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.GeneLociI;
27 import jalview.datamodel.Sequence;
28 import jalview.datamodel.SequenceFeature;
29 import jalview.datamodel.SequenceI;
30 import jalview.datamodel.features.SequenceFeatures;
31 import jalview.io.gff.SequenceOntologyFactory;
32 import jalview.io.gff.SequenceOntologyI;
33 import jalview.schemes.FeatureColour;
34 import jalview.schemes.FeatureSettingsAdapter;
35 import jalview.util.MapList;
36 import jalview.util.Platform;
38 import java.awt.Color;
39 import java.io.UnsupportedEncodingException;
40 import java.net.URLDecoder;
41 import java.util.ArrayList;
42 import java.util.Arrays;
43 import java.util.List;
45 import com.stevesoft.pat.Regex;
48 * A class that fetches genomic sequence and all transcripts for an Ensembl gene
52 public class EnsemblGene extends EnsemblSeqProxy
55 * accepts anything as we will attempt lookup of gene or
56 * transcript id or gene name
58 private static final Regex ACCESSION_REGEX = Platform.newRegex(".*");
60 private static final EnsemblFeatureType[] FEATURES_TO_FETCH = {
61 EnsemblFeatureType.gene, EnsemblFeatureType.transcript,
62 EnsemblFeatureType.exon, EnsemblFeatureType.cds,
63 EnsemblFeatureType.variation };
65 private static final String CHROMOSOME = "chromosome";
68 * Default constructor (to use rest.ensembl.org)
76 * Constructor given the target domain to fetch data from
80 public EnsemblGene(String d)
86 public String getDbName()
92 protected EnsemblFeatureType[] getFeaturesToFetch()
94 return FEATURES_TO_FETCH;
98 protected EnsemblSeqType getSourceEnsemblType()
100 return EnsemblSeqType.GENOMIC;
104 protected String getObjectType()
106 return OBJECT_TYPE_GENE;
110 * Returns an alignment containing the gene(s) for the given gene or
111 * transcript identifier, or external identifier (e.g. Uniprot id). If given a
112 * gene name or external identifier, returns any related gene sequences found
113 * for model organisms. If only a single gene is queried for, then its
114 * transcripts are also retrieved and added to the alignment. <br>
117 * <li>resolves a transcript identifier by looking up its parent gene id</li>
118 * <li>resolves an external identifier by looking up xref-ed gene ids</li>
119 * <li>fetches the gene sequence</li>
120 * <li>fetches features on the sequence</li>
121 * <li>identifies "transcript" features whose Parent is the requested
123 * <li>fetches the transcript sequence for each transcript</li>
124 * <li>makes a mapping from the gene to each transcript</li>
125 * <li>copies features from gene to transcript sequences</li>
126 * <li>fetches the protein sequence for each transcript, maps and saves it as
127 * a cross-reference</li>
128 * <li>aligns each transcript against the gene sequence based on the position
133 * a single gene or transcript identifier or gene name
134 * @return an alignment containing a gene, and possibly transcripts, or null
137 public AlignmentI getSequenceRecords(String query) throws Exception
140 * convert to a non-duplicated list of gene identifiers
142 List<String> geneIds = getGeneIds(query);
143 AlignmentI al = null;
144 for (String geneId : geneIds)
147 * fetch the gene sequence(s) with features and xrefs
149 AlignmentI geneAlignment = super.getSequenceRecords(geneId);
150 if (geneAlignment == null)
155 if (geneAlignment.getHeight() == 1)
157 // ensure id has 'correct' case for the Ensembl identifier
158 geneId = geneAlignment.getSequenceAt(0).getName();
159 findGeneLoci(geneAlignment.getSequenceAt(0), geneId);
160 getTranscripts(geneAlignment, geneId);
168 al.append(geneAlignment);
175 * Calls the /lookup/id REST service, parses the response for gene
176 * coordinates, and if successful, adds these to the sequence. If this fails,
177 * fall back on trying to parse the sequence description in case it is in
178 * Ensembl-gene format e.g. chromosome:GRCh38:17:45051610:45109016:1.
183 void findGeneLoci(SequenceI seq, String geneId)
185 GeneLociI geneLoci = new EnsemblLookup(getDomain()).getGeneLoci(geneId);
186 if (geneLoci != null)
188 seq.setGeneLoci(geneLoci.getSpeciesId(), geneLoci.getAssemblyId(),
189 geneLoci.getChromosomeId(), geneLoci.getMapping());
193 parseChromosomeLocations(seq);
198 * Parses and saves fields of an Ensembl-style description e.g.
199 * chromosome:GRCh38:17:45051610:45109016:1
203 boolean parseChromosomeLocations(SequenceI seq)
205 String description = seq.getDescription();
206 if (description == null)
210 String[] tokens = description.split(":");
211 if (tokens.length == 6 && tokens[0].startsWith(CHROMOSOME))
213 String ref = tokens[1];
214 String chrom = tokens[2];
217 int chStart = Integer.parseInt(tokens[3]);
218 int chEnd = Integer.parseInt(tokens[4]);
219 boolean forwardStrand = "1".equals(tokens[5]);
220 String species = ""; // not known here
221 int[] from = new int[] { seq.getStart(), seq.getEnd() };
222 int[] to = new int[] { forwardStrand ? chStart : chEnd,
223 forwardStrand ? chEnd : chStart };
224 MapList map = new MapList(from, to, 1, 1);
225 seq.setGeneLoci(species, ref, chrom, map);
227 } catch (NumberFormatException e)
229 System.err.println("Bad integers in description " + description);
236 * Converts a query, which may contain one or more gene, transcript, or
237 * external (to Ensembl) identifiers, into a non-redundant list of gene
243 List<String> getGeneIds(String accessions)
245 List<String> geneIds = new ArrayList<>();
247 for (String acc : accessions.split(getAccessionSeparator()))
250 * First try lookup as an Ensembl (gene or transcript) identifier
252 String geneId = new EnsemblLookup(getDomain()).getGeneId(acc);
255 if (!geneIds.contains(geneId))
263 * if given a gene or other external name, lookup and fetch
264 * the corresponding gene for all model organisms
266 List<String> ids = new EnsemblSymbol(getDomain(), getDbSource(),
267 getDbVersion()).getGeneIds(acc);
268 for (String id : ids)
270 if (!geneIds.contains(id))
281 * Constructs all transcripts for the gene, as identified by "transcript"
282 * features whose Parent is the requested gene. The coding transcript
283 * sequences (i.e. with introns omitted) are added to the alignment.
289 protected void getTranscripts(AlignmentI al, String accId)
292 SequenceI gene = al.getSequenceAt(0);
293 List<SequenceFeature> transcriptFeatures = getTranscriptFeatures(accId,
296 for (SequenceFeature transcriptFeature : transcriptFeatures)
298 makeTranscript(transcriptFeature, al, gene);
301 clearGeneFeatures(gene);
305 * Remove unwanted features (transcript, exon, CDS) from the gene sequence
306 * after we have used them to derive transcripts and transfer features
310 protected void clearGeneFeatures(SequenceI gene)
313 * Note we include NMD_transcript_variant here because it behaves like
314 * 'transcript' in Ensembl, although strictly speaking it is not
315 * (it is a sub-type of sequence_variant)
317 String[] soTerms = new String[] {
318 SequenceOntologyI.NMD_TRANSCRIPT_VARIANT,
319 SequenceOntologyI.TRANSCRIPT, SequenceOntologyI.EXON,
320 SequenceOntologyI.CDS };
321 List<SequenceFeature> sfs = gene.getFeatures().getFeaturesByOntology(
323 for (SequenceFeature sf : sfs)
325 gene.deleteFeature(sf);
330 * Constructs a spliced transcript sequence by finding 'exon' features for the
331 * given id (or failing that 'CDS'). Copies features on to the new sequence.
332 * 'Aligns' the new sequence against the gene sequence by padding with gaps,
333 * and adds it to the alignment.
335 * @param transcriptFeature
337 * the alignment to which to add the new sequence
339 * the parent gene sequence, with features
342 SequenceI makeTranscript(SequenceFeature transcriptFeature, AlignmentI al,
345 String accId = getTranscriptId(transcriptFeature);
352 * NB we are mapping from gene sequence (not genome), so do not
353 * need to check for reverse strand (gene and transcript sequences
354 * are in forward sense)
358 * make a gene-length sequence filled with gaps
359 * we will fill in the bases for transcript regions
361 char[] seqChars = new char[gene.getLength()];
362 Arrays.fill(seqChars, al.getGapCharacter());
365 * look for exon features of the transcript, failing that for CDS
366 * (for example ENSG00000124610 has 1 CDS but no exon features)
368 String parentId = accId;
369 List<SequenceFeature> splices = findFeatures(gene,
370 SequenceOntologyI.EXON, parentId);
371 if (splices.isEmpty())
373 splices = findFeatures(gene, SequenceOntologyI.CDS, parentId);
375 SequenceFeatures.sortFeatures(splices, true);
377 int transcriptLength = 0;
378 final char[] geneChars = gene.getSequence();
379 int offset = gene.getStart(); // to convert to 0-based positions
380 List<int[]> mappedFrom = new ArrayList<>();
382 for (SequenceFeature sf : splices)
384 int start = sf.getBegin() - offset;
385 int end = sf.getEnd() - offset;
386 int spliceLength = end - start + 1;
387 System.arraycopy(geneChars, start, seqChars, start, spliceLength);
388 transcriptLength += spliceLength;
389 mappedFrom.add(new int[] { sf.getBegin(), sf.getEnd() });
392 Sequence transcript = new Sequence(accId, seqChars, 1,
396 * Ensembl has gene name as transcript Name
397 * EnsemblGenomes doesn't, but has a url-encoded description field
399 String description = transcriptFeature.getDescription();
400 if (description == null)
402 description = (String) transcriptFeature.getValue(DESCRIPTION);
404 if (description != null)
408 transcript.setDescription(URLDecoder.decode(description, "UTF-8"));
409 } catch (UnsupportedEncodingException e)
411 e.printStackTrace(); // as if
414 transcript.createDatasetSequence();
416 al.addSequence(transcript);
419 * transfer features to the new sequence; we use EnsemblCdna to do this,
420 * to filter out unwanted features types (see method retainFeature)
422 List<int[]> mapTo = new ArrayList<>();
423 mapTo.add(new int[] { 1, transcriptLength });
424 MapList mapping = new MapList(mappedFrom, mapTo, 1, 1);
425 EnsemblCdna cdna = new EnsemblCdna(getDomain());
426 cdna.transferFeatures(gene.getFeatures().getPositionalFeatures(),
427 transcript.getDatasetSequence(), mapping, parentId);
429 mapTranscriptToChromosome(transcript, gene, mapping);
432 * fetch and save cross-references
434 cdna.getCrossReferences(transcript);
437 * and finally fetch the protein product and save as a cross-reference
439 cdna.addProteinProduct(transcript);
445 * If the gene has a mapping to chromosome coordinates, derive the transcript
446 * chromosome regions and save on the transcript sequence
451 * the mapping from gene to transcript positions
453 protected void mapTranscriptToChromosome(SequenceI transcript,
454 SequenceI gene, MapList mapping)
456 GeneLociI loci = gene.getGeneLoci();
462 MapList geneMapping = loci.getMapping();
464 List<int[]> exons = mapping.getFromRanges();
465 List<int[]> transcriptLoci = new ArrayList<>();
467 for (int[] exon : exons)
469 transcriptLoci.add(geneMapping.locateInTo(exon[0], exon[1]));
472 List<int[]> transcriptRange = Arrays.asList(new int[] {
473 transcript.getStart(), transcript.getEnd() });
474 MapList mapList = new MapList(transcriptRange, transcriptLoci, 1, 1);
476 transcript.setGeneLoci(loci.getSpeciesId(), loci.getAssemblyId(),
477 loci.getChromosomeId(), mapList);
481 * Returns the 'transcript_id' property of the sequence feature (or null)
486 protected String getTranscriptId(SequenceFeature feature)
488 return (String) feature.getValue(JSON_ID);
492 * Returns a list of the transcript features on the sequence whose Parent is
493 * the gene for the accession id.
495 * Transcript features are those of type "transcript", or any of its sub-types
496 * in the Sequence Ontology e.g. "mRNA", "processed_transcript". We also
497 * include "NMD_transcript_variant", because this type behaves like a
498 * transcript identifier in Ensembl, although strictly speaking it is not in
502 * @param geneSequence
505 protected List<SequenceFeature> getTranscriptFeatures(String accId,
506 SequenceI geneSequence)
508 List<SequenceFeature> transcriptFeatures = new ArrayList<>();
510 String parentIdentifier = accId;
512 List<SequenceFeature> sfs = geneSequence.getFeatures()
513 .getFeaturesByOntology(SequenceOntologyI.TRANSCRIPT);
514 sfs.addAll(geneSequence.getFeatures().getPositionalFeatures(
515 SequenceOntologyI.NMD_TRANSCRIPT_VARIANT));
517 for (SequenceFeature sf : sfs)
519 String parent = (String) sf.getValue(PARENT);
520 if (parentIdentifier.equalsIgnoreCase(parent))
522 transcriptFeatures.add(sf);
526 return transcriptFeatures;
530 public String getDescription()
532 return "Fetches all transcripts and variant features for a gene or transcript";
536 * Default test query is a gene id (can also enter a transcript id)
539 public String getTestQuery()
541 return Platform.isJS() ? "ENSG00000123569" : "ENSG00000157764";
542 // ENSG00000123569 // H2BFWT histone, 2 transcripts, reverse strand
543 // ENSG00000157764 // BRAF, 5 transcripts, reverse strand
544 // ENSG00000090266 // NDUFB2, 15 transcripts, forward strand
545 // ENSG00000101812 // H2BFM histone, 3 transcripts, forward strand
549 * Answers a list of sequence features (if any) whose type is 'gene' (or a
550 * subtype of gene in the Sequence Ontology), and whose ID is the accession we
554 protected List<SequenceFeature> getIdentifyingFeatures(SequenceI seq,
557 List<SequenceFeature> result = new ArrayList<>();
558 List<SequenceFeature> sfs = seq.getFeatures()
559 .getFeaturesByOntology(SequenceOntologyI.GENE);
560 for (SequenceFeature sf : sfs)
562 String id = (String) sf.getValue(JSON_ID);
563 if (accId.equalsIgnoreCase(id))
572 * Answers true unless feature type is 'gene', or 'transcript' with a parent
573 * which is a different gene. We need the gene features to identify the range,
574 * but it is redundant information on the gene sequence. Checking the parent
575 * allows us to drop transcript features which belong to different
576 * (overlapping) genes.
579 protected boolean retainFeature(SequenceFeature sf, String accessionId)
581 SequenceOntologyI so = SequenceOntologyFactory.getSequenceOntology();
582 String type = sf.getType();
583 if (so.isA(type, SequenceOntologyI.GENE))
587 if (isTranscript(type))
589 String parent = (String) sf.getValue(PARENT);
590 if (!accessionId.equalsIgnoreCase(parent))
599 * Override to do nothing as Ensembl doesn't return a protein sequence for a
603 protected void addProteinProduct(SequenceI querySeq)
608 public Regex getAccessionValidator()
610 return ACCESSION_REGEX;
614 * Returns a descriptor for suitable feature display settings with
616 * <li>only exon or sequence_variant features (or their subtypes in the
617 * Sequence Ontology) visible</li>
618 * <li>variant features coloured red</li>
619 * <li>exon features coloured by label (exon name)</li>
620 * <li>variants displayed above (on top of) exons</li>
624 public FeatureSettingsModelI getFeatureColourScheme()
626 return new FeatureSettingsAdapter()
628 SequenceOntologyI so = SequenceOntologyFactory.getSequenceOntology();
631 public boolean isFeatureHidden(String type)
633 return (!so.isA(type, SequenceOntologyI.EXON)
634 && !so.isA(type, SequenceOntologyI.SEQUENCE_VARIANT));
638 public FeatureColourI getFeatureColour(String type)
640 if (so.isA(type, SequenceOntologyI.EXON))
642 return new FeatureColour()
645 public boolean isColourByLabel()
651 if (so.isA(type, SequenceOntologyI.SEQUENCE_VARIANT))
653 return new FeatureColour()
657 public Color getColour()
667 * order to render sequence_variant after exon after the rest
670 public int compare(String feature1, String feature2)
672 if (so.isA(feature1, SequenceOntologyI.SEQUENCE_VARIANT))
676 if (so.isA(feature2, SequenceOntologyI.SEQUENCE_VARIANT))
680 if (so.isA(feature1, SequenceOntologyI.EXON))
684 if (so.isA(feature2, SequenceOntologyI.EXON))