import jalview.io.gff.SequenceOntologyFactory;
import jalview.io.gff.SequenceOntologyI;
import jalview.util.MapList;
+import jalview.util.StringUtils;
-import java.io.IOException;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.List;
{
private static final String GENE_PREFIX = "gene:";
- // TODO modify to accept other species e.g. ENSMUSGnnn
- private static final Regex ACCESSION_REGEX = new Regex(
- "(ENSG|ENST)[0-9]{11}$");
+ /*
+ * accepts anything as we will attempt lookup of gene or
+ * transcript id or gene name
+ */
+ private static final Regex ACCESSION_REGEX = new Regex(".*");
private static final EnsemblFeatureType[] FEATURES_TO_FETCH = {
EnsemblFeatureType.gene, EnsemblFeatureType.transcript,
}
/**
- * Builds an alignment of all transcripts for the requested gene:
+ * Returns an alignment containing the gene(s) for the given gene or
+ * transcript identifier, or external identifier (e.g. Uniprot id). If given a
+ * gene name or external identifier, returns any related gene sequences found
+ * for model organisms. If only a single gene is queried for, then its
+ * transcripts are also retrieved and added to the alignment. <br>
+ * Method:
* <ul>
+ * <li>resolves a transcript identifier by looking up its parent gene id</li>
+ * <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>aligns each transcript against the gene sequence based on the position
* mappings</li>
* </ul>
+ *
+ * @param query
+ * one or more identifiers separated by a space
+ * @return an alignment containing one or more genes, and possibly
+ * transcripts, or null
*/
@Override
public AlignmentI getSequenceRecords(String query) throws Exception
{
- List<String> transcriptsWanted = null;
+ // todo: tidy up handling of one or multiple accession ids
+ String[] queries = query.split(getAccessionSeparator());
+ /*
+ * if given a transcript id, look up its gene parent
+ */
if (isTranscriptIdentifier(query))
{
- transcriptsWanted = Arrays.asList(query
- .split(getAccessionSeparator()));
- query = getGeneForTranscript(query);
+ // we are assuming all transcripts have the same gene parent here
+ query = new EnsemblLookup().getParent(queries[0]);
if (query == null)
{
return null;
}
}
+ /*
+ * if given a gene or other external name, lookup and fetch
+ * the corresponding gene for all model organisms
+ */
+ if (!isGeneIdentifier(query))
+ {
+ List<String> geneIds = new EnsemblSymbol().getIds(query);
+ if (geneIds.isEmpty())
+ {
+ return null;
+ }
+ String theIds = StringUtils.listToDelimitedString(geneIds,
+ getAccessionSeparator());
+ return getSequenceRecords(theIds);
+ }
+
AlignmentI al = super.getSequenceRecords(query);
- if (al.getHeight() > 0)
+
+ /*
+ * if we retrieved a single gene, get its transcripts as well
+ */
+ if (al.getHeight() == 1)
{
- getTranscripts(al, query, transcriptsWanted);
+ getTranscripts(al, query);
}
return al;
}
/**
- * Gets the parent gene identifier for a given transcript identifier, by
- * retrieving 'transcript' features overlapping the transcript, and finding
- * the Parent property of the feature whose id is the given identifier.
+ * 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 getGeneForTranscript(String transcriptId)
+ protected String getGeneIdentifiersForName(String query)
{
- String geneId = null;
-
- /*
- * reduce multiple transcripts (e.g. from Uniprot x-ref) to the first
- * one only as representative (they should all have the same gene)
- */
- transcriptId = transcriptId.split(getAccessionSeparator())[0];
-
- try
+ List<String> ids = new EnsemblSymbol().getIds(query);
+ if (ids != null)
{
- EnsemblFeatureType[] geneFeature = new EnsemblFeatureType[] { EnsemblFeatureType.transcript };
- AlignmentI al = new EnsemblFeatures().getSequenceRecords(
- transcriptId, geneFeature);
- if (al != null && al.getHeight() > 0)
+ for (String id : ids)
{
- SequenceFeature[] sfs = al.getSequenceAt(0).getSequenceFeatures();
- if (sfs != null)
+ if (isGeneIdentifier(id))
{
- for (SequenceFeature sf : sfs)
- {
- if (transcriptId.equals(getTranscriptId(sf)))
- {
- String parent = (String) sf.getValue(PARENT);
- if (parent != null && parent.startsWith(GENE_PREFIX))
- {
- geneId = parent.substring(5);
- }
- break;
- }
- }
+ return id;
}
}
- return geneId;
- } catch (IOException e)
- {
- System.err.println("Error retrieving gene id for " + transcriptId
- + ": " + e.getMessage());
- return null;
}
+ return null;
}
/**
*
* @param al
* @param accId
- * @param transcriptsWanted
- * optional list of transcript ids to filter by
* @throws Exception
*/
- protected void getTranscripts(AlignmentI al, String accId,
- List<String> transcriptsWanted)
+ protected void getTranscripts(AlignmentI al, String accId)
throws Exception
{
SequenceI gene = al.getSequenceAt(0);
List<SequenceFeature> transcriptFeatures = getTranscriptFeatures(accId,
- gene, transcriptsWanted);
+ gene);
for (SequenceFeature transcriptFeature : transcriptFeatures)
{
transcript.getDatasetSequence(), mapping, parentId);
/*
+ * fetch and save cross-references
+ */
+ super.getCrossReferences(transcript);
+
+ /*
* and finally fetch the protein product and save as a cross-reference
*/
new EnsemblCdna().addProteinProduct(transcript);
*
* @param accId
* @param geneSequence
- * @param transcriptsWanted
- * optional list of ids to filter on
* @return
*/
protected List<SequenceFeature> getTranscriptFeatures(String accId,
- SequenceI geneSequence, List<String> transcriptsWanted)
+ SequenceI geneSequence)
{
List<SequenceFeature> transcriptFeatures = new ArrayList<SequenceFeature>();
{
if (isTranscript(sf.getType()))
{
- if (transcriptsWanted != null)
- {
- String transcriptId = (String) sf.getValue("transcript_id");
- if (!transcriptsWanted.contains(transcriptId))
- {
- // continue;
- }
- }
String parent = (String) sf.getValue(PARENT);
if (parentIdentifier.equals(parent))
{