import jalview.exceptions.JalviewException;
import jalview.io.FastaFile;
import jalview.io.FileParse;
-import jalview.io.gff.SequenceOntology;
+import jalview.io.gff.SequenceOntologyFactory;
+import jalview.io.gff.SequenceOntologyI;
import jalview.schemes.ResidueProperties;
import jalview.util.DBRefUtils;
import jalview.util.MapList;
*/
public abstract class EnsemblSeqProxy extends EnsemblRestClient
{
+ private static final List<String> CROSS_REFERENCES = Arrays
+ .asList(new String[] { "CCDS", "Uniprot/SWISSPROT" });
+
protected static final String CONSEQUENCE_TYPE = "consequence_type";
protected static final String PARENT = "Parent";
protected static final String ID = "ID";
- /*
- * this needs special handling, as it isA sequence_variant in the
- * Sequence Ontology, but behaves in Ensembl as if it isA transcript
- */
- protected static final String NMD_VARIANT = "NMD_transcript_variant";
-
protected static final String NAME = "Name";
+ /*
+ * enum for 'type' parameter to the /sequence REST service
+ */
public enum EnsemblSeqType
{
/**
- * type=genomic for the full dna including introns
+ * type=genomic to fetch full dna including introns
*/
GENOMIC("genomic"),
/**
- * type=cdna for transcribed dna including UTRs
+ * type=cdna to fetch dna including UTRs
*/
CDNA("cdna"),
/**
- * type=cds for coding dna excluding UTRs
+ * type=cds to fetch coding dna excluding UTRs
*/
CDS("cds"),
/**
- * type=protein for the peptide product sequence
+ * type=protein to fetch peptide product sequence
*/
PROTEIN("protein");
@Override
public AlignmentI getSequenceRecords(String query) throws Exception
{
- long now = System.currentTimeMillis();
// TODO use a String... query vararg instead?
// danger: accession separator used as a regex here, a string elsewhere
+ " chunks. Unexpected problem (" + r.getLocalizedMessage()
+ ")";
System.err.println(msg);
- if (alignment != null)
- {
- break; // return what we got
- }
- else
- {
- throw new JalviewException(msg, r);
- }
+ break;
}
}
+ if (alignment == null)
+ {
+ return null;
+ }
+
/*
* fetch and transfer genomic sequence features,
* fetch protein product and add as cross-reference
addFeaturesAndProduct(accId, alignment);
}
- inProgress = false;
- System.out.println(getClass().getName() + " took "
- + (System.currentTimeMillis() - now) + "ms to fetch");
+ for (SequenceI seq : alignment.getSequences())
+ {
+ getCrossReferences(seq);
+ }
+
return alignment;
}
* get 'dummy' genomic sequence with exon, cds and variation features
*/
SequenceI genomicSequence = null;
- EnsemblOverlap gffFetcher = new EnsemblOverlap();
+ EnsemblFeatures gffFetcher = new EnsemblFeatures();
EnsemblFeatureType[] features = getFeaturesToFetch();
AlignmentI geneFeatures = gffFetcher.getSequenceRecords(accId,
features);
MapList mapList = mapCdsToProtein(querySeq, proteinSeq);
if (mapList != null)
{
- Mapping map = new Mapping(proteinSeq.getDatasetSequence(), mapList);
+ // clunky: ensure Uniprot xref if we have one is on mapped sequence
+ SequenceI ds = proteinSeq.getDatasetSequence();
+ ds.setSourceDBRef(proteinSeq.getSourceDBRef());
+ Mapping map = new Mapping(ds, mapList);
DBRefEntry dbr = new DBRefEntry(getDbSource(), getDbVersion(),
accId, map);
querySeq.getDatasetSequence().addDBRef(dbr);
}
/**
+ * Get database xrefs from Ensembl, and attach them to the sequence
+ *
+ * @param seq
+ */
+ protected void getCrossReferences(SequenceI seq)
+ {
+ while (seq.getDatasetSequence() != null)
+ {
+ seq = seq.getDatasetSequence();
+ }
+
+ EnsemblXref xrefFetcher = new EnsemblXref();
+ List<DBRefEntry> xrefs = xrefFetcher.getCrossReferences(seq.getName(),
+ getCrossReferenceDatabases());
+ for (DBRefEntry xref : xrefs)
+ {
+ seq.addDBRef(xref);
+ /*
+ * Save any Uniprot xref to be the reference for SIFTS mapping
+ */
+ if (DBRefSource.UNIPROT.equals(xref.getSource()))
+ {
+ seq.setSourceDBRef(xref);
+ }
+ }
+ }
+
+ /**
+ * Returns a list of database names to be used when fetching cross-references.
+ *
+ * @return
+ */
+ protected List<String> getCrossReferenceDatabases()
+ {
+ return CROSS_REFERENCES;
+ }
+
+ /**
* Returns a mapping from dna to protein by inspecting sequence features of
* type "CDS" on the dna.
*
int mappedDnaLength = getCdsRanges(dnaSeq, ranges);
int proteinLength = proteinSeq.getLength();
- List<int[]> proteinRange = new ArrayList<int[]>();
+ int proteinEnd = proteinLength;
int proteinStart = 1;
/*
proteinStart = 2;
proteinLength--;
}
- proteinRange.add(new int[] { proteinStart, proteinLength });
+ List<int[]> proteinRange = new ArrayList<int[]>();
/*
* dna length should map to protein (or protein plus stop codon)
*/
int codesForResidues = mappedDnaLength / 3;
- if (codesForResidues == proteinLength
- || codesForResidues == (proteinLength + 1))
+ if (codesForResidues == (proteinLength + 1))
{
+ MappingUtils.unmapStopCodon(ranges, mappedDnaLength);
+ codesForResidues--;
+ }
+ if (codesForResidues == proteinLength)
+ {
+ proteinRange.add(new int[] { proteinStart, proteinEnd });
return new MapList(ranges, proteinRange, 3, 1);
}
return null;
}
/**
- * Adds CDS ranges to the ranges list, and returns the total length mapped.
+ * Adds CDS ranges to the ranges list, and returns the total length mapped
+ * from.
*
- * No need to worry about reverse strand dna here since the retrieved sequence
- * is as transcribed (reverse complement for reverse strand), i.e in the same
- * sense as the peptide.
+ * No need to worry about reverse strand dna, here since the retrieved
+ * sequence is as transcribed (reverse complement for reverse strand), i.e in
+ * the same sense as the peptide.
*
* @param dnaSeq
* @param ranges
{
return 0;
}
+ SequenceOntologyI so = SequenceOntologyFactory.getInstance();
int mappedDnaLength = 0;
for (SequenceFeature sf : sfs)
{
/*
* process a CDS feature (or a sub-type of CDS)
*/
- if (SequenceOntology.getInstance().isA(sf.getType(), SequenceOntology.CDS))
+ if (so.isA(sf.getType(), SequenceOntologyI.CDS))
{
int phase = 0;
try {
*/
int begin = sf.getBegin();
int end = sf.getEnd();
- if (ranges.isEmpty() && phase > 0)
+ if (ranges.isEmpty())
{
begin += phase;
if (begin > end)
* the start position of the sequence we are mapping to
* @return
*/
- protected MapList getGenomicRanges(SequenceI sourceSequence,
+ protected MapList getGenomicRangesFromFeatures(SequenceI sourceSequence,
String accId, int start)
{
SequenceFeature[] sfs = sourceSequence.getSequenceFeatures();
*/
if (identifiesSequence(sf, accId))
{
- int strand = sf.getStrand();
-
- if (directionSet && strand != direction)
- {
- // abort - mix of forward and backward
+ int strand = sf.getStrand();
+ strand = strand == 0 ? 1 : strand; // treat unknown as forward
+
+ if (directionSet && strand != direction)
+ {
+ // abort - mix of forward and backward
System.err.println("Error: forward and backward strand for "
+ accId);
return null;
*/
Collections.sort(regions, new RangeSorter(direction == 1));
- List<int[]> to = new ArrayList<int[]>();
- to.add(new int[] { start, start + mappedLength - 1 });
+ List<int[]> to = Arrays.asList(new int[] { start,
+ start + mappedLength - 1 });
return new MapList(regions, to, 1, 1);
}
/*
* for sequence_variant, make an additional feature with consequence
*/
- if (SequenceOntology.getInstance().isSequenceVariant(sf.getType()))
- {
- String consequence = (String) sf.getValue(CONSEQUENCE_TYPE);
- if (consequence != null)
- {
- SequenceFeature sf2 = new SequenceFeature("consequence",
- consequence, copy.getBegin(), copy.getEnd(), 0f,
- null);
- targetSequence.addSequenceFeature(sf2);
- }
- }
+ // if (SequenceOntologyFactory.getInstance().isA(sf.getType(),
+ // SequenceOntologyI.SEQUENCE_VARIANT))
+ // {
+ // String consequence = (String) sf.getValue(CONSEQUENCE_TYPE);
+ // if (consequence != null)
+ // {
+ // SequenceFeature sf2 = new SequenceFeature("consequence",
+ // consequence, copy.getBegin(), copy.getEnd(), 0f,
+ // null);
+ // targetSequence.addSequenceFeature(sf2);
+ // }
+ // }
}
}
return false;
}
+ // long start = System.currentTimeMillis();
SequenceFeature[] sfs = sourceSequence.getSequenceFeatures();
- MapList mapping = getGenomicRanges(sourceSequence, accessionId,
+ MapList mapping = getGenomicRangesFromFeatures(sourceSequence, accessionId,
targetSequence.getStart());
if (mapping == null)
{
return false;
}
- return transferFeatures(sfs, targetSequence, mapping, accessionId);
+ boolean result = transferFeatures(sfs, targetSequence, mapping,
+ accessionId);
+ // System.out.println("transferFeatures (" + (sfs.length) + " --> "
+ // + targetSequence.getSequenceFeatures().length + ") to "
+ // + targetSequence.getName()
+ // + " took " + (System.currentTimeMillis() - start) + "ms");
+ return result;
}
/**
SequenceFeature[] sfs = sequence.getSequenceFeatures();
if (sfs != null) {
- SequenceOntology so = SequenceOntology.getInstance();
+ SequenceOntologyI so = SequenceOntologyFactory.getInstance();
for (SequenceFeature sf :sfs) {
if (so.isA(sf.getType(), type))
{
}
AlignmentUtils.transferFeatures(dnaSeq, peptide, dnaToProtein,
- SequenceOntology.EXON);
+ SequenceOntologyI.EXON);
LinkedHashMap<Integer, String[][]> variants = buildDnaVariantsMap(
dnaSeq, dnaToProtein);
String desc = StringUtils.listToDelimitedString(peptideVariants,
", ");
SequenceFeature sf = new SequenceFeature(
- SequenceOntology.SEQUENCE_VARIANT, desc, peptidePos,
+ SequenceOntologyI.SEQUENCE_VARIANT, desc, peptidePos,
peptidePos, 0f, null);
peptide.addSequenceFeature(sf);
count++;
}
}
+
+ /*
+ * ugly sort to get sequence features in start position order
+ * - would be better to store in Sequence as a TreeSet instead?
+ */
+ Arrays.sort(peptide.getSequenceFeatures(),
+ new Comparator<SequenceFeature>()
+ {
+ @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;
}
* LinkedHashMap ensures we add the peptide features in sequence order
*/
LinkedHashMap<Integer, String[][]> variants = new LinkedHashMap<Integer, String[][]>();
- SequenceOntology so = SequenceOntology.getInstance();
+ SequenceOntologyI so = SequenceOntologyFactory.getInstance();
SequenceFeature[] dnaFeatures = dnaSeq.getSequenceFeatures();
if (dnaFeatures == null)
// not handling multi-locus variant features
continue;
}
- if (so.isSequenceVariant(sf.getType()))
+ if (so.isA(sf.getType(), SequenceOntologyI.SEQUENCE_VARIANT))
{
int[] mapsTo = dnaToProtein.locateInTo(dnaCol, dnaCol);
if (mapsTo == null)
*/
public static boolean isTranscript(String featureType)
{
- return NMD_VARIANT.equals(featureType)
- || SequenceOntology.getInstance().isA(featureType, SequenceOntology.TRANSCRIPT);
+ return SequenceOntologyI.NMD_TRANSCRIPT_VARIANT.equals(featureType)
+ || SequenceOntologyFactory.getInstance().isA(featureType,
+ SequenceOntologyI.TRANSCRIPT);
}
}