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 protected String getObjectType()
104 return OBJECT_TYPE_GENE;
108 * Returns an alignment containing the gene(s) for the given gene or
109 * transcript identifier, or external identifier (e.g. Uniprot id). If given a
110 * gene name or external identifier, returns any related gene sequences found
111 * for model organisms. If only a single gene is queried for, then its
112 * transcripts are also retrieved and added to the alignment. <br>
115 * <li>resolves a transcript identifier by looking up its parent gene id</li>
116 * <li>resolves an external identifier by looking up xref-ed gene ids</li>
117 * <li>fetches the gene sequence</li>
118 * <li>fetches features on the sequence</li>
119 * <li>identifies "transcript" features whose Parent is the requested
121 * <li>fetches the transcript sequence for each transcript</li>
122 * <li>makes a mapping from the gene to each transcript</li>
123 * <li>copies features from gene to transcript sequences</li>
124 * <li>fetches the protein sequence for each transcript, maps and saves it as
125 * a cross-reference</li>
126 * <li>aligns each transcript against the gene sequence based on the position
131 * a single gene or transcript identifier or gene name
132 * @return an alignment containing a gene, and possibly transcripts, or null
135 public AlignmentI getSequenceRecords(String query) throws Exception
138 * convert to a non-duplicated list of gene identifiers
140 List<String> geneIds = getGeneIds(query);
142 AlignmentI al = null;
143 for (String geneId : geneIds)
146 * fetch the gene sequence(s) with features and xrefs
148 AlignmentI geneAlignment = super.getSequenceRecords(geneId);
149 if (geneAlignment == null)
153 if (geneAlignment.getHeight() == 1)
155 getTranscripts(geneAlignment, geneId);
163 al.append(geneAlignment);
170 * Converts a query, which may contain one or more gene, transcript, or
171 * external (to Ensembl) identifiers, into a non-redundant list of gene
177 List<String> getGeneIds(String accessions)
179 List<String> geneIds = new ArrayList<>();
181 for (String acc : accessions.split(getAccessionSeparator()))
184 * First try lookup as an Ensembl (gene or transcript) identifier
186 String geneId = new EnsemblLookup(getDomain()).getGeneId(acc);
189 if (!geneIds.contains(geneId))
197 * if given a gene or other external name, lookup and fetch
198 * the corresponding gene for all model organisms
200 List<String> ids = new EnsemblSymbol(getDomain(), getDbSource(),
201 getDbVersion()).getGeneIds(acc);
202 for (String id : ids)
204 if (!geneIds.contains(id))
215 * Constructs all transcripts for the gene, as identified by "transcript"
216 * features whose Parent is the requested gene. The coding transcript
217 * sequences (i.e. with introns omitted) are added to the alignment.
223 protected void getTranscripts(AlignmentI al, String accId)
226 SequenceI gene = al.getSequenceAt(0);
227 List<SequenceFeature> transcriptFeatures = getTranscriptFeatures(accId,
230 for (SequenceFeature transcriptFeature : transcriptFeatures)
232 makeTranscript(transcriptFeature, al, gene);
235 clearGeneFeatures(gene);
239 * Remove unwanted features (transcript, exon, CDS) from the gene sequence
240 * after we have used them to derive transcripts and transfer features
244 protected void clearGeneFeatures(SequenceI gene)
247 * Note we include NMD_transcript_variant here because it behaves like
248 * 'transcript' in Ensembl, although strictly speaking it is not
249 * (it is a sub-type of sequence_variant)
251 String[] soTerms = new String[] {
252 SequenceOntologyI.NMD_TRANSCRIPT_VARIANT,
253 SequenceOntologyI.TRANSCRIPT, SequenceOntologyI.EXON,
254 SequenceOntologyI.CDS };
255 List<SequenceFeature> sfs = gene.getFeatures().getFeaturesByOntology(
257 for (SequenceFeature sf : sfs)
259 gene.deleteFeature(sf);
264 * Constructs a spliced transcript sequence by finding 'exon' features for the
265 * given id (or failing that 'CDS'). Copies features on to the new sequence.
266 * 'Aligns' the new sequence against the gene sequence by padding with gaps,
267 * and adds it to the alignment.
269 * @param transcriptFeature
271 * the alignment to which to add the new sequence
273 * the parent gene sequence, with features
276 SequenceI makeTranscript(SequenceFeature transcriptFeature, AlignmentI al,
279 String accId = getTranscriptId(transcriptFeature);
286 * NB we are mapping from gene sequence (not genome), so do not
287 * need to check for reverse strand (gene and transcript sequences
288 * are in forward sense)
292 * make a gene-length sequence filled with gaps
293 * we will fill in the bases for transcript regions
295 char[] seqChars = new char[gene.getLength()];
296 Arrays.fill(seqChars, al.getGapCharacter());
299 * look for exon features of the transcript, failing that for CDS
300 * (for example ENSG00000124610 has 1 CDS but no exon features)
302 String parentId = "transcript:" + accId;
303 List<SequenceFeature> splices = findFeatures(gene,
304 SequenceOntologyI.EXON, parentId);
305 if (splices.isEmpty())
307 splices = findFeatures(gene, SequenceOntologyI.CDS, parentId);
309 SequenceFeatures.sortFeatures(splices, true);
311 int transcriptLength = 0;
312 final char[] geneChars = gene.getSequence();
313 int offset = gene.getStart(); // to convert to 0-based positions
314 List<int[]> mappedFrom = new ArrayList<>();
316 for (SequenceFeature sf : splices)
318 int start = sf.getBegin() - offset;
319 int end = sf.getEnd() - offset;
320 int spliceLength = end - start + 1;
321 System.arraycopy(geneChars, start, seqChars, start, spliceLength);
322 transcriptLength += spliceLength;
323 mappedFrom.add(new int[] { sf.getBegin(), sf.getEnd() });
326 Sequence transcript = new Sequence(accId, seqChars, 1,
330 * Ensembl has gene name as transcript Name
331 * EnsemblGenomes doesn't, but has a url-encoded description field
333 String description = (String) transcriptFeature.getValue(NAME);
334 if (description == null)
336 description = (String) transcriptFeature.getValue(DESCRIPTION);
338 if (description != null)
342 transcript.setDescription(URLDecoder.decode(description, "UTF-8"));
343 } catch (UnsupportedEncodingException e)
345 e.printStackTrace(); // as if
348 transcript.createDatasetSequence();
350 al.addSequence(transcript);
353 * transfer features to the new sequence; we use EnsemblCdna to do this,
354 * to filter out unwanted features types (see method retainFeature)
356 List<int[]> mapTo = new ArrayList<>();
357 mapTo.add(new int[] { 1, transcriptLength });
358 MapList mapping = new MapList(mappedFrom, mapTo, 1, 1);
359 EnsemblCdna cdna = new EnsemblCdna(getDomain());
360 cdna.transferFeatures(gene.getFeatures().getPositionalFeatures(),
361 transcript.getDatasetSequence(), mapping, parentId);
364 * fetch and save cross-references
366 cdna.getCrossReferences(transcript);
369 * and finally fetch the protein product and save as a cross-reference
371 cdna.addProteinProduct(transcript);
377 * Returns the 'transcript_id' property of the sequence feature (or null)
382 protected String getTranscriptId(SequenceFeature feature)
384 return (String) feature.getValue("transcript_id");
388 * Returns a list of the transcript features on the sequence whose Parent is
389 * the gene for the accession id.
391 * Transcript features are those of type "transcript", or any of its sub-types
392 * in the Sequence Ontology e.g. "mRNA", "processed_transcript". We also
393 * include "NMD_transcript_variant", because this type behaves like a
394 * transcript identifier in Ensembl, although strictly speaking it is not in
398 * @param geneSequence
401 protected List<SequenceFeature> getTranscriptFeatures(String accId,
402 SequenceI geneSequence)
404 List<SequenceFeature> transcriptFeatures = new ArrayList<>();
406 String parentIdentifier = GENE_PREFIX + accId;
408 List<SequenceFeature> sfs = geneSequence.getFeatures()
409 .getFeaturesByOntology(SequenceOntologyI.TRANSCRIPT);
410 sfs.addAll(geneSequence.getFeatures().getPositionalFeatures(
411 SequenceOntologyI.NMD_TRANSCRIPT_VARIANT));
413 for (SequenceFeature sf : sfs)
415 String parent = (String) sf.getValue(PARENT);
416 if (parentIdentifier.equals(parent))
418 transcriptFeatures.add(sf);
422 return transcriptFeatures;
426 public String getDescription()
428 return "Fetches all transcripts and variant features for a gene or transcript";
432 * Default test query is a gene id (can also enter a transcript id)
435 public String getTestQuery()
437 return "ENSG00000157764"; // BRAF, 5 transcripts, reverse strand
438 // ENSG00000090266 // NDUFB2, 15 transcripts, forward strand
439 // ENSG00000101812 // H2BFM histone, 3 transcripts, forward strand
440 // ENSG00000123569 // H2BFWT histone, 2 transcripts, reverse strand
444 * Answers true for a feature of type 'gene' (or a sub-type of gene in the
445 * Sequence Ontology), whose ID is the accession we are retrieving
448 protected boolean identifiesSequence(SequenceFeature sf, String accId)
450 if (SequenceOntologyFactory.getInstance().isA(sf.getType(),
451 SequenceOntologyI.GENE))
453 // NB features as gff use 'ID'; rest services return as 'id'
454 String id = (String) sf.getValue("ID");
455 if ((GENE_PREFIX + accId).equals(id))
464 * Answers true unless feature type is 'gene', or 'transcript' with a parent
465 * which is a different gene. We need the gene features to identify the range,
466 * but it is redundant information on the gene sequence. Checking the parent
467 * allows us to drop transcript features which belong to different
468 * (overlapping) genes.
471 protected boolean retainFeature(SequenceFeature sf, String accessionId)
473 SequenceOntologyI so = SequenceOntologyFactory.getInstance();
474 String type = sf.getType();
475 if (so.isA(type, SequenceOntologyI.GENE))
479 if (isTranscript(type))
481 String parent = (String) sf.getValue(PARENT);
482 if (!(GENE_PREFIX + accessionId).equals(parent))
491 * Answers false. This allows an optimisation - a single 'gene' feature is all
492 * that is needed to identify the positions of the gene on the genomic
496 protected boolean isSpliceable()
502 * Override to do nothing as Ensembl doesn't return a protein sequence for a
506 protected void addProteinProduct(SequenceI querySeq)
511 public Regex getAccessionValidator()
513 return ACCESSION_REGEX;
517 * Returns a descriptor for suitable feature display settings with
519 * <li>only exon or sequence_variant features (or their subtypes in the
520 * Sequence Ontology) visible</li>
521 * <li>variant features coloured red</li>
522 * <li>exon features coloured by label (exon name)</li>
523 * <li>variants displayed above (on top of) exons</li>
527 public FeatureSettingsModelI getFeatureColourScheme()
529 return new FeatureSettingsAdapter()
531 SequenceOntologyI so = SequenceOntologyFactory.getInstance();
534 public boolean isFeatureDisplayed(String type)
536 return (so.isA(type, SequenceOntologyI.EXON)
537 || so.isA(type, SequenceOntologyI.SEQUENCE_VARIANT));
541 public FeatureColourI getFeatureColour(String type)
543 if (so.isA(type, SequenceOntologyI.EXON))
545 return new FeatureColour()
548 public boolean isColourByLabel()
554 if (so.isA(type, SequenceOntologyI.SEQUENCE_VARIANT))
556 return new FeatureColour()
560 public Color getColour()
570 * order to render sequence_variant after exon after the rest
573 public int compare(String feature1, String feature2)
575 if (so.isA(feature1, SequenceOntologyI.SEQUENCE_VARIANT))
579 if (so.isA(feature2, SequenceOntologyI.SEQUENCE_VARIANT))
583 if (so.isA(feature1, SequenceOntologyI.EXON))
587 if (so.isA(feature2, SequenceOntologyI.EXON))