2 * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
3 * Copyright (C) $$Year-Rel$$ The Jalview Authors
5 * This file is part of Jalview.
7 * Jalview is free software: you can redistribute it and/or
8 * modify it under the terms of the GNU General Public License
9 * as published by the Free Software Foundation, either version 3
10 * of the License, or (at your option) any later version.
12 * Jalview is distributed in the hope that it will be useful, but
13 * WITHOUT ANY WARRANTY; without even the implied warranty
14 * of MERCHANTABILITY or FITNESS FOR A PARTICULAR
15 * PURPOSE. See the GNU General Public License for more details.
17 * You should have received a copy of the GNU General Public License
18 * along with Jalview. If not, see <http://www.gnu.org/licenses/>.
19 * The Jalview Authors are detailed in the 'AUTHORS' file.
21 package jalview.ext.ensembl;
23 import jalview.api.FeatureColourI;
24 import jalview.api.FeatureSettingsModelI;
25 import jalview.datamodel.AlignmentI;
26 import jalview.datamodel.Sequence;
27 import jalview.datamodel.SequenceFeature;
28 import jalview.datamodel.SequenceI;
29 import jalview.datamodel.features.SequenceFeatures;
30 import jalview.io.gff.SequenceOntologyFactory;
31 import jalview.io.gff.SequenceOntologyI;
32 import jalview.schemes.FeatureColour;
33 import jalview.schemes.FeatureSettingsAdapter;
34 import jalview.util.MapList;
36 import java.awt.Color;
37 import java.io.UnsupportedEncodingException;
38 import java.net.URLDecoder;
39 import java.util.ArrayList;
40 import java.util.Arrays;
41 import java.util.List;
43 import com.stevesoft.pat.Regex;
46 * A class that fetches genomic sequence and all transcripts for an Ensembl gene
50 public class EnsemblGene extends EnsemblSeqProxy
52 private static final String GENE_PREFIX = "gene:";
55 * accepts anything as we will attempt lookup of gene or
56 * transcript id or gene name
58 private static final Regex ACCESSION_REGEX = new Regex(".*");
60 private static final EnsemblFeatureType[] FEATURES_TO_FETCH = {
61 EnsemblFeatureType.gene, EnsemblFeatureType.transcript,
62 EnsemblFeatureType.exon, EnsemblFeatureType.cds,
63 EnsemblFeatureType.variation };
66 * Default constructor (to use rest.ensembl.org)
74 * Constructor given the target domain to fetch data from
78 public EnsemblGene(String d)
84 public String getDbName()
90 protected EnsemblFeatureType[] getFeaturesToFetch()
92 return FEATURES_TO_FETCH;
96 protected EnsemblSeqType getSourceEnsemblType()
98 return EnsemblSeqType.GENOMIC;
102 * Returns an alignment containing the gene(s) for the given gene or
103 * transcript identifier, or external identifier (e.g. Uniprot id). If given a
104 * gene name or external identifier, returns any related gene sequences found
105 * for model organisms. If only a single gene is queried for, then its
106 * transcripts are also retrieved and added to the alignment. <br>
109 * <li>resolves a transcript identifier by looking up its parent gene id</li>
110 * <li>resolves an external identifier by looking up xref-ed gene ids</li>
111 * <li>fetches the gene sequence</li>
112 * <li>fetches features on the sequence</li>
113 * <li>identifies "transcript" features whose Parent is the requested gene</li>
114 * <li>fetches the transcript sequence for each transcript</li>
115 * <li>makes a mapping from the gene to each transcript</li>
116 * <li>copies features from gene to transcript sequences</li>
117 * <li>fetches the protein sequence for each transcript, maps and saves it as
118 * a cross-reference</li>
119 * <li>aligns each transcript against the gene sequence based on the position
124 * a single gene or transcript identifier or gene name
125 * @return an alignment containing a gene, and possibly transcripts, or null
128 public AlignmentI getSequenceRecords(String query) throws Exception
131 * convert to a non-duplicated list of gene identifiers
133 List<String> geneIds = getGeneIds(query);
135 AlignmentI al = null;
136 for (String geneId : geneIds)
139 * fetch the gene sequence(s) with features and xrefs
141 AlignmentI geneAlignment = super.getSequenceRecords(geneId);
142 if (geneAlignment == null)
146 if (geneAlignment.getHeight() == 1)
148 getTranscripts(geneAlignment, geneId);
156 al.append(geneAlignment);
163 * Converts a query, which may contain one or more gene or transcript
164 * identifiers, into a non-redundant list of gene identifiers.
169 List<String> getGeneIds(String accessions)
171 List<String> geneIds = new ArrayList<String>();
173 for (String acc : accessions.split(getAccessionSeparator()))
175 if (isGeneIdentifier(acc))
177 if (!geneIds.contains(acc))
184 * if given a transcript id, look up its gene parent
186 else if (isTranscriptIdentifier(acc))
188 String geneId = new EnsemblLookup(getDomain()).getParent(acc);
189 if (geneId != null && !geneIds.contains(geneId))
196 * if given a gene or other external name, lookup and fetch
197 * the corresponding gene for all model organisms
201 List<String> ids = new EnsemblSymbol(getDomain(), getDbSource(),
202 getDbVersion()).getIds(acc);
203 for (String geneId : ids)
205 if (!geneIds.contains(geneId))
216 * Attempts to get Ensembl stable identifiers for model organisms for a gene
217 * name by calling the xrefs symbol REST service to resolve the gene name.
222 protected String getGeneIdentifiersForName(String query)
224 List<String> ids = new EnsemblSymbol(getDomain(), getDbSource(),
225 getDbVersion()).getIds(query);
228 for (String id : ids)
230 if (isGeneIdentifier(id))
240 * Constructs all transcripts for the gene, as identified by "transcript"
241 * features whose Parent is the requested gene. The coding transcript
242 * sequences (i.e. with introns omitted) are added to the alignment.
248 protected void getTranscripts(AlignmentI al, String accId)
251 SequenceI gene = al.getSequenceAt(0);
252 List<SequenceFeature> transcriptFeatures = getTranscriptFeatures(accId,
255 for (SequenceFeature transcriptFeature : transcriptFeatures)
257 makeTranscript(transcriptFeature, al, gene);
260 clearGeneFeatures(gene);
264 * Remove unwanted features (transcript, exon, CDS) from the gene sequence
265 * after we have used them to derive transcripts and transfer features
269 protected void clearGeneFeatures(SequenceI gene)
272 * Note we include NMD_transcript_variant here because it behaves like
273 * 'transcript' in Ensembl, although strictly speaking it is not
274 * (it is a sub-type of sequence_variant)
276 String[] soTerms = new String[] {
277 SequenceOntologyI.NMD_TRANSCRIPT_VARIANT,
278 SequenceOntologyI.TRANSCRIPT, SequenceOntologyI.EXON,
279 SequenceOntologyI.CDS };
280 List<SequenceFeature> sfs = gene.getFeatures().getFeaturesByOntology(
282 for (SequenceFeature sf : sfs)
284 gene.deleteFeature(sf);
289 * Constructs a spliced transcript sequence by finding 'exon' features for the
290 * given id (or failing that 'CDS'). Copies features on to the new sequence.
291 * 'Aligns' the new sequence against the gene sequence by padding with gaps,
292 * and adds it to the alignment.
294 * @param transcriptFeature
296 * the alignment to which to add the new sequence
298 * the parent gene sequence, with features
301 SequenceI makeTranscript(SequenceFeature transcriptFeature,
302 AlignmentI al, SequenceI gene)
304 String accId = getTranscriptId(transcriptFeature);
311 * NB we are mapping from gene sequence (not genome), so do not
312 * need to check for reverse strand (gene and transcript sequences
313 * are in forward sense)
317 * make a gene-length sequence filled with gaps
318 * we will fill in the bases for transcript regions
320 char[] seqChars = new char[gene.getLength()];
321 Arrays.fill(seqChars, al.getGapCharacter());
324 * look for exon features of the transcript, failing that for CDS
325 * (for example ENSG00000124610 has 1 CDS but no exon features)
327 String parentId = "transcript:" + accId;
328 List<SequenceFeature> splices = findFeatures(gene,
329 SequenceOntologyI.EXON, parentId);
330 if (splices.isEmpty())
332 splices = findFeatures(gene, SequenceOntologyI.CDS, parentId);
334 SequenceFeatures.sortFeatures(splices, true);
336 int transcriptLength = 0;
337 final char[] geneChars = gene.getSequence();
338 int offset = gene.getStart(); // to convert to 0-based positions
339 List<int[]> mappedFrom = new ArrayList<int[]>();
341 for (SequenceFeature sf : splices)
343 int start = sf.getBegin() - offset;
344 int end = sf.getEnd() - offset;
345 int spliceLength = end - start + 1;
346 System.arraycopy(geneChars, start, seqChars, start, spliceLength);
347 transcriptLength += spliceLength;
348 mappedFrom.add(new int[] { sf.getBegin(), sf.getEnd() });
351 Sequence transcript = new Sequence(accId, seqChars, 1, transcriptLength);
354 * Ensembl has gene name as transcript Name
355 * EnsemblGenomes doesn't, but has a url-encoded description field
357 String description = (String) transcriptFeature.getValue(NAME);
358 if (description == null)
360 description = (String) transcriptFeature.getValue(DESCRIPTION);
362 if (description != null)
366 transcript.setDescription(URLDecoder.decode(description, "UTF-8"));
367 } catch (UnsupportedEncodingException e)
369 e.printStackTrace(); // as if
372 transcript.createDatasetSequence();
374 al.addSequence(transcript);
377 * transfer features to the new sequence; we use EnsemblCdna to do this,
378 * to filter out unwanted features types (see method retainFeature)
380 List<int[]> mapTo = new ArrayList<int[]>();
381 mapTo.add(new int[] { 1, transcriptLength });
382 MapList mapping = new MapList(mappedFrom, mapTo, 1, 1);
383 EnsemblCdna cdna = new EnsemblCdna(getDomain());
384 cdna.transferFeatures(gene.getFeatures().getPositionalFeatures(),
385 transcript.getDatasetSequence(), mapping, parentId);
388 * fetch and save cross-references
390 cdna.getCrossReferences(transcript);
393 * and finally fetch the protein product and save as a cross-reference
395 cdna.addProteinProduct(transcript);
401 * Returns the 'transcript_id' property of the sequence feature (or null)
406 protected String getTranscriptId(SequenceFeature feature)
408 return (String) feature.getValue("transcript_id");
412 * Returns a list of the transcript features on the sequence whose Parent is
413 * the gene for the accession id.
416 * @param geneSequence
419 protected List<SequenceFeature> getTranscriptFeatures(String accId,
420 SequenceI geneSequence)
422 List<SequenceFeature> transcriptFeatures = new ArrayList<SequenceFeature>();
424 String parentIdentifier = GENE_PREFIX + accId;
425 // todo optimise here by transcript type!
426 List<SequenceFeature> sfs = geneSequence.getFeatures()
427 .getPositionalFeatures();
429 for (SequenceFeature sf : sfs)
431 if (isTranscript(sf.getType()))
433 String parent = (String) sf.getValue(PARENT);
434 if (parentIdentifier.equals(parent))
436 transcriptFeatures.add(sf);
441 return transcriptFeatures;
445 public String getDescription()
447 return "Fetches all transcripts and variant features for a gene or transcript";
451 * Default test query is a gene id (can also enter a transcript id)
454 public String getTestQuery()
456 return "ENSG00000157764"; // BRAF, 5 transcripts, reverse strand
457 // ENSG00000090266 // NDUFB2, 15 transcripts, forward strand
458 // ENSG00000101812 // H2BFM histone, 3 transcripts, forward strand
459 // ENSG00000123569 // H2BFWT histone, 2 transcripts, reverse strand
463 * Answers true for a feature of type 'gene' (or a sub-type of gene in the
464 * Sequence Ontology), whose ID is the accession we are retrieving
467 protected boolean identifiesSequence(SequenceFeature sf, String accId)
469 if (SequenceOntologyFactory.getInstance().isA(sf.getType(),
470 SequenceOntologyI.GENE))
472 String id = (String) sf.getValue(ID);
473 if ((GENE_PREFIX + accId).equals(id))
482 * Answers true unless feature type is 'gene', or 'transcript' with a parent
483 * which is a different gene. We need the gene features to identify the range,
484 * but it is redundant information on the gene sequence. Checking the parent
485 * allows us to drop transcript features which belong to different
486 * (overlapping) genes.
489 protected boolean retainFeature(SequenceFeature sf, String accessionId)
491 SequenceOntologyI so = SequenceOntologyFactory.getInstance();
492 String type = sf.getType();
493 if (so.isA(type, SequenceOntologyI.GENE))
497 if (isTranscript(type))
499 String parent = (String) sf.getValue(PARENT);
500 if (!(GENE_PREFIX + accessionId).equals(parent))
509 * Answers false. This allows an optimisation - a single 'gene' feature is all
510 * that is needed to identify the positions of the gene on the genomic
514 protected boolean isSpliceable()
520 * Override to do nothing as Ensembl doesn't return a protein sequence for a
524 protected void addProteinProduct(SequenceI querySeq)
529 public Regex getAccessionValidator()
531 return ACCESSION_REGEX;
535 * Returns a descriptor for suitable feature display settings with
537 * <li>only exon or sequence_variant features (or their subtypes in the
538 * Sequence Ontology) visible</li>
539 * <li>variant features coloured red</li>
540 * <li>exon features coloured by label (exon name)</li>
541 * <li>variants displayed above (on top of) exons</li>
545 public FeatureSettingsModelI getFeatureColourScheme()
547 return new FeatureSettingsAdapter()
549 SequenceOntologyI so = SequenceOntologyFactory.getInstance();
552 public boolean isFeatureDisplayed(String type)
554 return (so.isA(type, SequenceOntologyI.EXON) || so.isA(type,
555 SequenceOntologyI.SEQUENCE_VARIANT));
559 public FeatureColourI getFeatureColour(String type)
561 if (so.isA(type, SequenceOntologyI.EXON))
563 return new FeatureColour()
566 public boolean isColourByLabel()
572 if (so.isA(type, SequenceOntologyI.SEQUENCE_VARIANT))
574 return new FeatureColour()
578 public Color getColour()
588 * order to render sequence_variant after exon after the rest
591 public int compare(String feature1, String feature2)
593 if (so.isA(feature1, SequenceOntologyI.SEQUENCE_VARIANT))
597 if (so.isA(feature2, SequenceOntologyI.SEQUENCE_VARIANT))
601 if (so.isA(feature1, SequenceOntologyI.EXON))
605 if (so.isA(feature2, SequenceOntologyI.EXON))