1 package jalview.ext.ensembl;
3 import jalview.analysis.AlignmentUtils;
4 import jalview.datamodel.Alignment;
5 import jalview.datamodel.AlignmentI;
6 import jalview.datamodel.DBRefEntry;
7 import jalview.datamodel.DBRefSource;
8 import jalview.datamodel.Mapping;
9 import jalview.datamodel.SequenceFeature;
10 import jalview.datamodel.SequenceI;
11 import jalview.exceptions.JalviewException;
12 import jalview.io.FastaFile;
13 import jalview.io.FileParse;
14 import jalview.io.gff.SequenceOntology;
15 import jalview.schemes.ResidueProperties;
16 import jalview.util.DBRefUtils;
17 import jalview.util.MapList;
18 import jalview.util.MappingUtils;
19 import jalview.util.StringUtils;
21 import java.io.IOException;
22 import java.net.MalformedURLException;
24 import java.util.ArrayList;
25 import java.util.Arrays;
26 import java.util.Collections;
27 import java.util.Comparator;
28 import java.util.LinkedHashMap;
29 import java.util.List;
30 import java.util.Map.Entry;
33 * Base class for Ensembl sequence fetchers
37 public abstract class EnsemblSeqProxy extends EnsemblRestClient
39 private static final List<String> CROSS_REFERENCES = Arrays
40 .asList(new String[] { "CCDS" });
42 protected static final String CONSEQUENCE_TYPE = "consequence_type";
44 protected static final String PARENT = "Parent";
46 protected static final String ID = "ID";
49 * this needs special handling, as it isA sequence_variant in the
50 * Sequence Ontology, but behaves in Ensembl as if it isA transcript
52 protected static final String NMD_VARIANT = "NMD_transcript_variant";
54 protected static final String NAME = "Name";
56 public enum EnsemblSeqType
59 * type=genomic for the full dna including introns
64 * type=cdna for transcribed dna including UTRs
69 * type=cds for coding dna excluding UTRs
74 * type=protein for the peptide product sequence
79 * the value of the 'type' parameter to fetch this version of
84 EnsemblSeqType(String t)
89 public String getType()
97 * A comparator to sort ranges into ascending start position order
99 private class RangeSorter implements Comparator<int[]>
103 RangeSorter(boolean forward)
109 public int compare(int[] o1, int[] o2)
111 return (forwards ? 1 : -1) * Integer.compare(o1[0], o2[0]);
119 public EnsemblSeqProxy()
124 * Makes the sequence queries to Ensembl's REST service and returns an
125 * alignment consisting of the returned sequences.
128 public AlignmentI getSequenceRecords(String query) throws Exception
130 long now = System.currentTimeMillis();
131 // TODO use a String... query vararg instead?
133 // danger: accession separator used as a regex here, a string elsewhere
134 // in this case it is ok (it is just a space), but (e.g.) '\' would not be
135 List<String> allIds = Arrays.asList(query
136 .split(getAccessionSeparator()));
137 AlignmentI alignment = null;
141 * execute queries, if necessary in batches of the
142 * maximum allowed number of ids
144 int maxQueryCount = getMaximumQueryCount();
145 for (int v = 0, vSize = allIds.size(); v < vSize; v += maxQueryCount)
147 int p = Math.min(vSize, v + maxQueryCount);
148 List<String> ids = allIds.subList(v, p);
151 alignment = fetchSequences(ids, alignment);
152 } catch (Throwable r)
155 String msg = "Aborting ID retrieval after " + v
156 + " chunks. Unexpected problem (" + r.getLocalizedMessage()
158 System.err.println(msg);
159 if (alignment != null)
161 break; // return what we got
165 throw new JalviewException(msg, r);
171 * fetch and transfer genomic sequence features,
172 * fetch protein product and add as cross-reference
174 for (String accId : allIds)
176 addFeaturesAndProduct(accId, alignment);
179 for (SequenceI seq : alignment.getSequences())
181 getCrossReferences(seq);
184 System.out.println(getClass().getName() + " took "
185 + (System.currentTimeMillis() - now) + "ms to fetch");
190 * Fetches Ensembl features using the /overlap REST endpoint, and adds them to
191 * the sequence in the alignment. Also fetches the protein product, maps it
192 * from the CDS features of the sequence, and saves it as a cross-reference of
198 protected void addFeaturesAndProduct(String accId, AlignmentI alignment)
200 if (alignment == null)
208 * get 'dummy' genomic sequence with exon, cds and variation features
210 SequenceI genomicSequence = null;
211 EnsemblOverlap gffFetcher = new EnsemblOverlap();
212 EnsemblFeatureType[] features = getFeaturesToFetch();
213 AlignmentI geneFeatures = gffFetcher.getSequenceRecords(accId,
215 if (geneFeatures.getHeight() > 0)
217 genomicSequence = geneFeatures.getSequenceAt(0);
219 if (genomicSequence != null)
222 * transfer features to the query sequence
224 SequenceI querySeq = alignment.findName(accId);
225 if (transferFeatures(accId, genomicSequence, querySeq))
229 * fetch and map protein product, and add it as a cross-reference
230 * of the retrieved sequence
232 addProteinProduct(querySeq);
235 } catch (IOException e)
237 System.err.println("Error transferring Ensembl features: "
243 * Returns those sequence feature types to fetch from Ensembl. We may want
244 * features either because they are of interest to the user, or as means to
245 * identify the locations of the sequence on the genomic sequence (CDS
246 * features identify CDS, exon features identify cDNA etc).
250 protected abstract EnsemblFeatureType[] getFeaturesToFetch();
253 * Fetches and maps the protein product, and adds it as a cross-reference of
254 * the retrieved sequence
256 protected void addProteinProduct(SequenceI querySeq)
258 String accId = querySeq.getName();
261 AlignmentI protein = new EnsemblProtein().getSequenceRecords(accId);
262 if (protein == null || protein.getHeight() == 0)
264 System.out.println("Failed to retrieve protein for " + accId);
267 SequenceI proteinSeq = protein.getSequenceAt(0);
270 * need dataset sequences (to be the subject of mappings)
272 proteinSeq.createDatasetSequence();
273 querySeq.createDatasetSequence();
275 MapList mapList = mapCdsToProtein(querySeq, proteinSeq);
278 Mapping map = new Mapping(proteinSeq.getDatasetSequence(), mapList);
279 DBRefEntry dbr = new DBRefEntry(getDbSource(), getDbVersion(),
281 querySeq.getDatasetSequence().addDBRef(dbr);
284 * compute peptide variants from dna variants and add as
285 * sequence features on the protein sequence ta-da
287 computeProteinFeatures(querySeq, proteinSeq, mapList);
289 } catch (Exception e)
292 .println(String.format("Error retrieving protein for %s: %s",
293 accId, e.getMessage()));
298 * Get Uniprot and PDB xrefs from Ensembl, and attach them to the protein
303 protected void getCrossReferences(SequenceI seq)
305 while (seq.getDatasetSequence() != null)
307 seq = seq.getDatasetSequence();
310 EnsemblXref xrefFetcher = new EnsemblXref();
311 List<DBRefEntry> xrefs = xrefFetcher.getCrossReferences(seq.getName(),
312 getCrossReferenceDatabases());
313 for (DBRefEntry xref : xrefs)
320 * Returns a list of database names to be used when fetching cross-references.
324 protected List<String> getCrossReferenceDatabases()
326 return CROSS_REFERENCES;
330 * Returns a mapping from dna to protein by inspecting sequence features of
331 * type "CDS" on the dna.
337 protected MapList mapCdsToProtein(SequenceI dnaSeq, SequenceI proteinSeq)
339 List<int[]> ranges = new ArrayList<int[]>(50);
341 int mappedDnaLength = getCdsRanges(dnaSeq, ranges);
343 int proteinLength = proteinSeq.getLength();
344 List<int[]> proteinRange = new ArrayList<int[]>();
345 int proteinStart = 1;
348 * incomplete start codon may mean X at start of peptide
349 * we ignore both for mapping purposes
351 if (proteinSeq.getCharAt(0) == 'X')
356 proteinRange.add(new int[] { proteinStart, proteinLength });
359 * dna length should map to protein (or protein plus stop codon)
361 int codesForResidues = mappedDnaLength / 3;
362 if (codesForResidues == proteinLength
363 || codesForResidues == (proteinLength + 1))
365 return new MapList(ranges, proteinRange, 3, 1);
371 * Adds CDS ranges to the ranges list, and returns the total length mapped.
373 * No need to worry about reverse strand dna here since the retrieved sequence
374 * is as transcribed (reverse complement for reverse strand), i.e in the same
375 * sense as the peptide.
381 protected int getCdsRanges(SequenceI dnaSeq, List<int[]> ranges)
383 SequenceFeature[] sfs = dnaSeq.getSequenceFeatures();
388 int mappedDnaLength = 0;
389 for (SequenceFeature sf : sfs)
392 * process a CDS feature (or a sub-type of CDS)
394 if (SequenceOntology.getInstance().isA(sf.getType(), SequenceOntology.CDS))
398 phase = Integer.parseInt(sf.getPhase());
399 } catch (NumberFormatException e)
404 * phase > 0 on first codon means 5' incomplete - skip to the start
405 * of the next codon; example ENST00000496384
407 int begin = sf.getBegin();
408 int end = sf.getEnd();
409 if (ranges.isEmpty() && phase > 0)
414 continue; // shouldn't happen?
417 ranges.add(new int[] { begin, end });
418 mappedDnaLength += Math.abs(end - begin) + 1;
421 return mappedDnaLength;
425 * Fetches sequences for the list of accession ids and adds them to the
426 * alignment. Returns the extended (or created) alignment.
431 * @throws JalviewException
432 * @throws IOException
434 protected AlignmentI fetchSequences(List<String> ids, AlignmentI alignment)
435 throws JalviewException, IOException
437 if (!isEnsemblAvailable())
440 throw new JalviewException("ENSEMBL Rest API not available.");
442 FileParse fp = getSequenceReader(ids);
443 FastaFile fr = new FastaFile(fp);
444 if (fr.hasWarningMessage())
446 System.out.println(String.format(
447 "Warning when retrieving %d ids %s\n%s", ids.size(),
448 ids.toString(), fr.getWarningMessage()));
450 else if (fr.getSeqs().size() != ids.size())
452 System.out.println(String.format(
453 "Only retrieved %d sequences for %d query strings", fr
454 .getSeqs().size(), ids.size()));
457 if (fr.getSeqs().size() == 1 && fr.getSeqs().get(0).getLength() == 0)
460 * POST request has returned an empty FASTA file e.g. for invalid id
462 throw new IOException("No data returned for " + ids);
465 if (fr.getSeqs().size() > 0)
467 AlignmentI seqal = new Alignment(
468 fr.getSeqsAsArray());
469 for (SequenceI sq:seqal.getSequences())
471 if (sq.getDescription() == null)
473 sq.setDescription(getDbName());
475 String name = sq.getName();
476 if (ids.contains(name)
477 || ids.contains(name.replace("ENSP", "ENST")))
479 DBRefUtils.parseToDbRef(sq, DBRefSource.ENSEMBL, "0", name);
482 if (alignment == null)
488 alignment.append(seqal);
495 * Returns the URL for the REST call
498 * @throws MalformedURLException
501 protected URL getUrl(List<String> ids) throws MalformedURLException
504 * a single id is included in the URL path
505 * multiple ids go in the POST body instead
507 StringBuffer urlstring = new StringBuffer(128);
508 urlstring.append(SEQUENCE_ID_URL);
511 urlstring.append("/").append(ids.get(0));
513 // @see https://github.com/Ensembl/ensembl-rest/wiki/Output-formats
514 urlstring.append("?type=").append(getSourceEnsemblType().getType());
515 urlstring.append(("&Accept=text/x-fasta"));
517 URL url = new URL(urlstring.toString());
522 * A sequence/id POST request currently allows up to 50 queries
524 * @see http://rest.ensembl.org/documentation/info/sequence_id_post
527 public int getMaximumQueryCount()
533 protected boolean useGetRequest()
539 protected String getRequestMimeType(boolean multipleIds)
541 return multipleIds ? "application/json" : "text/x-fasta";
545 protected String getResponseMimeType()
547 return "text/x-fasta";
552 * @return the configured sequence return type for this source
554 protected abstract EnsemblSeqType getSourceEnsemblType();
557 * Returns a list of [start, end] genomic ranges corresponding to the sequence
560 * The correspondence between the frames of reference is made by locating
561 * those features on the genomic sequence which identify the retrieved
562 * sequence. Specifically
564 * <li>genomic sequence is identified by "transcript" features with
565 * ID=transcript:transcriptId</li>
566 * <li>cdna sequence is identified by "exon" features with
567 * Parent=transcript:transcriptId</li>
568 * <li>cds sequence is identified by "CDS" features with
569 * Parent=transcript:transcriptId</li>
572 * The returned ranges are sorted to run forwards (for positive strand) or
573 * backwards (for negative strand). Aborts and returns null if both positive
574 * and negative strand are found (this should not normally happen).
576 * @param sourceSequence
579 * the start position of the sequence we are mapping to
582 protected MapList getGenomicRanges(SequenceI sourceSequence,
583 String accId, int start)
585 SequenceFeature[] sfs = sourceSequence.getSequenceFeatures();
592 * generously initial size for number of cds regions
593 * (worst case titin Q8WZ42 has c. 313 exons)
595 List<int[]> regions = new ArrayList<int[]>(100);
596 int mappedLength = 0;
597 int direction = 1; // forward
598 boolean directionSet = false;
600 for (SequenceFeature sf : sfs)
603 * accept the target feature type or a specialisation of it
604 * (e.g. coding_exon for exon)
606 if (identifiesSequence(sf, accId))
608 int strand = sf.getStrand();
610 if (directionSet && strand != direction)
612 // abort - mix of forward and backward
613 System.err.println("Error: forward and backward strand for "
621 * add to CDS ranges, semi-sorted forwards/backwards
625 regions.add(0, new int[] { sf.getEnd(), sf.getBegin() });
629 regions.add(new int[] { sf.getBegin(), sf.getEnd() });
631 mappedLength += Math.abs(sf.getEnd() - sf.getBegin() + 1);
636 * 'gene' sequence is contiguous so we can stop as soon as its
637 * identifying feature has been found
644 if (regions.isEmpty())
646 System.out.println("Failed to identify target sequence for " + accId
647 + " from genomic features");
652 * a final sort is needed since Ensembl returns CDS sorted within source
653 * (havana / ensembl_havana)
655 Collections.sort(regions, new RangeSorter(direction == 1));
657 List<int[]> to = new ArrayList<int[]>();
658 to.add(new int[] { start, start + mappedLength - 1 });
660 return new MapList(regions, to, 1, 1);
664 * Answers true if the sequence being retrieved may occupy discontiguous
665 * regions on the genomic sequence.
667 protected boolean isSpliceable()
673 * Returns true if the sequence feature marks positions of the genomic
674 * sequence feature which are within the sequence being retrieved. For
675 * example, an 'exon' feature whose parent is the target transcript marks the
676 * cdna positions of the transcript.
682 protected abstract boolean identifiesSequence(SequenceFeature sf,
686 * Transfers the sequence feature to the target sequence, locating its start
687 * and end range based on the mapping. Features which do not overlap the
688 * target sequence are ignored.
691 * @param targetSequence
693 * mapping from the sequence feature's coordinates to the target
696 protected void transferFeature(SequenceFeature sf,
697 SequenceI targetSequence, MapList mapping)
699 int start = sf.getBegin();
700 int end = sf.getEnd();
701 int[] mappedRange = mapping.locateInTo(start, end);
703 if (mappedRange != null)
705 SequenceFeature copy = new SequenceFeature(sf);
706 copy.setBegin(Math.min(mappedRange[0], mappedRange[1]));
707 copy.setEnd(Math.max(mappedRange[0], mappedRange[1]));
708 targetSequence.addSequenceFeature(copy);
711 * for sequence_variant, make an additional feature with consequence
713 if (SequenceOntology.getInstance().isSequenceVariant(sf.getType()))
715 String consequence = (String) sf.getValue(CONSEQUENCE_TYPE);
716 if (consequence != null)
718 SequenceFeature sf2 = new SequenceFeature("consequence",
719 consequence, copy.getBegin(), copy.getEnd(), 0f,
721 targetSequence.addSequenceFeature(sf2);
728 * Transfers features from sourceSequence to targetSequence
731 * @param sourceSequence
732 * @param targetSequence
733 * @return true if any features were transferred, else false
735 protected boolean transferFeatures(String accessionId,
736 SequenceI sourceSequence, SequenceI targetSequence)
738 if (sourceSequence == null || targetSequence == null)
743 SequenceFeature[] sfs = sourceSequence.getSequenceFeatures();
744 MapList mapping = getGenomicRanges(sourceSequence, accessionId,
745 targetSequence.getStart());
751 return transferFeatures(sfs, targetSequence, mapping, accessionId);
755 * Transfer features to the target sequence. The start/end positions are
756 * converted using the mapping. Features which do not overlap are ignored.
757 * Features whose parent is not the specified identifier are also ignored.
760 * @param targetSequence
765 protected boolean transferFeatures(SequenceFeature[] features,
766 SequenceI targetSequence, MapList mapping, String parentId)
768 final boolean forwardStrand = mapping.isFromForwardStrand();
771 * sort features by start position (descending if reverse strand)
772 * before transferring (in forwards order) to the target sequence
774 Arrays.sort(features, new Comparator<SequenceFeature>()
777 public int compare(SequenceFeature o1, SequenceFeature o2)
779 int c = Integer.compare(o1.getBegin(), o2.getBegin());
780 return forwardStrand ? c : -c;
784 boolean transferred = false;
785 for (SequenceFeature sf : features)
787 if (retainFeature(sf, parentId))
789 transferFeature(sf, targetSequence, mapping);
797 * Answers true if the feature type is one we want to keep for the sequence.
798 * Some features are only retrieved in order to identify the sequence range,
799 * and may then be discarded as redundant information (e.g. "CDS" feature for
802 @SuppressWarnings("unused")
803 protected boolean retainFeature(SequenceFeature sf, String accessionId)
805 return true; // override as required
809 * Answers true if the feature has a Parent which refers to the given
810 * accession id, or if the feature has no parent. Answers false if the
811 * feature's Parent is for a different accession id.
817 protected boolean featureMayBelong(SequenceFeature sf, String identifier)
819 String parent = (String) sf.getValue(PARENT);
820 // using contains to allow for prefix "gene:", "transcript:" etc
821 if (parent != null && !parent.contains(identifier))
823 // this genomic feature belongs to a different transcript
830 public String getDescription()
832 return "Ensembl " + getSourceEnsemblType().getType()
833 + " sequence with variant features";
837 * Returns a (possibly empty) list of features on the sequence which have the
838 * specified sequence ontology type (or a sub-type of it), and the given
839 * identifier as parent
846 protected List<SequenceFeature> findFeatures(SequenceI sequence,
847 String type, String parentId)
849 List<SequenceFeature> result = new ArrayList<SequenceFeature>();
851 SequenceFeature[] sfs = sequence.getSequenceFeatures();
853 SequenceOntology so = SequenceOntology.getInstance();
854 for (SequenceFeature sf :sfs) {
855 if (so.isA(sf.getType(), type))
857 String parent = (String) sf.getValue(PARENT);
858 if (parent.equals(parentId))
869 * Maps exon features from dna to protein, and computes variants in peptide
870 * product generated by variants in dna, and adds them as sequence_variant
871 * features on the protein sequence. Returns the number of variant features
876 * @param dnaToProtein
878 static int computeProteinFeatures(SequenceI dnaSeq,
879 SequenceI peptide, MapList dnaToProtein)
881 while (dnaSeq.getDatasetSequence() != null)
883 dnaSeq = dnaSeq.getDatasetSequence();
885 while (peptide.getDatasetSequence() != null)
887 peptide = peptide.getDatasetSequence();
890 AlignmentUtils.transferFeatures(dnaSeq, peptide, dnaToProtein,
891 SequenceOntology.EXON);
893 LinkedHashMap<Integer, String[][]> variants = buildDnaVariantsMap(
894 dnaSeq, dnaToProtein);
897 * scan codon variations, compute peptide variants and add to peptide sequence
900 for (Entry<Integer, String[][]> variant : variants.entrySet())
902 int peptidePos = variant.getKey();
903 String[][] codonVariants = variant.getValue();
904 String residue = String.valueOf(peptide.getCharAt(peptidePos - 1)); // 0-based
905 List<String> peptideVariants = computePeptideVariants(codonVariants,
907 if (!peptideVariants.isEmpty())
909 String desc = StringUtils.listToDelimitedString(peptideVariants,
911 SequenceFeature sf = new SequenceFeature(
912 SequenceOntology.SEQUENCE_VARIANT, desc, peptidePos,
913 peptidePos, 0f, null);
914 peptide.addSequenceFeature(sf);
922 * Builds a map whose key is position in the protein sequence, and value is an
923 * array of all variants for the coding codon positions
926 * @param dnaToProtein
929 static LinkedHashMap<Integer, String[][]> buildDnaVariantsMap(
930 SequenceI dnaSeq, MapList dnaToProtein)
933 * map from peptide position to all variant features of the codon for it
934 * LinkedHashMap ensures we add the peptide features in sequence order
936 LinkedHashMap<Integer, String[][]> variants = new LinkedHashMap<Integer, String[][]>();
937 SequenceOntology so = SequenceOntology.getInstance();
939 SequenceFeature[] dnaFeatures = dnaSeq.getSequenceFeatures();
940 if (dnaFeatures == null)
945 int dnaStart = dnaSeq.getStart();
946 int[] lastCodon = null;
947 int lastPeptidePostion = 0;
950 * build a map of codon variations for peptides
952 for (SequenceFeature sf : dnaFeatures)
954 int dnaCol = sf.getBegin();
955 if (dnaCol != sf.getEnd())
957 // not handling multi-locus variant features
960 if (so.isSequenceVariant(sf.getType()))
962 int[] mapsTo = dnaToProtein.locateInTo(dnaCol, dnaCol);
965 // feature doesn't lie within coding region
968 int peptidePosition = mapsTo[0];
969 String[][] codonVariants = variants.get(peptidePosition);
970 if (codonVariants == null)
972 codonVariants = new String[3][];
973 variants.put(peptidePosition, codonVariants);
977 * extract dna variants to a string array
979 String alls = (String) sf.getValue("alleles");
984 String[] alleles = alls.split(",");
987 * get this peptides codon positions e.g. [3, 4, 5] or [4, 7, 10]
989 int[] codon = peptidePosition == lastPeptidePostion ? lastCodon
990 : MappingUtils.flattenRanges(dnaToProtein.locateInFrom(
991 peptidePosition, peptidePosition));
992 lastPeptidePostion = peptidePosition;
996 * save nucleotide (and this variant) for each codon position
998 for (int codonPos = 0; codonPos < 3; codonPos++)
1000 String nucleotide = String.valueOf(dnaSeq
1001 .getCharAt(codon[codonPos] - dnaStart));
1002 if (codon[codonPos] == dnaCol)
1005 * record current dna base and its alleles
1007 String[] dnaVariants = new String[alleles.length + 1];
1008 dnaVariants[0] = nucleotide;
1009 System.arraycopy(alleles, 0, dnaVariants, 1, alleles.length);
1010 codonVariants[codonPos] = dnaVariants;
1012 else if (codonVariants[codonPos] == null)
1015 * record current dna base only
1016 * (at least until we find any variation and overwrite it)
1018 codonVariants[codonPos] = new String[] { nucleotide };
1027 * Returns a sorted, non-redundant list of all peptide translations generated
1028 * by the given dna variants, excluding the current residue value
1030 * @param codonVariants
1031 * an array of base values (acgtACGT) for codon positions 1, 2, 3
1033 * the current residue translation
1036 static List<String> computePeptideVariants(
1037 String[][] codonVariants, String residue)
1039 List<String> result = new ArrayList<String>();
1040 for (String base1 : codonVariants[0])
1042 for (String base2 : codonVariants[1])
1044 for (String base3 : codonVariants[2])
1046 String codon = base1 + base2 + base3;
1047 // TODO: report frameshift/insertion/deletion
1048 // and multiple-base variants?!
1049 String peptide = codon.contains("-") ? "-" : ResidueProperties
1050 .codonTranslate(codon);
1051 if (peptide != null && !result.contains(peptide)
1052 && !peptide.equalsIgnoreCase(residue))
1054 result.add(peptide);
1061 * sort alphabetically with STOP at the end
1063 Collections.sort(result, new Comparator<String>()
1067 public int compare(String o1, String o2)
1069 if ("STOP".equals(o1))
1073 else if ("STOP".equals(o2))
1079 return o1.compareTo(o2);
1087 * Answers true if the feature type is either 'NMD_transcript_variant' or
1088 * 'transcript' or one of its sub-types in the Sequence Ontology. This is
1089 * needed because NMD_transcript_variant behaves like 'transcript' in Ensembl
1090 * although strictly speaking it is not (it is a sub-type of
1091 * sequence_variant).
1093 * @param featureType
1096 public static boolean isTranscript(String featureType)
1098 return NMD_VARIANT.equals(featureType)
1099 || SequenceOntology.getInstance().isA(featureType, SequenceOntology.TRANSCRIPT);