the two proteins are retrieved but not duplicated
+ * - mappings are built from transcript (CDS) to proteins
+ * - no mappings from gene to proteins
+ */
+ SequenceI[] seqs = new SequenceI[] { gene, braf001, braf002 };
+ AlignmentI al = new Alignment(seqs);
+ Alignment xrefs = new CrossRef(seqs, al)
+ .findXrefSequences("UNIPROT");
+ assertEquals(2, xrefs.getHeight());
+ assertSame(pep1, xrefs.getSequenceAt(0));
+ assertSame(pep2, xrefs.getSequenceAt(1));
+ }
+
+ /**
+ * <pre>
+ * Test that emulates this (real but simplified) case:
+ * Alignment: DBrefs
+ * UNIPROT|P0CE19 EMBL|J03321, EMBL|X06707, EMBL|M19487
+ * UNIPROT|P0CE20 EMBL|J03321, EMBL|X06707, EMBL|X07547
+ * Find cross-references for EMBL. These are mocked here as
+ * EMBL|J03321 with mappings to P0CE18, P0CE19, P0CE20
+ * EMBL|X06707 with mappings to P0CE17, P0CE19, P0CE20
+ * EMBL|M19487 with mappings to P0CE19, Q46432
+ * EMBL|X07547 with mappings to P0CE20, B0BCM4
+ * EMBL sequences are first 'fetched' (mocked here) for P0CE19.
+ * The 3 EMBL sequences are added to the alignment dataset.
+ * Their dbrefs to Uniprot products P0CE19 and P0CE20 should be matched in the
+ * alignment dataset and updated to reference the original Uniprot sequences.
+ * For the second Uniprot sequence, the J03321 and X06707 xrefs should be
+ * resolved from the dataset, and only the X07547 dbref fetched.
+ * So the end state to verify is:
+ * - 4 cross-ref sequences returned: J03321, X06707, M19487, X07547
+ * - P0CE19/20 dbrefs to EMBL sequences now have mappings
+ * - J03321 dbrefs to P0CE19/20 mapped to original Uniprot sequences
+ * - X06707 dbrefs to P0CE19/20 mapped to original Uniprot sequences
+ * </pre>
+ */
+ @Test(groups = { "Functional" })
+ public void testFindXrefSequences_uniprotEmblManyToMany()
+ {
+ /*
+ * Uniprot sequences, both with xrefs to EMBL|J03321
+ * and EMBL|X07547
+ */
+ SequenceI p0ce19 = new Sequence("UNIPROT|P0CE19", "KPFG");
+ p0ce19.addDBRef(new DBRefEntry("EMBL", "0", "J03321"));
+ p0ce19.addDBRef(new DBRefEntry("EMBL", "0", "X06707"));
+ p0ce19.addDBRef(new DBRefEntry("EMBL", "0", "M19487"));
+ SequenceI p0ce20 = new Sequence("UNIPROT|P0CE20", "PFGK");
+ p0ce20.addDBRef(new DBRefEntry("EMBL", "0", "J03321"));
+ p0ce20.addDBRef(new DBRefEntry("EMBL", "0", "X06707"));
+ p0ce20.addDBRef(new DBRefEntry("EMBL", "0", "X07547"));
+
+ /*
+ * EMBL sequences to be 'fetched', complete with dbrefs and mappings
+ * to their protein products (CDS location and translations are provided
+ * in EMBL XML); these should be matched to, and replaced with,
+ * the corresponding uniprot sequences after fetching
+ */
+
+ /*
+ * J03321 with mappings to P0CE19 and P0CE20
+ */
+ final SequenceI j03321 = new Sequence("EMBL|J03321", "AAACCCTTTGGGAAAA");
+ DBRefEntry dbref1 = new DBRefEntry("UNIPROT", "0", "P0CE19");
+ MapList mapList = new MapList(new int[] { 1, 12 }, new int[] { 1, 4 },
+ 3, 1);
+ Mapping map = new Mapping(new Sequence("UNIPROT|P0CE19", "KPFG"), mapList);
+ // add a dbref to the mapped to sequence - should get copied to p0ce19
+ map.getTo().addDBRef(new DBRefEntry("PIR", "0", "S01875"));
+ dbref1.setMap(map);
+ j03321.addDBRef(dbref1);
+ DBRefEntry dbref2 = new DBRefEntry("UNIPROT", "0", "P0CE20");
+ mapList = new MapList(new int[] { 4, 15 }, new int[] { 2, 5 }, 3, 1);
+ dbref2.setMap(new Mapping(new Sequence("UNIPROT|P0CE20", "PFGK"),
+ new MapList(mapList)));
+ j03321.addDBRef(dbref2);
+
+ /*
+ * X06707 with mappings to P0CE19 and P0CE20
+ */
+ final SequenceI x06707 = new Sequence("EMBL|X06707", "atgAAACCCTTTGGG");
+ DBRefEntry dbref3 = new DBRefEntry("UNIPROT", "0", "P0CE19");
+ MapList map2 = new MapList(new int[] { 4, 15 }, new int[] { 1, 4 }, 3,
+ 1);
+ dbref3.setMap(new Mapping(new Sequence("UNIPROT|P0CE19", "KPFG"), map2));
+ x06707.addDBRef(dbref3);
+ DBRefEntry dbref4 = new DBRefEntry("UNIPROT", "0", "P0CE20");
+ MapList map3 = new MapList(new int[] { 4, 15 }, new int[] { 1, 4 }, 3,
+ 1);
+ dbref4.setMap(new Mapping(new Sequence("UNIPROT|P0CE20", "PFGK"), map3));
+ x06707.addDBRef(dbref4);
+
+ /*
+ * M19487 with mapping to P0CE19 and Q46432
+ */
+ final SequenceI m19487 = new Sequence("EMBL|M19487", "AAACCCTTTGGG");
+ DBRefEntry dbref5 = new DBRefEntry("UNIPROT", "0", "P0CE19");
+ dbref5.setMap(new Mapping(new Sequence("UNIPROT|P0CE19", "KPFG"),
+ new MapList(mapList)));
+ m19487.addDBRef(dbref5);
+ DBRefEntry dbref6 = new DBRefEntry("UNIPROT", "0", "Q46432");
+ dbref6.setMap(new Mapping(new Sequence("UNIPROT|Q46432", "KPFG"),
+ new MapList(mapList)));
+ m19487.addDBRef(dbref6);
+
+ /*
+ * X07547 with mapping to P0CE20 and B0BCM4
+ */
+ final SequenceI x07547 = new Sequence("EMBL|X07547", "cccAAACCCTTTGGG");
+ DBRefEntry dbref7 = new DBRefEntry("UNIPROT", "0", "P0CE20");
+ dbref7.setMap(new Mapping(new Sequence("UNIPROT|P0CE19", "KPFG"),
+ new MapList(map2)));
+ x07547.addDBRef(dbref7);
+ DBRefEntry dbref8 = new DBRefEntry("UNIPROT", "0", "B0BCM4");
+ dbref8.setMap(new Mapping(new Sequence("UNIPROT|B0BCM4", "KPFG"),
+ new MapList(map2)));
+ x07547.addDBRef(dbref8);
+
+ /*
+ * mock sequence fetcher to 'return' the EMBL sequences
+ * TODO: Mockito would allow .thenReturn().thenReturn() here,
+ * and also capture and verification of the parameters
+ * passed in calls to getSequences() - important to verify that
+ * duplicate sequence fetches are not requested
+ */
+ SequenceFetcher mockFetcher = new SequenceFetcher(false)
+ {
+ int call = 0;
+ @Override
+ public boolean isFetchable(String source)
+ {
+ return true;
+ }
+ @Override
+ public SequenceI[] getSequences(List<DBRefEntry> refs, boolean dna)
+ {
+ call++;
+ if (call == 1) {
+ assertEquals("Expected 3 embl seqs in first fetch", 3, refs.size());
+ return new SequenceI[] { j03321, x06707, m19487 };
+ } else {
+ assertEquals("Expected 1 embl seq in second fetch", 1, refs.size());
+ return new SequenceI[] { x07547 };
+ }
+ }
+ };
+ SequenceFetcherFactory.setSequenceFetcher(mockFetcher);
+
+ /*
+ * find EMBL xrefs for Uniprot seqs and verify that
+ * - the EMBL xref'd sequences are retrieved without duplicates
+ * - mappings are added to the Uniprot dbrefs
+ * - mappings in the EMBL-to-Uniprot dbrefs are updated to the
+ * alignment sequences
+ * - dbrefs on the EMBL sequences are added to the original dbrefs
+ */
+ SequenceI[] seqs = new SequenceI[] { p0ce19, p0ce20 };
+ AlignmentI al = new Alignment(seqs);
+ Alignment xrefs = new CrossRef(seqs, al)
+ .findXrefSequences("EMBL");
+
+ /*
+ * verify retrieved sequences
+ */
+ assertNotNull(xrefs);
+ assertEquals(4, xrefs.getHeight());
+ assertSame(j03321, xrefs.getSequenceAt(0));
+ assertSame(x06707, xrefs.getSequenceAt(1));
+ assertSame(m19487, xrefs.getSequenceAt(2));
+ assertSame(x07547, xrefs.getSequenceAt(3));
+
+ /*
+ * verify mappings added to Uniprot-to-EMBL dbrefs
+ */
+ Mapping mapping = p0ce19.getDBRefs()[0].getMap();
+ assertSame(j03321, mapping.getTo());
+ mapping = p0ce19.getDBRefs()[1].getMap();
+ assertSame(x06707, mapping.getTo());
+ mapping = p0ce20.getDBRefs()[0].getMap();
+ assertSame(j03321, mapping.getTo());
+ mapping = p0ce20.getDBRefs()[1].getMap();
+ assertSame(x06707, mapping.getTo());
+
+ /*
+ * verify dbrefs on EMBL are mapped to alignment seqs
+ */
+ assertSame(p0ce19, j03321.getDBRefs()[0].getMap().getTo());
+ assertSame(p0ce20, j03321.getDBRefs()[1].getMap().getTo());
+ assertSame(p0ce19, x06707.getDBRefs()[0].getMap().getTo());
+ assertSame(p0ce20, x06707.getDBRefs()[1].getMap().getTo());
+
+ /*
+ * verify new dbref on EMBL dbref mapping is copied to the
+ * original Uniprot sequence
+ */
+ assertEquals(4, p0ce19.getDBRefs().length);
+ assertEquals("PIR", p0ce19.getDBRefs()[3].getSource());
+ assertEquals("S01875", p0ce19.getDBRefs()[3].getAccessionId());
+ }
+
+ @Test(groups = "Functional")
+ public void testSameSequence()
+ {
+ assertTrue(CrossRef.sameSequence(null, null));
+ SequenceI seq1 = new Sequence("seq1", "ABCDEF");
+ assertFalse(CrossRef.sameSequence(seq1, null));
+ assertFalse(CrossRef.sameSequence(null, seq1));
+ assertTrue(CrossRef.sameSequence(seq1, new Sequence("seq2", "ABCDEF")));
+ assertTrue(CrossRef.sameSequence(seq1, new Sequence("seq2", "abcdef")));
+ assertFalse(CrossRef
+ .sameSequence(seq1, new Sequence("seq2", "ABCDE-F")));
+ assertFalse(CrossRef.sameSequence(seq1, new Sequence("seq2", "BCDEF")));
+ }
}
* case insensitive matching
*/
assertTrue(testee.equals("a12345"));
+
+ testee = sequenceIdMatcher.new SeqIdName("UNIPROT|A12345");
+ assertFalse(testee.equals("A12345"));
+ assertFalse(testee.equals("UNIPROT|B98765"));
+ assertFalse(testee.equals("UNIPROT|"));
+ assertTrue(testee.equals("UNIPROT"));
}
}
assertArrayEquals(new int[] { 2, 2 },
acf.getMappedRegion(seq2, seq1, 6));
}
+
+ /**
+ * Tests for addMap. See also tests for MapList.addMapList
+ */
+ @Test(groups = { "Functional" })
+ public void testAddMap()
+ {
+ final Sequence seq1 = new Sequence("Seq1", "c-G-TA-gC-gT-T");
+ seq1.createDatasetSequence();
+ final Sequence aseq1 = new Sequence("Seq1", "-V-L");
+ aseq1.createDatasetSequence();
+
+ AlignedCodonFrame acf = new AlignedCodonFrame();
+ MapList map = new MapList(new int[] { 2, 4, 6, 6, 8, 9 }, new int[] {
+ 1, 2 }, 3, 1);
+ acf.addMap(seq1.getDatasetSequence(), aseq1.getDatasetSequence(), map);
+ assertEquals(1, acf.getMappingsFromSequence(seq1).size());
+ Mapping before = acf.getMappingsFromSequence(seq1).get(0);
+
+ /*
+ * add the same map again, verify it doesn't get duplicated
+ */
+ acf.addMap(seq1.getDatasetSequence(), aseq1.getDatasetSequence(), map);
+ assertEquals(1, acf.getMappingsFromSequence(seq1).size());
+ assertSame(before, acf.getMappingsFromSequence(seq1).get(0));
+ }
}
import jalview.datamodel.Sequence;
import jalview.datamodel.SequenceI;
+import java.util.List;
+
import org.testng.annotations.Test;
public class DBRefUtilsTest
ref5.setMap(new Mapping(new MapList(new int[] { 1, 1 }, new int[] { 1,
1 }, 1, 1)));
- DBRefEntry[] matches = DBRefUtils.searchRefs(new DBRefEntry[] { ref1,
+ List<DBRefEntry> matches = DBRefUtils.searchRefs(new DBRefEntry[] {
+ ref1,
ref2, ref3, ref4, ref5 }, target);
- assertEquals(3, matches.length);
- assertSame(ref1, matches[0]);
- assertSame(ref2, matches[1]);
- assertSame(ref5, matches[2]);
+ assertEquals(3, matches.size());
+ assertSame(ref1, matches.get(0));
+ assertSame(ref2, matches.get(1));
+ assertSame(ref5, matches.get(2));
}
/**
new int[] { 1, 1 }, 2, 2));
ref3.setMap(map3);
- DBRefEntry[] matches = DBRefUtils.searchRefs(new DBRefEntry[] { ref1,
+ List<DBRefEntry> matches = DBRefUtils.searchRefs(new DBRefEntry[] {
+ ref1,
ref2, ref3 }, target);
- assertEquals(2, matches.length);
- assertSame(ref1, matches[0]);
- assertSame(ref2, matches[1]);
+ assertEquals(2, matches.size());
+ assertSame(ref1, matches.get(0));
+ assertSame(ref2, matches.get(1));
}
/**
ref5.setMap(new Mapping(new MapList(new int[] { 1, 1 }, new int[] { 1,
1 }, 1, 1)));
- DBRefEntry[] matches = DBRefUtils.searchRefs(new DBRefEntry[] { ref1,
- ref2, ref3, ref4, ref5 }, "A1234");
- assertEquals(3, matches.length);
- assertSame(ref1, matches[0]);
- assertSame(ref2, matches[1]);
- assertSame(ref5, matches[2]);
+ DBRefEntry[] dbrefs = new DBRefEntry[] { ref1,
+ ref2, ref3, ref4, ref5 };
+ List<DBRefEntry> matches = DBRefUtils.searchRefs(dbrefs, "A1234");
+ assertEquals(3, matches.size());
+ assertSame(ref1, matches.get(0));
+ assertSame(ref2, matches.get(1));
+ assertSame(ref5, matches.get(2));
+ }
+
+ /**
+ * Test the method that searches for matches references - case when we are
+ * matching a reference with null (any) accession id
+ */
+ @Test(groups = { "Functional" })
+ public void testSearchRefs_wildcardAccessionid()
+ {
+ DBRefEntry target = new DBRefEntry("EMBL", "2", null);
+
+ DBRefEntry ref1 = new DBRefEntry("EMBL", "1", "A1234"); // matches
+ // constructor changes embl to EMBL
+ DBRefEntry ref2 = new DBRefEntry("embl", "1", "A1235"); // matches
+ // constructor does not upper-case accession id
+ DBRefEntry ref3 = new DBRefEntry("EMBL", "1", "A1236"); // matches
+ DBRefEntry ref4 = new DBRefEntry("EMBLCDS", "1", "A1234"); // no match
+ // ref5 matches although it has a mapping - ignored
+ DBRefEntry ref5 = new DBRefEntry("EMBL", "1", "A1237");
+ ref5.setMap(new Mapping(new MapList(new int[] { 1, 1 }, new int[] { 1,
+ 1 }, 1, 1)));
+
+ List<DBRefEntry> matches = DBRefUtils.searchRefs(new DBRefEntry[] {
+ ref1,
+ ref2, ref3, ref4, ref5 }, target);
+ assertEquals(4, matches.size());
+ assertSame(ref1, matches.get(0));
+ assertSame(ref2, matches.get(1));
+ assertSame(ref3, matches.get(2));
+ assertSame(ref5, matches.get(3));
}
}
assertEquals(87064, ranges.get(1)[1]);
/*
- * beyond 5' or 3' locus
- */
- ranges = DnaUtils.parseLocation("<34..126");
- assertEquals(1, ranges.size());
- assertEquals(34, ranges.get(0)[0]);
- assertEquals(126, ranges.get(0)[1]);
- ranges = DnaUtils.parseLocation("35..>127");
- assertEquals(1, ranges.size());
- assertEquals(35, ranges.get(0)[0]);
- assertEquals(127, ranges.get(0)[1]);
-
- /*
* valid things we don't yet handle
*/
+ assertNull(DnaUtils.parseLocation("<34..126"));
+ assertNull(DnaUtils.parseLocation("35..>126"));
assertNull(DnaUtils.parseLocation("34.126"));
assertNull(DnaUtils.parseLocation("34^126"));
assertNull(DnaUtils.parseLocation("order(34..126,130..180)"));
s);
}
+ /**
+ * Test that confirms adding a map twice does nothing
+ */
+ @Test(groups = { "Functional" })
+ public void testAddMapList_sameMap()
+ {
+ MapList ml = new MapList(new int[] { 11, 15, 20, 25, 35, 30 },
+ new int[] { 72, 22 }, 1, 3);
+ String before = ml.toString();
+ ml.addMapList(ml);
+ assertEquals(before, ml.toString());
+ ml.addMapList(new MapList(ml));
+ assertEquals(before, ml.toString());
+ }
+
@Test(groups = { "Functional" })
public void testAddMapList_contiguous()
{
package jalview.ws;
+import jalview.analysis.CrossRef;
import jalview.datamodel.Alignment;
import jalview.datamodel.AlignmentI;
import jalview.datamodel.DBRefSource;
{
boolean dna = sp.isDnaCoding();
// try and find products
- String types[] = jalview.analysis.CrossRef
- .findSequenceXrefTypes(dna, al.getSequencesArray());
+ CrossRef crossRef = new CrossRef(al.getSequencesArray(),
+ al);
+ List<String> types = crossRef.findXrefSourcesForSequences();
if (types != null)
{
System.out.println("Xref Types for: "
+ (dna ? "dna" : "prot"));
- for (int t = 0; t < types.length; t++)
+ for (String source : types)
{
- System.out.println("Type: " + types[t]);
- SequenceI[] prod = jalview.analysis.CrossRef
- .findXrefSequences(al.getSequencesArray(), dna,
- types[t], null)
+ System.out.println("Type: " + source);
+ SequenceI[] prod = crossRef.findXrefSequences(source)
.getSequencesArray();
System.out.println("Found "
+ ((prod == null) ? "no" : "" + prod.length)
// have a bash at finding the products amongst all the retrieved
// sequences.
SequenceI[] seqs = al.getSequencesArray();
- Alignment prodal = jalview.analysis.CrossRef.findXrefSequences(
- seqs, dna, null, ds);
+ Alignment prodal = new CrossRef(seqs, ds)
+ .findXrefSequences(null);
System.out.println("Found "
+ ((prodal == null) ? "no" : "" + prodal.getHeight())
+ " products");
.getMap().getMappedWidth(), 1);
assertEquals("Expected local reference map to be 3 nucleotides", dr[0]
.getMap().getWidth(), 3);
- AlignmentI sprods = CrossRef.findXrefSequences(
- alsq.getSequencesArray(), true, dr[0].getSource(), alsq);
+ AlignmentI sprods = new CrossRef(alsq.getSequencesArray(), alsq)
+ .findXrefSequences(dr[0].getSource());
assertNotNull(
"Couldn't recover cross reference sequence from dataset. Was it ever added ?",
sprods);