X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fext%2Fensembl%2FEnsemblGene.java;h=36b19e25375aaa31ed24ae1cf694e0906ab0fd11;hb=a4f2a7f356b8edab17a9a5bb6f2e71a1419792a9;hp=45e8e345a2416e4b6ff39a56c38cdf87096e0032;hpb=f5ddcb62310a9f32545231d0c5e1d7c364bfbb53;p=jalview.git diff --git a/src/jalview/ext/ensembl/EnsemblGene.java b/src/jalview/ext/ensembl/EnsemblGene.java index 45e8e34..36b19e2 100644 --- a/src/jalview/ext/ensembl/EnsemblGene.java +++ b/src/jalview/ext/ensembl/EnsemblGene.java @@ -23,7 +23,8 @@ package jalview.ext.ensembl; import jalview.api.FeatureColourI; import jalview.api.FeatureSettingsModelI; import jalview.datamodel.AlignmentI; -import jalview.datamodel.GeneLoci; +import jalview.datamodel.DBRefEntry; +import jalview.datamodel.GeneLociI; import jalview.datamodel.Sequence; import jalview.datamodel.SequenceFeature; import jalview.datamodel.SequenceI; @@ -99,6 +100,12 @@ public class EnsemblGene extends EnsemblSeqProxy return EnsemblSeqType.GENOMIC; } + @Override + protected String getObjectType() + { + return OBJECT_TYPE_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 @@ -145,8 +152,14 @@ public class EnsemblGene extends EnsemblSeqProxy { continue; } + if (geneAlignment.getHeight() == 1) { + // ensure id has 'correct' case for the Ensembl identifier + geneId = geneAlignment.getSequenceAt(0).getName(); + + findGeneLoci(geneAlignment.getSequenceAt(0), geneId); + getTranscripts(geneAlignment, geneId); } if (al == null) @@ -162,66 +175,104 @@ public class EnsemblGene extends EnsemblSeqProxy } /** - * 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. If this fails, + * fall back on trying to parse the sequence description in case it is in + * Ensembl-gene format e.g. chromosome:GRCh38:17:45051610:45109016:1. + * + * @param seq + * @param geneId + */ + void findGeneLoci(SequenceI seq, String geneId) + { + GeneLociI geneLoci = new EnsemblLookup(getDomain()).getGeneLoci(geneId); + if (geneLoci != null) + { + seq.setGeneLoci(geneLoci.getSpeciesId(), geneLoci.getAssemblyId(), + geneLoci.getChromosomeId(), geneLoci.getMap()); + } + else + { + parseChromosomeLocations(seq); + } + } + + /** + * 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 */ List getGeneIds(String accessions) { - List geneIds = new ArrayList(); + List geneIds = new ArrayList<>(); 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 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); } } } @@ -230,30 +281,6 @@ public class EnsemblGene extends EnsemblSeqProxy } /** - * 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 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. @@ -353,7 +380,7 @@ public class EnsemblGene extends EnsemblSeqProxy int transcriptLength = 0; final char[] geneChars = gene.getSequence(); int offset = gene.getStart(); // to convert to 0-based positions - List mappedFrom = new ArrayList(); + List mappedFrom = new ArrayList<>(); for (SequenceFeature sf : splices) { @@ -395,7 +422,7 @@ public class EnsemblGene extends EnsemblSeqProxy * transfer features to the new sequence; we use EnsemblCdna to do this, * to filter out unwanted features types (see method retainFeature) */ - List mapTo = new ArrayList(); + List mapTo = new ArrayList<>(); mapTo.add(new int[] { 1, transcriptLength }); MapList mapping = new MapList(mappedFrom, mapTo, 1, 1); EnsemblCdna cdna = new EnsemblCdna(getDomain()); @@ -426,25 +453,16 @@ public class EnsemblGene extends EnsemblSeqProxy * @param mapping * the mapping from gene to transcript positions */ - protected void mapTranscriptToChromosome(Sequence transcript, + protected void mapTranscriptToChromosome(SequenceI transcript, SequenceI gene, MapList mapping) { - GeneLoci loci = gene.getGeneLoci(); + GeneLociI loci = gene.getGeneLoci(); if (loci == null) { return; } - /* - * patch to ensure gene to chromosome mapping is complete - * (in case created before gene length was known) - */ - MapList geneMapping = loci.mapping; - if (geneMapping.getFromRanges().get(0)[1] == 0) - { - geneMapping.getFromRanges().get(0)[0] = gene.getStart(); - geneMapping.getFromRanges().get(0)[1] = gene.getEnd(); - } + MapList geneMapping = loci.getMap(); List exons = mapping.getFromRanges(); List transcriptLoci = new ArrayList<>(); @@ -457,10 +475,9 @@ public class EnsemblGene extends EnsemblSeqProxy List transcriptRange = Arrays.asList(new int[] { transcript.getStart(), transcript.getEnd() }); MapList mapList = new MapList(transcriptRange, transcriptLoci, 1, 1); - GeneLoci gl = new GeneLoci(loci.species, loci.assembly, - loci.chromosome, mapList); - transcript.setGeneLoci(gl); + transcript.setGeneLoci(loci.getSpeciesId(), loci.getAssemblyId(), + loci.getChromosomeId(), mapList); } /** @@ -477,6 +494,12 @@ public class EnsemblGene extends EnsemblSeqProxy /** * Returns a list of the transcript features on the sequence whose Parent is * the gene for the accession id. + *

+ * 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 @@ -485,22 +508,21 @@ public class EnsemblGene extends EnsemblSeqProxy protected List getTranscriptFeatures(String accId, SequenceI geneSequence) { - List transcriptFeatures = new ArrayList(); + List transcriptFeatures = new ArrayList<>(); String parentIdentifier = GENE_PREFIX + accId; - // todo optimise here by transcript type! + List 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.equalsIgnoreCase(parent)) { - String parent = (String) sf.getValue(PARENT); - if (parentIdentifier.equals(parent)) - { - transcriptFeatures.add(sf); - } + transcriptFeatures.add(sf); } } @@ -526,22 +548,27 @@ public class EnsemblGene extends EnsemblSeqProxy } /** - * Answers true for a feature of type 'gene' (or a sub-type of gene in the - * Sequence Ontology), whose ID is the accession we are retrieving + * Answers a list of sequence features (if any) whose type is 'gene' (or a + * subtype of gene in the Sequence Ontology), and whose ID is the accession we + * are retrieving */ @Override - protected boolean identifiesSequence(SequenceFeature sf, String accId) + protected List getIdentifyingFeatures(SequenceI seq, + String accId) { - if (SequenceOntologyFactory.getInstance().isA(sf.getType(), - SequenceOntologyI.GENE)) + List result = new ArrayList<>(); + List sfs = seq.getFeatures() + .getFeaturesByOntology(SequenceOntologyI.GENE); + for (SequenceFeature sf : sfs) { - String id = (String) sf.getValue(ID); - if ((GENE_PREFIX + accId).equals(id)) + // NB features as gff use 'ID'; rest services return as 'id' + String id = (String) sf.getValue("ID"); + if ((GENE_PREFIX + accId).equalsIgnoreCase(id)) { - return true; + result.add(sf); } } - return false; + return result; } /** @@ -563,7 +590,7 @@ public class EnsemblGene extends EnsemblSeqProxy if (isTranscript(type)) { String parent = (String) sf.getValue(PARENT); - if (!(GENE_PREFIX + accessionId).equals(parent)) + if (!(GENE_PREFIX + accessionId).equalsIgnoreCase(parent)) { return false; } @@ -572,17 +599,6 @@ public class EnsemblGene extends EnsemblSeqProxy } /** - * Answers false. This allows an optimisation - a single 'gene' feature is all - * that is needed to identify the positions of the gene on the genomic - * sequence. - */ - @Override - protected boolean isSpliceable() - { - return false; - } - - /** * Override to do nothing as Ensembl doesn't return a protein sequence for a * gene identifier */