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.DBRefEntry;
27 import jalview.datamodel.GeneLociI;
28 import jalview.datamodel.Sequence;
29 import jalview.datamodel.SequenceFeature;
30 import jalview.datamodel.SequenceI;
31 import jalview.datamodel.features.SequenceFeatures;
32 import jalview.io.gff.SequenceOntologyFactory;
33 import jalview.io.gff.SequenceOntologyI;
34 import jalview.schemes.FeatureColour;
35 import jalview.schemes.FeatureSettingsAdapter;
36 import jalview.util.MapList;
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
54 private static final String GENE_PREFIX = "gene:";
57 * accepts anything as we will attempt lookup of gene or
58 * transcript id or gene name
60 private static final Regex ACCESSION_REGEX = new Regex(".*");
62 private static final EnsemblFeatureType[] FEATURES_TO_FETCH = {
63 EnsemblFeatureType.gene, EnsemblFeatureType.transcript,
64 EnsemblFeatureType.exon, EnsemblFeatureType.cds,
65 EnsemblFeatureType.variation };
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 * Returns an alignment containing the gene(s) for the given gene or
105 * transcript identifier, or external identifier (e.g. Uniprot id). If given a
106 * gene name or external identifier, returns any related gene sequences found
107 * for model organisms. If only a single gene is queried for, then its
108 * transcripts are also retrieved and added to the alignment. <br>
111 * <li>resolves a transcript identifier by looking up its parent gene id</li>
112 * <li>resolves an external identifier by looking up xref-ed gene ids</li>
113 * <li>fetches the gene sequence</li>
114 * <li>fetches features on the sequence</li>
115 * <li>identifies "transcript" features whose Parent is the requested
117 * <li>fetches the transcript sequence for each transcript</li>
118 * <li>makes a mapping from the gene to each transcript</li>
119 * <li>copies features from gene to transcript sequences</li>
120 * <li>fetches the protein sequence for each transcript, maps and saves it as
121 * a cross-reference</li>
122 * <li>aligns each transcript against the gene sequence based on the position
127 * a single gene or transcript identifier or gene name
128 * @return an alignment containing a gene, and possibly transcripts, or null
131 public AlignmentI getSequenceRecords(String query) throws Exception
134 * convert to a non-duplicated list of gene identifiers
136 List<String> geneIds = getGeneIds(query);
138 AlignmentI al = null;
139 for (String geneId : geneIds)
142 * fetch the gene sequence(s) with features and xrefs
144 AlignmentI geneAlignment = super.getSequenceRecords(geneId);
145 if (geneAlignment == null)
150 parseChromosomeLocations(geneAlignment);
152 if (geneAlignment.getHeight() == 1)
154 getTranscripts(geneAlignment, geneId);
162 al.append(geneAlignment);
169 * Parses and saves fields of an Ensembl-style description e.g.
170 * chromosome:GRCh38:17:45051610:45109016:1
174 private void parseChromosomeLocations(AlignmentI alignment)
176 for (SequenceI seq : alignment.getSequences())
178 String description = seq.getDescription();
179 if (description == null)
183 String[] tokens = description.split(":");
184 if (tokens.length == 6 && tokens[0].startsWith(DBRefEntry.CHROMOSOME))
186 String ref = tokens[1];
187 String chrom = tokens[2];
190 int chStart = Integer.parseInt(tokens[3]);
191 int chEnd = Integer.parseInt(tokens[4]);
192 boolean forwardStrand = "1".equals(tokens[5]);
193 String species = ""; // dunno yet!
194 int[] from = new int[] { seq.getStart(), seq.getEnd() };
195 int[] to = new int[] { forwardStrand ? chStart : chEnd,
196 forwardStrand ? chEnd : chStart };
197 MapList map = new MapList(from, to, 1, 1);
198 seq.setGeneLoci(species, ref, chrom, map);
199 } catch (NumberFormatException e)
201 System.err.println("Bad integers in description " + description);
208 * Converts a query, which may contain one or more gene or transcript
209 * identifiers, into a non-redundant list of gene identifiers.
214 List<String> getGeneIds(String accessions)
216 List<String> geneIds = new ArrayList<String>();
218 for (String acc : accessions.split(getAccessionSeparator()))
220 if (isGeneIdentifier(acc))
222 if (!geneIds.contains(acc))
229 * if given a transcript id, look up its gene parent
231 else if (isTranscriptIdentifier(acc))
233 String geneId = new EnsemblLookup(getDomain()).getParent(acc);
234 if (geneId != null && !geneIds.contains(geneId))
239 else if (isProteinIdentifier(acc))
241 String tscriptId = new EnsemblLookup(getDomain()).getParent(acc);
242 if (tscriptId != null)
244 String geneId = new EnsemblLookup(getDomain())
245 .getParent(tscriptId);
247 if (geneId != null && !geneIds.contains(geneId))
252 // NOTE - acc is lost if it resembles an ENS.+ ID but isn't actually
253 // resolving to one... e.g. ENSMICP00000009241
256 * if given a gene or other external name, lookup and fetch
257 * the corresponding gene for all model organisms
261 List<String> ids = new EnsemblSymbol(getDomain(), getDbSource(),
262 getDbVersion()).getIds(acc);
263 for (String geneId : ids)
265 if (!geneIds.contains(geneId))
276 * Attempts to get Ensembl stable identifiers for model organisms for a gene
277 * name by calling the xrefs symbol REST service to resolve the gene name.
282 protected String getGeneIdentifiersForName(String query)
284 List<String> ids = new EnsemblSymbol(getDomain(), getDbSource(),
285 getDbVersion()).getIds(query);
288 for (String id : ids)
290 if (isGeneIdentifier(id))
300 * Constructs all transcripts for the gene, as identified by "transcript"
301 * features whose Parent is the requested gene. The coding transcript
302 * sequences (i.e. with introns omitted) are added to the alignment.
308 protected void getTranscripts(AlignmentI al, String accId)
311 SequenceI gene = al.getSequenceAt(0);
312 List<SequenceFeature> transcriptFeatures = getTranscriptFeatures(accId,
315 for (SequenceFeature transcriptFeature : transcriptFeatures)
317 makeTranscript(transcriptFeature, al, gene);
320 clearGeneFeatures(gene);
324 * Remove unwanted features (transcript, exon, CDS) from the gene sequence
325 * after we have used them to derive transcripts and transfer features
329 protected void clearGeneFeatures(SequenceI gene)
332 * Note we include NMD_transcript_variant here because it behaves like
333 * 'transcript' in Ensembl, although strictly speaking it is not
334 * (it is a sub-type of sequence_variant)
336 String[] soTerms = new String[] {
337 SequenceOntologyI.NMD_TRANSCRIPT_VARIANT,
338 SequenceOntologyI.TRANSCRIPT, SequenceOntologyI.EXON,
339 SequenceOntologyI.CDS };
340 List<SequenceFeature> sfs = gene.getFeatures().getFeaturesByOntology(
342 for (SequenceFeature sf : sfs)
344 gene.deleteFeature(sf);
349 * Constructs a spliced transcript sequence by finding 'exon' features for the
350 * given id (or failing that 'CDS'). Copies features on to the new sequence.
351 * 'Aligns' the new sequence against the gene sequence by padding with gaps,
352 * and adds it to the alignment.
354 * @param transcriptFeature
356 * the alignment to which to add the new sequence
358 * the parent gene sequence, with features
361 SequenceI makeTranscript(SequenceFeature transcriptFeature, AlignmentI al,
364 String accId = getTranscriptId(transcriptFeature);
371 * NB we are mapping from gene sequence (not genome), so do not
372 * need to check for reverse strand (gene and transcript sequences
373 * are in forward sense)
377 * make a gene-length sequence filled with gaps
378 * we will fill in the bases for transcript regions
380 char[] seqChars = new char[gene.getLength()];
381 Arrays.fill(seqChars, al.getGapCharacter());
384 * look for exon features of the transcript, failing that for CDS
385 * (for example ENSG00000124610 has 1 CDS but no exon features)
387 String parentId = "transcript:" + accId;
388 List<SequenceFeature> splices = findFeatures(gene,
389 SequenceOntologyI.EXON, parentId);
390 if (splices.isEmpty())
392 splices = findFeatures(gene, SequenceOntologyI.CDS, parentId);
394 SequenceFeatures.sortFeatures(splices, true);
396 int transcriptLength = 0;
397 final char[] geneChars = gene.getSequence();
398 int offset = gene.getStart(); // to convert to 0-based positions
399 List<int[]> mappedFrom = new ArrayList<int[]>();
401 for (SequenceFeature sf : splices)
403 int start = sf.getBegin() - offset;
404 int end = sf.getEnd() - offset;
405 int spliceLength = end - start + 1;
406 System.arraycopy(geneChars, start, seqChars, start, spliceLength);
407 transcriptLength += spliceLength;
408 mappedFrom.add(new int[] { sf.getBegin(), sf.getEnd() });
411 Sequence transcript = new Sequence(accId, seqChars, 1,
415 * Ensembl has gene name as transcript Name
416 * EnsemblGenomes doesn't, but has a url-encoded description field
418 String description = (String) transcriptFeature.getValue(NAME);
419 if (description == null)
421 description = (String) transcriptFeature.getValue(DESCRIPTION);
423 if (description != null)
427 transcript.setDescription(URLDecoder.decode(description, "UTF-8"));
428 } catch (UnsupportedEncodingException e)
430 e.printStackTrace(); // as if
433 transcript.createDatasetSequence();
435 al.addSequence(transcript);
438 * transfer features to the new sequence; we use EnsemblCdna to do this,
439 * to filter out unwanted features types (see method retainFeature)
441 List<int[]> mapTo = new ArrayList<int[]>();
442 mapTo.add(new int[] { 1, transcriptLength });
443 MapList mapping = new MapList(mappedFrom, mapTo, 1, 1);
444 EnsemblCdna cdna = new EnsemblCdna(getDomain());
445 cdna.transferFeatures(gene.getFeatures().getPositionalFeatures(),
446 transcript.getDatasetSequence(), mapping, parentId);
448 mapTranscriptToChromosome(transcript, gene, mapping);
451 * fetch and save cross-references
453 cdna.getCrossReferences(transcript);
456 * and finally fetch the protein product and save as a cross-reference
458 cdna.addProteinProduct(transcript);
464 * If the gene has a mapping to chromosome coordinates, derive the transcript
465 * chromosome regions and save on the transcript sequence
470 * the mapping from gene to transcript positions
472 protected void mapTranscriptToChromosome(SequenceI transcript,
473 SequenceI gene, MapList mapping)
475 GeneLociI loci = gene.getGeneLoci();
481 MapList geneMapping = loci.getMap();
483 List<int[]> exons = mapping.getFromRanges();
484 List<int[]> transcriptLoci = new ArrayList<>();
486 for (int[] exon : exons)
488 transcriptLoci.add(geneMapping.locateInTo(exon[0], exon[1]));
491 List<int[]> transcriptRange = Arrays.asList(new int[] {
492 transcript.getStart(), transcript.getEnd() });
493 MapList mapList = new MapList(transcriptRange, transcriptLoci, 1, 1);
495 transcript.setGeneLoci(loci.getSpeciesId(), loci.getAssemblyId(),
496 loci.getChromosomeId(), mapList);
500 * Returns the 'transcript_id' property of the sequence feature (or null)
505 protected String getTranscriptId(SequenceFeature feature)
507 return (String) feature.getValue("transcript_id");
511 * Returns a list of the transcript features on the sequence whose Parent is
512 * the gene for the accession id.
515 * @param geneSequence
518 protected List<SequenceFeature> getTranscriptFeatures(String accId,
519 SequenceI geneSequence)
521 List<SequenceFeature> transcriptFeatures = new ArrayList<SequenceFeature>();
523 String parentIdentifier = GENE_PREFIX + accId;
524 // todo optimise here by transcript type!
525 List<SequenceFeature> sfs = geneSequence.getFeatures()
526 .getPositionalFeatures();
528 for (SequenceFeature sf : sfs)
530 if (isTranscript(sf.getType()))
532 String parent = (String) sf.getValue(PARENT);
533 if (parentIdentifier.equals(parent))
535 transcriptFeatures.add(sf);
540 return transcriptFeatures;
544 public String getDescription()
546 return "Fetches all transcripts and variant features for a gene or transcript";
550 * Default test query is a gene id (can also enter a transcript id)
553 public String getTestQuery()
555 return "ENSG00000157764"; // BRAF, 5 transcripts, reverse strand
556 // ENSG00000090266 // NDUFB2, 15 transcripts, forward strand
557 // ENSG00000101812 // H2BFM histone, 3 transcripts, forward strand
558 // ENSG00000123569 // H2BFWT histone, 2 transcripts, reverse strand
562 * Answers true for a feature of type 'gene' (or a sub-type of gene in the
563 * Sequence Ontology), whose ID is the accession we are retrieving
566 protected boolean identifiesSequence(SequenceFeature sf, String accId)
568 if (SequenceOntologyFactory.getInstance().isA(sf.getType(),
569 SequenceOntologyI.GENE))
571 String id = (String) sf.getValue(ID);
572 if ((GENE_PREFIX + accId).equals(id))
581 * Answers true unless feature type is 'gene', or 'transcript' with a parent
582 * which is a different gene. We need the gene features to identify the range,
583 * but it is redundant information on the gene sequence. Checking the parent
584 * allows us to drop transcript features which belong to different
585 * (overlapping) genes.
588 protected boolean retainFeature(SequenceFeature sf, String accessionId)
590 SequenceOntologyI so = SequenceOntologyFactory.getInstance();
591 String type = sf.getType();
592 if (so.isA(type, SequenceOntologyI.GENE))
596 if (isTranscript(type))
598 String parent = (String) sf.getValue(PARENT);
599 if (!(GENE_PREFIX + accessionId).equals(parent))
608 * Answers false. This allows an optimisation - a single 'gene' feature is all
609 * that is needed to identify the positions of the gene on the genomic
613 protected boolean isSpliceable()
619 * Override to do nothing as Ensembl doesn't return a protein sequence for a
623 protected void addProteinProduct(SequenceI querySeq)
628 public Regex getAccessionValidator()
630 return ACCESSION_REGEX;
634 * Returns a descriptor for suitable feature display settings with
636 * <li>only exon or sequence_variant features (or their subtypes in the
637 * Sequence Ontology) visible</li>
638 * <li>variant features coloured red</li>
639 * <li>exon features coloured by label (exon name)</li>
640 * <li>variants displayed above (on top of) exons</li>
644 public FeatureSettingsModelI getFeatureColourScheme()
646 return new FeatureSettingsAdapter()
648 SequenceOntologyI so = SequenceOntologyFactory.getInstance();
651 public boolean isFeatureDisplayed(String type)
653 return (so.isA(type, SequenceOntologyI.EXON)
654 || so.isA(type, SequenceOntologyI.SEQUENCE_VARIANT));
658 public FeatureColourI getFeatureColour(String type)
660 if (so.isA(type, SequenceOntologyI.EXON))
662 return new FeatureColour()
665 public boolean isColourByLabel()
671 if (so.isA(type, SequenceOntologyI.SEQUENCE_VARIANT))
673 return new FeatureColour()
677 public Color getColour()
687 * order to render sequence_variant after exon after the rest
690 public int compare(String feature1, String feature2)
692 if (so.isA(feature1, SequenceOntologyI.SEQUENCE_VARIANT))
696 if (so.isA(feature2, SequenceOntologyI.SEQUENCE_VARIANT))
700 if (so.isA(feature1, SequenceOntologyI.EXON))
704 if (so.isA(feature2, SequenceOntologyI.EXON))