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
115 * <li>fetches the transcript sequence for each transcript</li>
116 * <li>makes a mapping from the gene to each transcript</li>
117 * <li>copies features from gene to transcript sequences</li>
118 * <li>fetches the protein sequence for each transcript, maps and saves it as
119 * a cross-reference</li>
120 * <li>aligns each transcript against the gene sequence based on the position
125 * a single gene or transcript identifier or gene name
126 * @return an alignment containing a gene, and possibly transcripts, or null
129 public AlignmentI getSequenceRecords(String query) throws Exception
132 * convert to a non-duplicated list of gene identifiers
134 List<String> geneIds = getGeneIds(query);
136 AlignmentI al = null;
137 for (String geneId : geneIds)
140 * fetch the gene sequence(s) with features and xrefs
142 AlignmentI geneAlignment = super.getSequenceRecords(geneId);
143 if (geneAlignment == null)
147 if (geneAlignment.getHeight() == 1)
149 getTranscripts(geneAlignment, geneId);
157 al.append(geneAlignment);
164 * Converts a query, which may contain one or more gene, transcript, or
165 * external (to Ensembl) identifiers, into a non-redundant list of gene
171 List<String> getGeneIds(String accessions)
173 List<String> geneIds = new ArrayList<String>();
175 for (String acc : accessions.split(getAccessionSeparator()))
178 * First try lookup as an Ensembl (gene or transcript) identifier
180 String geneId = new EnsemblLookup(getDomain()).getGeneId(acc);
183 if (!geneIds.contains(geneId))
191 * if given a gene or other external name, lookup and fetch
192 * the corresponding gene for all model organisms
194 List<String> ids = new EnsemblSymbol(getDomain(), getDbSource(),
195 getDbVersion()).getGeneIds(acc);
196 for (String id : ids)
198 if (!geneIds.contains(id))
209 * Constructs all transcripts for the gene, as identified by "transcript"
210 * features whose Parent is the requested gene. The coding transcript
211 * sequences (i.e. with introns omitted) are added to the alignment.
217 protected void getTranscripts(AlignmentI al, String accId)
220 SequenceI gene = al.getSequenceAt(0);
221 List<SequenceFeature> transcriptFeatures = getTranscriptFeatures(accId,
224 for (SequenceFeature transcriptFeature : transcriptFeatures)
226 makeTranscript(transcriptFeature, al, gene);
229 clearGeneFeatures(gene);
233 * Remove unwanted features (transcript, exon, CDS) from the gene sequence
234 * after we have used them to derive transcripts and transfer features
238 protected void clearGeneFeatures(SequenceI gene)
241 * Note we include NMD_transcript_variant here because it behaves like
242 * 'transcript' in Ensembl, although strictly speaking it is not
243 * (it is a sub-type of sequence_variant)
245 String[] soTerms = new String[] {
246 SequenceOntologyI.NMD_TRANSCRIPT_VARIANT,
247 SequenceOntologyI.TRANSCRIPT, SequenceOntologyI.EXON,
248 SequenceOntologyI.CDS };
249 List<SequenceFeature> sfs = gene.getFeatures().getFeaturesByOntology(
251 for (SequenceFeature sf : sfs)
253 gene.deleteFeature(sf);
258 * Constructs a spliced transcript sequence by finding 'exon' features for the
259 * given id (or failing that 'CDS'). Copies features on to the new sequence.
260 * 'Aligns' the new sequence against the gene sequence by padding with gaps,
261 * and adds it to the alignment.
263 * @param transcriptFeature
265 * the alignment to which to add the new sequence
267 * the parent gene sequence, with features
270 SequenceI makeTranscript(SequenceFeature transcriptFeature, AlignmentI al,
273 String accId = getTranscriptId(transcriptFeature);
280 * NB we are mapping from gene sequence (not genome), so do not
281 * need to check for reverse strand (gene and transcript sequences
282 * are in forward sense)
286 * make a gene-length sequence filled with gaps
287 * we will fill in the bases for transcript regions
289 char[] seqChars = new char[gene.getLength()];
290 Arrays.fill(seqChars, al.getGapCharacter());
293 * look for exon features of the transcript, failing that for CDS
294 * (for example ENSG00000124610 has 1 CDS but no exon features)
296 String parentId = "transcript:" + accId;
297 List<SequenceFeature> splices = findFeatures(gene,
298 SequenceOntologyI.EXON, parentId);
299 if (splices.isEmpty())
301 splices = findFeatures(gene, SequenceOntologyI.CDS, parentId);
303 SequenceFeatures.sortFeatures(splices, true);
305 int transcriptLength = 0;
306 final char[] geneChars = gene.getSequence();
307 int offset = gene.getStart(); // to convert to 0-based positions
308 List<int[]> mappedFrom = new ArrayList<int[]>();
310 for (SequenceFeature sf : splices)
312 int start = sf.getBegin() - offset;
313 int end = sf.getEnd() - offset;
314 int spliceLength = end - start + 1;
315 System.arraycopy(geneChars, start, seqChars, start, spliceLength);
316 transcriptLength += spliceLength;
317 mappedFrom.add(new int[] { sf.getBegin(), sf.getEnd() });
320 Sequence transcript = new Sequence(accId, seqChars, 1,
324 * Ensembl has gene name as transcript Name
325 * EnsemblGenomes doesn't, but has a url-encoded description field
327 String description = (String) transcriptFeature.getValue(NAME);
328 if (description == null)
330 description = (String) transcriptFeature.getValue(DESCRIPTION);
332 if (description != null)
336 transcript.setDescription(URLDecoder.decode(description, "UTF-8"));
337 } catch (UnsupportedEncodingException e)
339 e.printStackTrace(); // as if
342 transcript.createDatasetSequence();
344 al.addSequence(transcript);
347 * transfer features to the new sequence; we use EnsemblCdna to do this,
348 * to filter out unwanted features types (see method retainFeature)
350 List<int[]> mapTo = new ArrayList<int[]>();
351 mapTo.add(new int[] { 1, transcriptLength });
352 MapList mapping = new MapList(mappedFrom, mapTo, 1, 1);
353 EnsemblCdna cdna = new EnsemblCdna(getDomain());
354 cdna.transferFeatures(gene.getFeatures().getPositionalFeatures(),
355 transcript.getDatasetSequence(), mapping, parentId);
358 * fetch and save cross-references
360 cdna.getCrossReferences(transcript);
363 * and finally fetch the protein product and save as a cross-reference
365 cdna.addProteinProduct(transcript);
371 * Returns the 'transcript_id' property of the sequence feature (or null)
376 protected String getTranscriptId(SequenceFeature feature)
378 return (String) feature.getValue("transcript_id");
382 * Returns a list of the transcript features on the sequence whose Parent is
383 * the gene for the accession id.
385 * Transcript features are those of type "transcript", or any of its sub-types
386 * in the Sequence Ontology e.g. "mRNA", "processed_transcript". We also
387 * include "NMD_transcript_variant", because this type behaves like a
388 * transcript identifier in Ensembl, although strictly speaking it is not in
392 * @param geneSequence
395 protected List<SequenceFeature> getTranscriptFeatures(String accId,
396 SequenceI geneSequence)
398 List<SequenceFeature> transcriptFeatures = new ArrayList<SequenceFeature>();
400 String parentIdentifier = GENE_PREFIX + accId;
402 List<SequenceFeature> sfs = geneSequence.getFeatures()
403 .getFeaturesByOntology(SequenceOntologyI.TRANSCRIPT);
404 sfs.addAll(geneSequence.getFeatures().getPositionalFeatures(
405 SequenceOntologyI.NMD_TRANSCRIPT_VARIANT));
407 for (SequenceFeature sf : sfs)
409 String parent = (String) sf.getValue(PARENT);
410 if (parentIdentifier.equals(parent))
412 transcriptFeatures.add(sf);
416 return transcriptFeatures;
420 public String getDescription()
422 return "Fetches all transcripts and variant features for a gene or transcript";
426 * Default test query is a gene id (can also enter a transcript id)
429 public String getTestQuery()
431 return "ENSG00000157764"; // BRAF, 5 transcripts, reverse strand
432 // ENSG00000090266 // NDUFB2, 15 transcripts, forward strand
433 // ENSG00000101812 // H2BFM histone, 3 transcripts, forward strand
434 // ENSG00000123569 // H2BFWT histone, 2 transcripts, reverse strand
438 * Answers true for a feature of type 'gene' (or a sub-type of gene in the
439 * Sequence Ontology), whose ID is the accession we are retrieving
442 protected boolean identifiesSequence(SequenceFeature sf, String accId)
444 if (SequenceOntologyFactory.getInstance().isA(sf.getType(),
445 SequenceOntologyI.GENE))
447 String id = (String) sf.getValue(ID);
448 if ((GENE_PREFIX + accId).equals(id))
457 * Answers true unless feature type is 'gene', or 'transcript' with a parent
458 * which is a different gene. We need the gene features to identify the range,
459 * but it is redundant information on the gene sequence. Checking the parent
460 * allows us to drop transcript features which belong to different
461 * (overlapping) genes.
464 protected boolean retainFeature(SequenceFeature sf, String accessionId)
466 SequenceOntologyI so = SequenceOntologyFactory.getInstance();
467 String type = sf.getType();
468 if (so.isA(type, SequenceOntologyI.GENE))
472 if (isTranscript(type))
474 String parent = (String) sf.getValue(PARENT);
475 if (!(GENE_PREFIX + accessionId).equals(parent))
484 * Answers false. This allows an optimisation - a single 'gene' feature is all
485 * that is needed to identify the positions of the gene on the genomic
489 protected boolean isSpliceable()
495 * Override to do nothing as Ensembl doesn't return a protein sequence for a
499 protected void addProteinProduct(SequenceI querySeq)
504 public Regex getAccessionValidator()
506 return ACCESSION_REGEX;
510 * Returns a descriptor for suitable feature display settings with
512 * <li>only exon or sequence_variant features (or their subtypes in the
513 * Sequence Ontology) visible</li>
514 * <li>variant features coloured red</li>
515 * <li>exon features coloured by label (exon name)</li>
516 * <li>variants displayed above (on top of) exons</li>
520 public FeatureSettingsModelI getFeatureColourScheme()
522 return new FeatureSettingsAdapter()
524 SequenceOntologyI so = SequenceOntologyFactory.getInstance();
527 public boolean isFeatureDisplayed(String type)
529 return (so.isA(type, SequenceOntologyI.EXON)
530 || so.isA(type, SequenceOntologyI.SEQUENCE_VARIANT));
534 public FeatureColourI getFeatureColour(String type)
536 if (so.isA(type, SequenceOntologyI.EXON))
538 return new FeatureColour()
541 public boolean isColourByLabel()
547 if (so.isA(type, SequenceOntologyI.SEQUENCE_VARIANT))
549 return new FeatureColour()
553 public Color getColour()
563 * order to render sequence_variant after exon after the rest
566 public int compare(String feature1, String feature2)
568 if (so.isA(feature1, SequenceOntologyI.SEQUENCE_VARIANT))
572 if (so.isA(feature2, SequenceOntologyI.SEQUENCE_VARIANT))
576 if (so.isA(feature1, SequenceOntologyI.EXON))
580 if (so.isA(feature2, SequenceOntologyI.EXON))