package jalview.ext.ensembl;
import jalview.analysis.AlignmentUtils;
+import jalview.analysis.Dna;
import jalview.datamodel.Alignment;
import jalview.datamodel.AlignmentI;
import jalview.datamodel.DBRefEntry;
-import jalview.datamodel.DBRefSource;
import jalview.datamodel.Mapping;
import jalview.datamodel.SequenceFeature;
import jalview.datamodel.SequenceI;
import jalview.io.FileParse;
import jalview.io.gff.SequenceOntologyFactory;
import jalview.io.gff.SequenceOntologyI;
+import jalview.util.Comparison;
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",
- "Uniprot/SPTREMBL" });
-
- protected static final String CONSEQUENCE_TYPE = "consequence_type";
+ private static final String ALLELES = "alleles";
protected static final String PARENT = "Parent";
GENOMIC("genomic"),
/**
- * type=cdna to fetch dna including UTRs
+ * type=cdna to fetch coding dna including UTRs
*/
CDNA("cdna"),
{
// clunky: ensure Uniprot xref if we have one is on mapped sequence
SequenceI ds = proteinSeq.getDatasetSequence();
- ds.setSourceDBRef(proteinSeq.getSourceDBRef());
+ // TODO: Verify ensp primary ref is on proteinSeq.getDatasetSequence()
Mapping map = new Mapping(ds, mapList);
- DBRefEntry dbr = new DBRefEntry(getDbSource(), getDbVersion(),
- accId, map);
+ DBRefEntry dbr = new DBRefEntry(getDbSource(),
+ getEnsemblDataVersion(), proteinSeq.getName(), map);
querySeq.getDatasetSequence().addDBRef(dbr);
/*
seq = seq.getDatasetSequence();
}
- EnsemblXref xrefFetcher = new EnsemblXref(getDomain());
- List<DBRefEntry> xrefs = xrefFetcher.getCrossReferences(seq.getName(),
- getCrossReferenceDatabases());
+ EnsemblXref xrefFetcher = new EnsemblXref(getDomain(), getDbSource(),
+ getEnsemblDataVersion());
+ List<DBRefEntry> xrefs = xrefFetcher.getCrossReferences(seq.getName());
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.
- * Specifically, the names are used to filter data returned by the Ensembl
- * xrefs REST service on the value in field 'dbname'.
- *
- * @return
- */
- protected List<String> getCrossReferenceDatabases()
- {
- return CROSS_REFERENCES;
+ /*
+ * and add a reference to itself
+ */
+ DBRefEntry self = new DBRefEntry(getDbSource(),
+ getEnsemblDataVersion(), seq.getName());
+ seq.addDBRef(self);
}
/**
if (ids.contains(name)
|| ids.contains(name.replace("ENSP", "ENST")))
{
- DBRefUtils.parseToDbRef(sq, DBRefSource.ENSEMBL, "0", name);
+ DBRefEntry dbref = DBRefUtils.parseToDbRef(sq, getDbSource(),
+ getEnsemblDataVersion(), name);
+ sq.addDBRef(dbref);
}
}
if (alignment == null)
* @param mapping
* mapping from the sequence feature's coordinates to the target
* sequence
+ * @param forwardStrand
*/
protected void transferFeature(SequenceFeature sf,
- SequenceI targetSequence, MapList mapping)
+ SequenceI targetSequence, MapList mapping, boolean forwardStrand)
{
int start = sf.getBegin();
int end = sf.getEnd();
SequenceFeature copy = new SequenceFeature(sf);
copy.setBegin(Math.min(mappedRange[0], mappedRange[1]));
copy.setEnd(Math.max(mappedRange[0], mappedRange[1]));
+ if (".".equals(copy.getFeatureGroup()))
+ {
+ copy.setFeatureGroup(getDbSource());
+ }
targetSequence.addSequenceFeature(copy);
/*
- * for sequence_variant, make an additional feature with consequence
+ * for sequence_variant on reverse strand, have to convert the allele
+ * values to their complements
*/
- // 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);
- // }
- // }
+ if (!forwardStrand
+ && SequenceOntologyFactory.getInstance().isA(sf.getType(),
+ SequenceOntologyI.SEQUENCE_VARIANT))
+ {
+ reverseComplementAlleles(copy);
+ }
+ }
+ }
+
+ /**
+ * Change the 'alleles' value of a feature by converting to complementary
+ * bases, and also update the feature description to match
+ *
+ * @param sf
+ */
+ static void reverseComplementAlleles(SequenceFeature sf)
+ {
+ final String alleles = (String) sf.getValue(ALLELES);
+ if (alleles == null)
+ {
+ return;
+ }
+ StringBuilder complement = new StringBuilder(alleles.length());
+ for (String allele : alleles.split(","))
+ {
+ reverseComplementAllele(complement, allele);
+ }
+ String comp = complement.toString();
+ sf.setValue(ALLELES, comp);
+ sf.setDescription(comp);
+
+ /*
+ * replace value of "alleles=" in sf.ATTRIBUTES as well
+ * so 'output as GFF' shows reverse complement alleles
+ */
+ String atts = sf.getAttributes();
+ if (atts != null)
+ {
+ atts = atts.replace(ALLELES + "=" + alleles, ALLELES + "=" + comp);
+ sf.setAttributes(atts);
+ }
+ }
+
+ /**
+ * Makes the 'reverse complement' of the given allele and appends it to the
+ * buffer, after a comma separator if not the first
+ *
+ * @param complement
+ * @param allele
+ */
+ static void reverseComplementAllele(StringBuilder complement,
+ String allele)
+ {
+ if (complement.length() > 0)
+ {
+ complement.append(",");
+ }
+
+ /*
+ * some 'alleles' are actually descriptive terms
+ * e.g. HGMD_MUTATION, PhenCode_variation
+ * - we don't want to 'reverse complement' these
+ */
+ if (!Comparison.isNucleotideSequence(allele, true))
+ {
+ complement.append(allele);
+ }
+ else
+ {
+ for (int i = allele.length() - 1; i >= 0; i--)
+ {
+ complement.append(Dna.getComplement(allele.charAt(i)));
+ }
}
}
final boolean forwardStrand = mapping.isFromForwardStrand();
/*
- * sort features by start position (descending if reverse strand)
- * before transferring (in forwards order) to the target sequence
+ * sort features by start position (which corresponds to end
+ * position descending if reverse strand) so as to add them in
+ * 'forwards' order to the target sequence
*/
- Arrays.sort(features, new Comparator<SequenceFeature>()
- {
- @Override
- public int compare(SequenceFeature o1, SequenceFeature o2)
- {
- int c = Integer.compare(o1.getBegin(), o2.getBegin());
- return forwardStrand ? c : -c;
- }
- });
+ sortFeatures(features, forwardStrand);
boolean transferred = false;
for (SequenceFeature sf : features)
{
if (retainFeature(sf, parentId))
{
- transferFeature(sf, targetSequence, mapping);
+ transferFeature(sf, targetSequence, mapping, forwardStrand);
transferred = true;
}
}
}
/**
+ * Sort features by start position ascending (if on forward strand), or end
+ * position descending (if on reverse strand)
+ *
+ * @param features
+ * @param forwardStrand
+ */
+ protected static void sortFeatures(SequenceFeature[] features,
+ final boolean forwardStrand)
+ {
+ Arrays.sort(features, new Comparator<SequenceFeature>()
+ {
+ @Override
+ public int compare(SequenceFeature o1, SequenceFeature o2)
+ {
+ if (forwardStrand)
+ {
+ return Integer.compare(o1.getBegin(), o2.getBegin());
+ }
+ else
+ {
+ return Integer.compare(o2.getEnd(), o1.getEnd());
+ }
+ }
+ });
+ }
+
+ /**
* Answers true if the feature type is one we want to keep for the sequence.
* Some features are only retrieved in order to identify the sequence range,
* and may then be discarded as redundant information (e.g. "CDS" feature for