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.FeatureColourAdapter;
12 import jalview.schemes.FeatureSettingsAdapter;
13 import jalview.util.MapList;
14 import jalview.util.StringUtils;
16 import java.awt.Color;
17 import java.io.UnsupportedEncodingException;
18 import java.net.URLDecoder;
19 import java.util.ArrayList;
20 import java.util.Arrays;
21 import java.util.List;
23 import com.stevesoft.pat.Regex;
26 * A class that fetches genomic sequence and all transcripts for an Ensembl gene
30 public class EnsemblGene extends EnsemblSeqProxy
32 private static final String GENE_PREFIX = "gene:";
35 * accepts anything as we will attempt lookup of gene or
36 * transcript id or gene name
38 private static final Regex ACCESSION_REGEX = new Regex(".*");
40 private static final EnsemblFeatureType[] FEATURES_TO_FETCH = {
41 EnsemblFeatureType.gene, EnsemblFeatureType.transcript,
42 EnsemblFeatureType.exon, EnsemblFeatureType.cds,
43 EnsemblFeatureType.variation };
46 * Default constructor (to use rest.ensembl.org)
54 * Constructor given the target domain to fetch data from
58 public EnsemblGene(String d)
64 public String getDbName()
70 protected EnsemblFeatureType[] getFeaturesToFetch()
72 return FEATURES_TO_FETCH;
76 protected EnsemblSeqType getSourceEnsemblType()
78 return EnsemblSeqType.GENOMIC;
82 * Returns an alignment containing the gene(s) for the given gene or
83 * transcript identifier, or external identifier (e.g. Uniprot id). If given a
84 * gene name or external identifier, returns any related gene sequences found
85 * for model organisms. If only a single gene is queried for, then its
86 * transcripts are also retrieved and added to the alignment. <br>
89 * <li>resolves a transcript identifier by looking up its parent gene id</li>
90 * <li>resolves an external identifier by looking up xref-ed gene ids</li>
91 * <li>fetches the gene sequence</li>
92 * <li>fetches features on the sequence</li>
93 * <li>identifies "transcript" features whose Parent is the requested gene</li>
94 * <li>fetches the transcript sequence for each transcript</li>
95 * <li>makes a mapping from the gene to each transcript</li>
96 * <li>copies features from gene to transcript sequences</li>
97 * <li>fetches the protein sequence for each transcript, maps and saves it as
98 * a cross-reference</li>
99 * <li>aligns each transcript against the gene sequence based on the position
104 * a single gene or transcript identifier or gene name
105 * @return an alignment containing a gene, and possibly transcripts, or null
108 public AlignmentI getSequenceRecords(String query) throws Exception
111 * if given a transcript id, look up its gene parent
113 if (isTranscriptIdentifier(query))
115 query = new EnsemblLookup(getDomain()).getParent(query);
123 * if given a gene or other external name, lookup and fetch
124 * the corresponding gene for all model organisms
126 if (!isGeneIdentifier(query))
128 List<String> geneIds = new EnsemblSymbol(getDomain()).getIds(query);
129 if (geneIds.isEmpty())
133 String theIds = StringUtils.listToDelimitedString(geneIds,
134 getAccessionSeparator());
135 return getSequenceRecords(theIds);
139 * fetch the gene sequence(s) with features and xrefs
141 AlignmentI al = super.getSequenceRecords(query);
144 * if we retrieved a single gene, get its transcripts as well
146 if (al.getHeight() == 1)
148 getTranscripts(al, query);
155 * Attempts to get Ensembl stable identifiers for model organisms for a gene
156 * name by calling the xrefs symbol REST service to resolve the gene name.
161 protected String getGeneIdentifiersForName(String query)
163 List<String> ids = new EnsemblSymbol(getDomain()).getIds(query);
166 for (String id : ids)
168 if (isGeneIdentifier(id))
178 * Constructs all transcripts for the gene, as identified by "transcript"
179 * features whose Parent is the requested gene. The coding transcript
180 * sequences (i.e. with introns omitted) are added to the alignment.
186 protected void getTranscripts(AlignmentI al, String accId)
189 SequenceI gene = al.getSequenceAt(0);
190 List<SequenceFeature> transcriptFeatures = getTranscriptFeatures(accId,
193 for (SequenceFeature transcriptFeature : transcriptFeatures)
195 makeTranscript(transcriptFeature, al, gene);
198 clearGeneFeatures(gene);
202 * Remove unwanted features (transcript, exon, CDS) from the gene sequence
203 * after we have used them to derive transcripts and transfer features
207 protected void clearGeneFeatures(SequenceI gene)
209 SequenceFeature[] sfs = gene.getSequenceFeatures();
212 SequenceOntologyI so = SequenceOntologyFactory.getInstance();
213 List<SequenceFeature> filtered = new ArrayList<SequenceFeature>();
214 for (SequenceFeature sf : sfs)
216 String type = sf.getType();
217 if (!isTranscript(type) && !so.isA(type, SequenceOntologyI.EXON)
218 && !so.isA(type, SequenceOntologyI.CDS))
223 gene.setSequenceFeatures(filtered
224 .toArray(new SequenceFeature[filtered
230 * Constructs a spliced transcript sequence by finding 'exon' features for the
231 * given id (or failing that 'CDS'). Copies features on to the new sequence.
232 * 'Aligns' the new sequence against the gene sequence by padding with gaps,
233 * and adds it to the alignment.
235 * @param transcriptFeature
237 * the alignment to which to add the new sequence
239 * the parent gene sequence, with features
242 SequenceI makeTranscript(SequenceFeature transcriptFeature,
243 AlignmentI al, SequenceI gene)
245 String accId = getTranscriptId(transcriptFeature);
252 * NB we are mapping from gene sequence (not genome), so do not
253 * need to check for reverse strand (gene and transcript sequences
254 * are in forward sense)
258 * make a gene-length sequence filled with gaps
259 * we will fill in the bases for transcript regions
261 char[] seqChars = new char[gene.getLength()];
262 Arrays.fill(seqChars, al.getGapCharacter());
265 * look for exon features of the transcript, failing that for CDS
266 * (for example ENSG00000124610 has 1 CDS but no exon features)
268 String parentId = "transcript:" + accId;
269 List<SequenceFeature> splices = findFeatures(gene,
270 SequenceOntologyI.EXON, parentId);
271 if (splices.isEmpty())
273 splices = findFeatures(gene, SequenceOntologyI.CDS, parentId);
276 int transcriptLength = 0;
277 final char[] geneChars = gene.getSequence();
278 int offset = gene.getStart(); // to convert to 0-based positions
279 List<int[]> mappedFrom = new ArrayList<int[]>();
281 for (SequenceFeature sf : splices)
283 int start = sf.getBegin() - offset;
284 int end = sf.getEnd() - offset;
285 int spliceLength = end - start + 1;
286 System.arraycopy(geneChars, start, seqChars, start, spliceLength);
287 transcriptLength += spliceLength;
288 mappedFrom.add(new int[] { sf.getBegin(), sf.getEnd() });
291 Sequence transcript = new Sequence(accId, seqChars, 1, transcriptLength);
294 * Ensembl has gene name as transcript Name
295 * EnsemblGenomes doesn't, but has a url-encoded description field
297 String description = (String) transcriptFeature.getValue(NAME);
298 if (description == null)
300 description = (String) transcriptFeature.getValue(DESCRIPTION);
302 if (description != null)
306 transcript.setDescription(URLDecoder.decode(description, "UTF-8"));
307 } catch (UnsupportedEncodingException e)
309 e.printStackTrace(); // as if
312 transcript.createDatasetSequence();
314 al.addSequence(transcript);
317 * transfer features to the new sequence; we use EnsemblCdna to do this,
318 * to filter out unwanted features types (see method retainFeature)
320 List<int[]> mapTo = new ArrayList<int[]>();
321 mapTo.add(new int[] { 1, transcriptLength });
322 MapList mapping = new MapList(mappedFrom, mapTo, 1, 1);
323 EnsemblCdna cdna = new EnsemblCdna(getDomain());
324 cdna.transferFeatures(gene.getSequenceFeatures(),
325 transcript.getDatasetSequence(), mapping, parentId);
328 * fetch and save cross-references
330 cdna.getCrossReferences(transcript);
333 * and finally fetch the protein product and save as a cross-reference
335 cdna.addProteinProduct(transcript);
341 * Returns the 'transcript_id' property of the sequence feature (or null)
346 protected String getTranscriptId(SequenceFeature feature)
348 return (String) feature.getValue("transcript_id");
352 * Returns a list of the transcript features on the sequence whose Parent is
353 * the gene for the accession id.
356 * @param geneSequence
359 protected List<SequenceFeature> getTranscriptFeatures(String accId,
360 SequenceI geneSequence)
362 List<SequenceFeature> transcriptFeatures = new ArrayList<SequenceFeature>();
364 String parentIdentifier = GENE_PREFIX + accId;
365 SequenceFeature[] sfs = geneSequence.getSequenceFeatures();
369 for (SequenceFeature sf : sfs)
371 if (isTranscript(sf.getType()))
373 String parent = (String) sf.getValue(PARENT);
374 if (parentIdentifier.equals(parent))
376 transcriptFeatures.add(sf);
382 return transcriptFeatures;
386 public String getDescription()
388 return "Fetches all transcripts and variant features for a gene or transcript";
392 * Default test query is a gene id (can also enter a transcript id)
395 public String getTestQuery()
397 return "ENSG00000157764"; // BRAF, 5 transcripts, reverse strand
398 // ENSG00000090266 // NDUFB2, 15 transcripts, forward strand
399 // ENSG00000101812 // H2BFM histone, 3 transcripts, forward strand
400 // ENSG00000123569 // H2BFWT histone, 2 transcripts, reverse strand
404 * Answers true for a feature of type 'gene' (or a sub-type of gene in the
405 * Sequence Ontology), whose ID is the accession we are retrieving
408 protected boolean identifiesSequence(SequenceFeature sf, String accId)
410 if (SequenceOntologyFactory.getInstance().isA(sf.getType(),
411 SequenceOntologyI.GENE))
413 String id = (String) sf.getValue(ID);
414 if ((GENE_PREFIX + accId).equals(id))
423 * Answers true unless feature type is 'gene', or 'transcript' with a parent
424 * which is a different gene. We need the gene features to identify the range,
425 * but it is redundant information on the gene sequence. Checking the parent
426 * allows us to drop transcript features which belong to different
427 * (overlapping) genes.
430 protected boolean retainFeature(SequenceFeature sf, String accessionId)
432 SequenceOntologyI so = SequenceOntologyFactory.getInstance();
433 String type = sf.getType();
434 if (so.isA(type, SequenceOntologyI.GENE))
438 if (isTranscript(type))
440 String parent = (String) sf.getValue(PARENT);
441 if (!(GENE_PREFIX + accessionId).equals(parent))
450 * Answers false. This allows an optimisation - a single 'gene' feature is all
451 * that is needed to identify the positions of the gene on the genomic
455 protected boolean isSpliceable()
461 * Override to do nothing as Ensembl doesn't return a protein sequence for a
465 protected void addProteinProduct(SequenceI querySeq)
470 public Regex getAccessionValidator()
472 return ACCESSION_REGEX;
476 * Returns a descriptor for suitable feature display settings with
478 * <li>only exon or sequence_variant features (or their subtypes in the
479 * Sequence Ontology) visible</li>
480 * <li>variant features coloured red</li>
481 * <li>exon features coloured by label (exon name)</li>
482 * <li>variants displayed above (on top of) exons</li>
486 public FeatureSettingsModelI getFeatureColourScheme()
488 return new FeatureSettingsAdapter()
490 SequenceOntologyI so = SequenceOntologyFactory.getInstance();
492 public boolean isFeatureDisplayed(String type)
494 return (so.isA(type, SequenceOntologyI.EXON) || so.isA(type,
495 SequenceOntologyI.SEQUENCE_VARIANT));
499 public FeatureColourI getFeatureColour(String type)
501 if (so.isA(type, SequenceOntologyI.EXON))
503 return new FeatureColourAdapter()
506 public boolean isColourByLabel()
512 if (so.isA(type, SequenceOntologyI.SEQUENCE_VARIANT))
514 return new FeatureColourAdapter()
518 public Color getColour()
528 * order to render sequence_variant after exon after the rest
531 public int compare(String feature1, String feature2)
533 if (so.isA(feature1, SequenceOntologyI.SEQUENCE_VARIANT))
537 if (so.isA(feature2, SequenceOntologyI.SEQUENCE_VARIANT))
541 if (so.isA(feature1, SequenceOntologyI.EXON))
545 if (so.isA(feature2, SequenceOntologyI.EXON))
555 public int getMaximumQueryCount()