import jalview.api.FeatureColourI;
import jalview.api.FeatureSettingsModelI;
import jalview.datamodel.AlignmentI;
+import jalview.datamodel.DBRefEntry;
+import jalview.datamodel.GeneLociI;
import jalview.datamodel.Sequence;
import jalview.datamodel.SequenceFeature;
import jalview.datamodel.SequenceI;
* <li>resolves an external identifier by looking up xref-ed gene ids</li>
* <li>fetches the gene sequence</li>
* <li>fetches features on the sequence</li>
- * <li>identifies "transcript" features whose Parent is the requested gene</li>
+ * <li>identifies "transcript" features whose Parent is the requested
+ * gene</li>
* <li>fetches the transcript sequence for each transcript</li>
* <li>makes a mapping from the gene to each transcript</li>
* <li>copies features from gene to transcript sequences</li>
{
continue;
}
+
if (geneAlignment.getHeight() == 1)
{
+ findGeneLoci(geneAlignment.getSequenceAt(0), geneId);
getTranscripts(geneAlignment, geneId);
}
if (al == null)
}
/**
- * Converts a query, which may contain one or more gene or transcript
- * identifiers, into a non-redundant list of gene identifiers.
+ * Calls the /lookup/id REST service, parses the response for gene
+ * coordinates, and if successful, adds these to the sequence
+ *
+ * @param seq
+ * @param geneId
+ */
+ void findGeneLoci(SequenceI seq, String geneId)
+ {
+ /*
+ * if sequence description is in gene loci format, parse this
+ * - but try remote lookup anyway as 'better' if available
+ */
+ parseChromosomeLocations(seq);
+
+ GeneLociI geneLoci = new EnsemblLookup(getDomain()).getGeneLoci(geneId);
+ if (geneLoci != null)
+ {
+ seq.setGeneLoci(geneLoci.getSpeciesId(), geneLoci.getAssemblyId(),
+ geneLoci.getChromosomeId(), geneLoci.getMap());
+ }
+ }
+
+ /**
+ * Parses and saves fields of an Ensembl-style description e.g.
+ * chromosome:GRCh38:17:45051610:45109016:1
+ *
+ * @param seq
+ */
+ boolean parseChromosomeLocations(SequenceI seq)
+ {
+ String description = seq.getDescription();
+ if (description == null)
+ {
+ return false;
+ }
+ String[] tokens = description.split(":");
+ if (tokens.length == 6 && tokens[0].startsWith(DBRefEntry.CHROMOSOME))
+ {
+ String ref = tokens[1];
+ String chrom = tokens[2];
+ try
+ {
+ int chStart = Integer.parseInt(tokens[3]);
+ int chEnd = Integer.parseInt(tokens[4]);
+ boolean forwardStrand = "1".equals(tokens[5]);
+ String species = ""; // not known here
+ int[] from = new int[] { seq.getStart(), seq.getEnd() };
+ int[] to = new int[] { forwardStrand ? chStart : chEnd,
+ forwardStrand ? chEnd : chStart };
+ MapList map = new MapList(from, to, 1, 1);
+ seq.setGeneLoci(species, ref, chrom, map);
+ return true;
+ } catch (NumberFormatException e)
+ {
+ System.err.println("Bad integers in description " + description);
+ }
+ }
+ return false;
+ }
+
+ /**
+ * Converts a query, which may contain one or more gene, transcript, or
+ * external (to Ensembl) identifiers, into a non-redundant list of gene
+ * identifiers.
*
* @param accessions
* @return
for (String acc : accessions.split(getAccessionSeparator()))
{
- if (isGeneIdentifier(acc))
- {
- if (!geneIds.contains(acc))
- {
- geneIds.add(acc);
- }
- }
-
/*
- * if given a transcript id, look up its gene parent
+ * First try lookup as an Ensembl (gene or transcript) identifier
*/
- else if (isTranscriptIdentifier(acc))
+ String geneId = new EnsemblLookup(getDomain()).getGeneId(acc);
+ if (geneId != null)
{
- String geneId = new EnsemblLookup(getDomain()).getParent(acc);
- if (geneId != null && !geneIds.contains(geneId))
+ if (!geneIds.contains(geneId))
{
geneIds.add(geneId);
}
}
- else if (isProteinIdentifier(acc))
- {
- String tscriptId = new EnsemblLookup(getDomain()).getParent(acc);
- if (tscriptId != null)
- {
- String geneId = new EnsemblLookup(getDomain())
- .getParent(tscriptId);
-
- if (geneId != null && !geneIds.contains(geneId))
- {
- geneIds.add(geneId);
- }
- }
- // NOTE - acc is lost if it resembles an ENS.+ ID but isn't actually
- // resolving to one... e.g. ENSMICP00000009241
- }
- /*
- * if given a gene or other external name, lookup and fetch
- * the corresponding gene for all model organisms
- */
else
{
+ /*
+ * if given a gene or other external name, lookup and fetch
+ * the corresponding gene for all model organisms
+ */
List<String> ids = new EnsemblSymbol(getDomain(), getDbSource(),
- getDbVersion()).getIds(acc);
- for (String geneId : ids)
+ getDbVersion()).getGeneIds(acc);
+ for (String id : ids)
{
- if (!geneIds.contains(geneId))
+ if (!geneIds.contains(id))
{
- geneIds.add(geneId);
+ geneIds.add(id);
}
}
}
}
/**
- * Attempts to get Ensembl stable identifiers for model organisms for a gene
- * name by calling the xrefs symbol REST service to resolve the gene name.
- *
- * @param query
- * @return
- */
- protected String getGeneIdentifiersForName(String query)
- {
- List<String> ids = new EnsemblSymbol(getDomain(), getDbSource(),
- getDbVersion()).getIds(query);
- if (ids != null)
- {
- for (String id : ids)
- {
- if (isGeneIdentifier(id))
- {
- return id;
- }
- }
- }
- return null;
- }
-
- /**
* Constructs all transcripts for the gene, as identified by "transcript"
* features whose Parent is the requested gene. The coding transcript
* sequences (i.e. with introns omitted) are added to the alignment.
* the parent gene sequence, with features
* @return
*/
- SequenceI makeTranscript(SequenceFeature transcriptFeature,
- AlignmentI al, SequenceI gene)
+ SequenceI makeTranscript(SequenceFeature transcriptFeature, AlignmentI al,
+ SequenceI gene)
{
String accId = getTranscriptId(transcriptFeature);
if (accId == null)
mappedFrom.add(new int[] { sf.getBegin(), sf.getEnd() });
}
- Sequence transcript = new Sequence(accId, seqChars, 1, transcriptLength);
+ Sequence transcript = new Sequence(accId, seqChars, 1,
+ transcriptLength);
/*
* Ensembl has gene name as transcript Name
cdna.transferFeatures(gene.getFeatures().getPositionalFeatures(),
transcript.getDatasetSequence(), mapping, parentId);
+ mapTranscriptToChromosome(transcript, gene, mapping);
+
/*
* fetch and save cross-references
*/
}
/**
+ * If the gene has a mapping to chromosome coordinates, derive the transcript
+ * chromosome regions and save on the transcript sequence
+ *
+ * @param transcript
+ * @param gene
+ * @param mapping
+ * the mapping from gene to transcript positions
+ */
+ protected void mapTranscriptToChromosome(SequenceI transcript,
+ SequenceI gene, MapList mapping)
+ {
+ GeneLociI loci = gene.getGeneLoci();
+ if (loci == null)
+ {
+ return;
+ }
+
+ MapList geneMapping = loci.getMap();
+
+ List<int[]> exons = mapping.getFromRanges();
+ List<int[]> transcriptLoci = new ArrayList<>();
+
+ for (int[] exon : exons)
+ {
+ transcriptLoci.add(geneMapping.locateInTo(exon[0], exon[1]));
+ }
+
+ List<int[]> transcriptRange = Arrays.asList(new int[] {
+ transcript.getStart(), transcript.getEnd() });
+ MapList mapList = new MapList(transcriptRange, transcriptLoci, 1, 1);
+
+ transcript.setGeneLoci(loci.getSpeciesId(), loci.getAssemblyId(),
+ loci.getChromosomeId(), mapList);
+ }
+
+ /**
* Returns the 'transcript_id' property of the sequence feature (or null)
*
* @param feature
/**
* Returns a list of the transcript features on the sequence whose Parent is
* the gene for the accession id.
+ * <p>
+ * Transcript features are those of type "transcript", or any of its sub-types
+ * in the Sequence Ontology e.g. "mRNA", "processed_transcript". We also
+ * include "NMD_transcript_variant", because this type behaves like a
+ * transcript identifier in Ensembl, although strictly speaking it is not in
+ * the SO.
*
* @param accId
* @param geneSequence
List<SequenceFeature> transcriptFeatures = new ArrayList<SequenceFeature>();
String parentIdentifier = GENE_PREFIX + accId;
- // todo optimise here by transcript type!
+
List<SequenceFeature> sfs = geneSequence.getFeatures()
- .getPositionalFeatures();
+ .getFeaturesByOntology(SequenceOntologyI.TRANSCRIPT);
+ sfs.addAll(geneSequence.getFeatures().getPositionalFeatures(
+ SequenceOntologyI.NMD_TRANSCRIPT_VARIANT));
for (SequenceFeature sf : sfs)
{
- if (isTranscript(sf.getType()))
+ String parent = (String) sf.getValue(PARENT);
+ if (parentIdentifier.equals(parent))
{
- String parent = (String) sf.getValue(PARENT);
- if (parentIdentifier.equals(parent))
- {
- transcriptFeatures.add(sf);
- }
+ transcriptFeatures.add(sf);
}
}
@Override
public boolean isFeatureDisplayed(String type)
{
- return (so.isA(type, SequenceOntologyI.EXON) || so.isA(type,
- SequenceOntologyI.SEQUENCE_VARIANT));
+ return (so.isA(type, SequenceOntologyI.EXON)
+ || so.isA(type, SequenceOntologyI.SEQUENCE_VARIANT));
}
@Override