1 package jalview.ext.ensembl;
3 import jalview.api.FeatureColourI;
4 import jalview.api.FeatureSettingsModelI;
5 import jalview.datamodel.AlignmentI;
6 import jalview.datamodel.Sequence;
7 import jalview.datamodel.SequenceFeature;
8 import jalview.datamodel.SequenceI;
9 import jalview.io.gff.SequenceOntologyFactory;
10 import jalview.io.gff.SequenceOntologyI;
11 import jalview.schemes.FeatureColour;
12 import jalview.schemes.FeatureSettingsAdapter;
13 import jalview.util.MapList;
15 import java.awt.Color;
16 import java.io.UnsupportedEncodingException;
17 import java.net.URLDecoder;
18 import java.util.ArrayList;
19 import java.util.Arrays;
20 import java.util.List;
22 import com.stevesoft.pat.Regex;
25 * A class that fetches genomic sequence and all transcripts for an Ensembl gene
29 public class EnsemblGene extends EnsemblSeqProxy
31 private static final String GENE_PREFIX = "gene:";
34 * accepts anything as we will attempt lookup of gene or
35 * transcript id or gene name
37 private static final Regex ACCESSION_REGEX = new Regex(".*");
39 private static final EnsemblFeatureType[] FEATURES_TO_FETCH = {
40 EnsemblFeatureType.gene, EnsemblFeatureType.transcript,
41 EnsemblFeatureType.exon, EnsemblFeatureType.cds,
42 EnsemblFeatureType.variation };
45 * Default constructor (to use rest.ensembl.org)
53 * Constructor given the target domain to fetch data from
57 public EnsemblGene(String d)
63 public String getDbName()
69 protected EnsemblFeatureType[] getFeaturesToFetch()
71 return FEATURES_TO_FETCH;
75 protected EnsemblSeqType getSourceEnsemblType()
77 return EnsemblSeqType.GENOMIC;
81 * Returns an alignment containing the gene(s) for the given gene or
82 * transcript identifier, or external identifier (e.g. Uniprot id). If given a
83 * gene name or external identifier, returns any related gene sequences found
84 * for model organisms. If only a single gene is queried for, then its
85 * transcripts are also retrieved and added to the alignment. <br>
88 * <li>resolves a transcript identifier by looking up its parent gene id</li>
89 * <li>resolves an external identifier by looking up xref-ed gene ids</li>
90 * <li>fetches the gene sequence</li>
91 * <li>fetches features on the sequence</li>
92 * <li>identifies "transcript" features whose Parent is the requested gene</li>
93 * <li>fetches the transcript sequence for each transcript</li>
94 * <li>makes a mapping from the gene to each transcript</li>
95 * <li>copies features from gene to transcript sequences</li>
96 * <li>fetches the protein sequence for each transcript, maps and saves it as
97 * a cross-reference</li>
98 * <li>aligns each transcript against the gene sequence based on the position
103 * a single gene or transcript identifier or gene name
104 * @return an alignment containing a gene, and possibly transcripts, or null
107 public AlignmentI getSequenceRecords(String query) throws Exception
110 * convert to a non-duplicated list of gene identifiers
112 List<String> geneIds = getGeneIds(query);
114 AlignmentI al = null;
115 for (String geneId : geneIds)
118 * fetch the gene sequence(s) with features and xrefs
120 AlignmentI geneAlignment = super.getSequenceRecords(geneId);
122 if (geneAlignment.getHeight() == 1)
124 getTranscripts(geneAlignment, geneId);
132 al.append(geneAlignment);
139 * Converts a query, which may contain one or more gene or transcript
140 * identifiers, into a non-redundant list of gene identifiers.
145 List<String> getGeneIds(String accessions)
147 List<String> geneIds = new ArrayList<String>();
149 for (String acc : accessions.split(getAccessionSeparator()))
151 if (isGeneIdentifier(acc))
153 if (!geneIds.contains(acc))
160 * if given a transcript id, look up its gene parent
162 else if (isTranscriptIdentifier(acc))
164 String geneId = new EnsemblLookup(getDomain()).getParent(acc);
165 if (geneId != null && !geneIds.contains(geneId))
172 * if given a gene or other external name, lookup and fetch
173 * the corresponding gene for all model organisms
177 List<String> ids = new EnsemblSymbol(getDomain(), getDbSource(),
178 getDbVersion()).getIds(acc);
179 for (String geneId : ids)
181 if (!geneIds.contains(geneId))
192 * Attempts to get Ensembl stable identifiers for model organisms for a gene
193 * name by calling the xrefs symbol REST service to resolve the gene name.
198 protected String getGeneIdentifiersForName(String query)
200 List<String> ids = new EnsemblSymbol(getDomain(), getDbSource(),
201 getDbVersion()).getIds(query);
204 for (String id : ids)
206 if (isGeneIdentifier(id))
216 * Constructs all transcripts for the gene, as identified by "transcript"
217 * features whose Parent is the requested gene. The coding transcript
218 * sequences (i.e. with introns omitted) are added to the alignment.
224 protected void getTranscripts(AlignmentI al, String accId)
227 SequenceI gene = al.getSequenceAt(0);
228 List<SequenceFeature> transcriptFeatures = getTranscriptFeatures(accId,
231 for (SequenceFeature transcriptFeature : transcriptFeatures)
233 makeTranscript(transcriptFeature, al, gene);
236 clearGeneFeatures(gene);
240 * Remove unwanted features (transcript, exon, CDS) from the gene sequence
241 * after we have used them to derive transcripts and transfer features
245 protected void clearGeneFeatures(SequenceI gene)
247 SequenceFeature[] sfs = gene.getSequenceFeatures();
250 SequenceOntologyI so = SequenceOntologyFactory.getInstance();
251 List<SequenceFeature> filtered = new ArrayList<SequenceFeature>();
252 for (SequenceFeature sf : sfs)
254 String type = sf.getType();
255 if (!isTranscript(type) && !so.isA(type, SequenceOntologyI.EXON)
256 && !so.isA(type, SequenceOntologyI.CDS))
261 gene.setSequenceFeatures(filtered
262 .toArray(new SequenceFeature[filtered
268 * Constructs a spliced transcript sequence by finding 'exon' features for the
269 * given id (or failing that 'CDS'). Copies features on to the new sequence.
270 * 'Aligns' the new sequence against the gene sequence by padding with gaps,
271 * and adds it to the alignment.
273 * @param transcriptFeature
275 * the alignment to which to add the new sequence
277 * the parent gene sequence, with features
280 SequenceI makeTranscript(SequenceFeature transcriptFeature,
281 AlignmentI al, SequenceI gene)
283 String accId = getTranscriptId(transcriptFeature);
290 * NB we are mapping from gene sequence (not genome), so do not
291 * need to check for reverse strand (gene and transcript sequences
292 * are in forward sense)
296 * make a gene-length sequence filled with gaps
297 * we will fill in the bases for transcript regions
299 char[] seqChars = new char[gene.getLength()];
300 Arrays.fill(seqChars, al.getGapCharacter());
303 * look for exon features of the transcript, failing that for CDS
304 * (for example ENSG00000124610 has 1 CDS but no exon features)
306 String parentId = "transcript:" + accId;
307 List<SequenceFeature> splices = findFeatures(gene,
308 SequenceOntologyI.EXON, parentId);
309 if (splices.isEmpty())
311 splices = findFeatures(gene, SequenceOntologyI.CDS, parentId);
314 int transcriptLength = 0;
315 final char[] geneChars = gene.getSequence();
316 int offset = gene.getStart(); // to convert to 0-based positions
317 List<int[]> mappedFrom = new ArrayList<int[]>();
319 for (SequenceFeature sf : splices)
321 int start = sf.getBegin() - offset;
322 int end = sf.getEnd() - offset;
323 int spliceLength = end - start + 1;
324 System.arraycopy(geneChars, start, seqChars, start, spliceLength);
325 transcriptLength += spliceLength;
326 mappedFrom.add(new int[] { sf.getBegin(), sf.getEnd() });
329 Sequence transcript = new Sequence(accId, seqChars, 1, transcriptLength);
332 * Ensembl has gene name as transcript Name
333 * EnsemblGenomes doesn't, but has a url-encoded description field
335 String description = (String) transcriptFeature.getValue(NAME);
336 if (description == null)
338 description = (String) transcriptFeature.getValue(DESCRIPTION);
340 if (description != null)
344 transcript.setDescription(URLDecoder.decode(description, "UTF-8"));
345 } catch (UnsupportedEncodingException e)
347 e.printStackTrace(); // as if
350 transcript.createDatasetSequence();
352 al.addSequence(transcript);
355 * transfer features to the new sequence; we use EnsemblCdna to do this,
356 * to filter out unwanted features types (see method retainFeature)
358 List<int[]> mapTo = new ArrayList<int[]>();
359 mapTo.add(new int[] { 1, transcriptLength });
360 MapList mapping = new MapList(mappedFrom, mapTo, 1, 1);
361 EnsemblCdna cdna = new EnsemblCdna(getDomain());
362 cdna.transferFeatures(gene.getSequenceFeatures(),
363 transcript.getDatasetSequence(), mapping, parentId);
366 * fetch and save cross-references
368 cdna.getCrossReferences(transcript);
371 * and finally fetch the protein product and save as a cross-reference
373 cdna.addProteinProduct(transcript);
379 * Returns the 'transcript_id' property of the sequence feature (or null)
384 protected String getTranscriptId(SequenceFeature feature)
386 return (String) feature.getValue("transcript_id");
390 * Returns a list of the transcript features on the sequence whose Parent is
391 * the gene for the accession id.
394 * @param geneSequence
397 protected List<SequenceFeature> getTranscriptFeatures(String accId,
398 SequenceI geneSequence)
400 List<SequenceFeature> transcriptFeatures = new ArrayList<SequenceFeature>();
402 String parentIdentifier = GENE_PREFIX + accId;
403 SequenceFeature[] sfs = geneSequence.getSequenceFeatures();
407 for (SequenceFeature sf : sfs)
409 if (isTranscript(sf.getType()))
411 String parent = (String) sf.getValue(PARENT);
412 if (parentIdentifier.equals(parent))
414 transcriptFeatures.add(sf);
420 return transcriptFeatures;
424 public String getDescription()
426 return "Fetches all transcripts and variant features for a gene or transcript";
430 * Default test query is a gene id (can also enter a transcript id)
433 public String getTestQuery()
435 return "ENSG00000157764"; // BRAF, 5 transcripts, reverse strand
436 // ENSG00000090266 // NDUFB2, 15 transcripts, forward strand
437 // ENSG00000101812 // H2BFM histone, 3 transcripts, forward strand
438 // ENSG00000123569 // H2BFWT histone, 2 transcripts, reverse strand
442 * Answers true for a feature of type 'gene' (or a sub-type of gene in the
443 * Sequence Ontology), whose ID is the accession we are retrieving
446 protected boolean identifiesSequence(SequenceFeature sf, String accId)
448 if (SequenceOntologyFactory.getInstance().isA(sf.getType(),
449 SequenceOntologyI.GENE))
451 String id = (String) sf.getValue(ID);
452 if ((GENE_PREFIX + accId).equals(id))
461 * Answers true unless feature type is 'gene', or 'transcript' with a parent
462 * which is a different gene. We need the gene features to identify the range,
463 * but it is redundant information on the gene sequence. Checking the parent
464 * allows us to drop transcript features which belong to different
465 * (overlapping) genes.
468 protected boolean retainFeature(SequenceFeature sf, String accessionId)
470 SequenceOntologyI so = SequenceOntologyFactory.getInstance();
471 String type = sf.getType();
472 if (so.isA(type, SequenceOntologyI.GENE))
476 if (isTranscript(type))
478 String parent = (String) sf.getValue(PARENT);
479 if (!(GENE_PREFIX + accessionId).equals(parent))
488 * Answers false. This allows an optimisation - a single 'gene' feature is all
489 * that is needed to identify the positions of the gene on the genomic
493 protected boolean isSpliceable()
499 * Override to do nothing as Ensembl doesn't return a protein sequence for a
503 protected void addProteinProduct(SequenceI querySeq)
508 public Regex getAccessionValidator()
510 return ACCESSION_REGEX;
514 * Returns a descriptor for suitable feature display settings with
516 * <li>only exon or sequence_variant features (or their subtypes in the
517 * Sequence Ontology) visible</li>
518 * <li>variant features coloured red</li>
519 * <li>exon features coloured by label (exon name)</li>
520 * <li>variants displayed above (on top of) exons</li>
524 public FeatureSettingsModelI getFeatureColourScheme()
526 return new FeatureSettingsAdapter()
528 SequenceOntologyI so = SequenceOntologyFactory.getInstance();
530 public boolean isFeatureDisplayed(String type)
532 return (so.isA(type, SequenceOntologyI.EXON) || so.isA(type,
533 SequenceOntologyI.SEQUENCE_VARIANT));
537 public FeatureColourI getFeatureColour(String type)
539 if (so.isA(type, SequenceOntologyI.EXON))
541 return new FeatureColour()
544 public boolean isColourByLabel()
550 if (so.isA(type, SequenceOntologyI.SEQUENCE_VARIANT))
552 return new FeatureColour()
556 public Color getColour()
566 * order to render sequence_variant after exon after the rest
569 public int compare(String feature1, String feature2)
571 if (so.isA(feature1, SequenceOntologyI.SEQUENCE_VARIANT))
575 if (so.isA(feature2, SequenceOntologyI.SEQUENCE_VARIANT))
579 if (so.isA(feature1, SequenceOntologyI.EXON))
583 if (so.isA(feature2, SequenceOntologyI.EXON))