1 package jalview.ext.ensembl;
3 import jalview.datamodel.Alignment;
4 import jalview.datamodel.AlignmentI;
5 import jalview.datamodel.DBRefEntry;
6 import jalview.datamodel.DBRefSource;
7 import jalview.datamodel.Mapping;
8 import jalview.datamodel.SequenceFeature;
9 import jalview.datamodel.SequenceI;
10 import jalview.exceptions.JalviewException;
11 import jalview.io.FastaFile;
12 import jalview.io.FileParse;
13 import jalview.io.gff.SequenceOntology;
14 import jalview.schemes.ResidueProperties;
15 import jalview.util.DBRefUtils;
16 import jalview.util.MapList;
17 import jalview.util.MappingUtils;
18 import jalview.util.StringUtils;
20 import java.io.IOException;
21 import java.net.MalformedURLException;
23 import java.util.ArrayList;
24 import java.util.Arrays;
25 import java.util.Collections;
26 import java.util.Comparator;
27 import java.util.LinkedHashMap;
28 import java.util.List;
29 import java.util.Map.Entry;
32 * Base class for Ensembl sequence fetchers
36 public abstract class EnsemblSeqProxy extends EnsemblRestClient
38 protected static final String CONSEQUENCE_TYPE = "consequence_type";
40 protected static final String PARENT = "Parent";
42 protected static final String ID = "ID";
45 * this needs special handling, as it isA sequence_variant in the
46 * Sequence Ontology, but behaves in Ensembl as if it isA transcript
48 protected static final String NMD_VARIANT = "NMD_transcript_variant";
50 protected static final String NAME = "Name";
52 public enum EnsemblSeqType
55 * type=genomic for the full dna including introns
60 * type=cdna for transcribed dna including UTRs
65 * type=cds for coding dna excluding UTRs
70 * type=protein for the peptide product sequence
75 * the value of the 'type' parameter to fetch this version of
80 EnsemblSeqType(String t)
85 public String getType()
93 * A comparator to sort ranges into ascending start position order
95 private class RangeSorter implements Comparator<int[]>
99 RangeSorter(boolean forward)
105 public int compare(int[] o1, int[] o2)
107 return (forwards ? 1 : -1) * Integer.compare(o1[0], o2[0]);
115 public EnsemblSeqProxy()
120 * Makes the sequence queries to Ensembl's REST service and returns an
121 * alignment consisting of the returned sequences.
124 public AlignmentI getSequenceRecords(String query) throws Exception
126 long now = System.currentTimeMillis();
127 // TODO use a String... query vararg instead?
129 // danger: accession separator used as a regex here, a string elsewhere
130 // in this case it is ok (it is just a space), but (e.g.) '\' would not be
131 List<String> allIds = Arrays.asList(query
132 .split(getAccessionSeparator()));
133 AlignmentI alignment = null;
137 * execute queries, if necessary in batches of the
138 * maximum allowed number of ids
140 int maxQueryCount = getMaximumQueryCount();
141 for (int v = 0, vSize = allIds.size(); v < vSize; v += maxQueryCount)
143 int p = Math.min(vSize, v + maxQueryCount);
144 List<String> ids = allIds.subList(v, p);
147 alignment = fetchSequences(ids, alignment);
148 } catch (Throwable r)
151 String msg = "Aborting ID retrieval after " + v
152 + " chunks. Unexpected problem (" + r.getLocalizedMessage()
154 System.err.println(msg);
155 if (alignment != null)
157 break; // return what we got
161 throw new JalviewException(msg, r);
167 * fetch and transfer genomic sequence features,
168 * fetch protein product and add as cross-reference
170 for (String accId : allIds)
172 addFeaturesAndProduct(accId, alignment);
176 System.out.println(getClass().getName() + " took "
177 + (System.currentTimeMillis() - now) + "ms to fetch");
182 * Fetches Ensembl features using the /overlap REST endpoint, and adds them to
183 * the sequence in the alignment. Also fetches the protein product, maps it
184 * from the CDS features of the sequence, and saves it as a cross-reference of
190 protected void addFeaturesAndProduct(String accId, AlignmentI alignment)
192 if (alignment == null)
200 * get 'dummy' genomic sequence with exon, cds and variation features
202 SequenceI genomicSequence = null;
203 EnsemblOverlap gffFetcher = new EnsemblOverlap();
204 EnsemblFeatureType[] features = getFeaturesToFetch();
205 AlignmentI geneFeatures = gffFetcher.getSequenceRecords(accId,
207 if (geneFeatures.getHeight() > 0)
209 genomicSequence = geneFeatures.getSequenceAt(0);
211 if (genomicSequence != null)
214 * transfer features to the query sequence
216 SequenceI querySeq = alignment.findName(accId);
217 if (transferFeatures(accId, genomicSequence, querySeq))
221 * fetch and map protein product, and add it as a cross-reference
222 * of the retrieved sequence
224 addProteinProduct(querySeq);
227 } catch (IOException e)
229 System.err.println("Error transferring Ensembl features: "
235 * Returns those sequence feature types to fetch from Ensembl. We may want
236 * features either because they are of interest to the user, or as means to
237 * identify the locations of the sequence on the genomic sequence (CDS
238 * features identify CDS, exon features identify cDNA etc).
242 protected abstract EnsemblFeatureType[] getFeaturesToFetch();
245 * Fetches and maps the protein product, and adds it as a cross-reference of
246 * the retrieved sequence
248 protected void addProteinProduct(SequenceI querySeq)
250 String accId = querySeq.getName();
253 AlignmentI protein = new EnsemblProtein().getSequenceRecords(accId);
254 if (protein == null || protein.getHeight() == 0)
256 System.out.println("Failed to retrieve protein for " + accId);
259 SequenceI proteinSeq = protein.getSequenceAt(0);
262 * need dataset sequences (to be the subject of mappings)
264 proteinSeq.createDatasetSequence();
265 querySeq.createDatasetSequence();
267 MapList mapList = mapCdsToProtein(querySeq, proteinSeq);
270 Mapping map = new Mapping(proteinSeq.getDatasetSequence(), mapList);
271 DBRefEntry dbr = new DBRefEntry(getDbSource(), getDbVersion(),
273 querySeq.getDatasetSequence().addDBRef(dbr);
276 * compute peptide variants from dna variants and add as
277 * sequence features on the protein sequence ta-da
279 computeProteinFeatures(querySeq, proteinSeq, mapList);
281 } catch (Exception e)
284 .println(String.format("Error retrieving protein for %s: %s",
285 accId, e.getMessage()));
290 * Returns a mapping from dna to protein by inspecting sequence features of
291 * type "CDS" on the dna.
297 protected MapList mapCdsToProtein(SequenceI dnaSeq, SequenceI proteinSeq)
299 List<int[]> ranges = new ArrayList<int[]>(50);
301 int mappedDnaLength = getCdsRanges(dnaSeq, ranges);
303 int proteinLength = proteinSeq.getLength();
304 List<int[]> proteinRange = new ArrayList<int[]>();
305 int proteinStart = 1;
308 * incomplete start codon may mean X at start of peptide
309 * we ignore both for mapping purposes
311 if (proteinSeq.getCharAt(0) == 'X')
316 proteinRange.add(new int[] { proteinStart, proteinLength });
319 * dna length should map to protein (or protein plus stop codon)
321 int codesForResidues = mappedDnaLength / 3;
322 if (codesForResidues == proteinLength
323 || codesForResidues == (proteinLength + 1))
325 return new MapList(ranges, proteinRange, 3, 1);
331 * Adds CDS ranges to the ranges list, and returns the total length mapped.
333 * No need to worry about reverse strand dna here since the retrieved sequence
334 * is as transcribed (reverse complement for reverse strand), i.e in the same
335 * sense as the peptide.
341 protected int getCdsRanges(SequenceI dnaSeq, List<int[]> ranges)
343 SequenceFeature[] sfs = dnaSeq.getSequenceFeatures();
348 int mappedDnaLength = 0;
349 for (SequenceFeature sf : sfs)
352 * process a CDS feature (or a sub-type of CDS)
354 if (SequenceOntology.getInstance().isA(sf.getType(), SequenceOntology.CDS))
358 phase = Integer.parseInt(sf.getPhase());
359 } catch (NumberFormatException e)
364 * phase > 0 on first codon means 5' incomplete - skip to the start
365 * of the next codon; example ENST00000496384
367 int begin = sf.getBegin();
368 int end = sf.getEnd();
369 if (ranges.isEmpty() && phase > 0)
374 continue; // shouldn't happen?
377 ranges.add(new int[] { begin, end });
378 mappedDnaLength += Math.abs(end - begin) + 1;
381 return mappedDnaLength;
385 * Fetches sequences for the list of accession ids and adds them to the
386 * alignment. Returns the extended (or created) alignment.
391 * @throws JalviewException
392 * @throws IOException
394 protected AlignmentI fetchSequences(List<String> ids, AlignmentI alignment)
395 throws JalviewException, IOException
397 if (!isEnsemblAvailable())
400 throw new JalviewException("ENSEMBL Rest API not available.");
402 FileParse fp = getSequenceReader(ids);
403 FastaFile fr = new FastaFile(fp);
404 if (fr.hasWarningMessage())
406 System.out.println(String.format(
407 "Warning when retrieving %d ids %s\n%s", ids.size(),
408 ids.toString(), fr.getWarningMessage()));
410 else if (fr.getSeqs().size() != ids.size())
412 System.out.println(String.format(
413 "Only retrieved %d sequences for %d query strings", fr
414 .getSeqs().size(), ids.size()));
417 if (fr.getSeqs().size() == 1 && fr.getSeqs().get(0).getLength() == 0)
420 * POST request has returned an empty FASTA file e.g. for invalid id
422 throw new IOException("No data returned for " + ids);
425 if (fr.getSeqs().size() > 0)
427 AlignmentI seqal = new Alignment(
428 fr.getSeqsAsArray());
429 for (SequenceI sq:seqal.getSequences())
431 if (sq.getDescription() == null)
433 sq.setDescription(getDbName());
435 String name = sq.getName();
436 if (ids.contains(name)
437 || ids.contains(name.replace("ENSP", "ENST")))
439 DBRefUtils.parseToDbRef(sq, DBRefSource.ENSEMBL, "0", name);
442 if (alignment == null)
448 alignment.append(seqal);
455 * Returns the URL for the REST call
458 * @throws MalformedURLException
461 protected URL getUrl(List<String> ids) throws MalformedURLException
464 * a single id is included in the URL path
465 * multiple ids go in the POST body instead
467 StringBuffer urlstring = new StringBuffer(128);
468 urlstring.append(SEQUENCE_ID_URL);
471 urlstring.append("/").append(ids.get(0));
473 // @see https://github.com/Ensembl/ensembl-rest/wiki/Output-formats
474 urlstring.append("?type=").append(getSourceEnsemblType().getType());
475 urlstring.append(("&Accept=text/x-fasta"));
477 URL url = new URL(urlstring.toString());
482 * A sequence/id POST request currently allows up to 50 queries
484 * @see http://rest.ensembl.org/documentation/info/sequence_id_post
487 public int getMaximumQueryCount()
493 protected boolean useGetRequest()
499 protected String getRequestMimeType(boolean multipleIds)
501 return multipleIds ? "application/json" : "text/x-fasta";
505 protected String getResponseMimeType()
507 return "text/x-fasta";
512 * @return the configured sequence return type for this source
514 protected abstract EnsemblSeqType getSourceEnsemblType();
517 * Returns a list of [start, end] genomic ranges corresponding to the sequence
520 * The correspondence between the frames of reference is made by locating
521 * those features on the genomic sequence which identify the retrieved
522 * sequence. Specifically
524 * <li>genomic sequence is identified by "transcript" features with
525 * ID=transcript:transcriptId</li>
526 * <li>cdna sequence is identified by "exon" features with
527 * Parent=transcript:transcriptId</li>
528 * <li>cds sequence is identified by "CDS" features with
529 * Parent=transcript:transcriptId</li>
532 * The returned ranges are sorted to run forwards (for positive strand) or
533 * backwards (for negative strand). Aborts and returns null if both positive
534 * and negative strand are found (this should not normally happen).
536 * @param sourceSequence
539 * the start position of the sequence we are mapping to
542 protected MapList getGenomicRanges(SequenceI sourceSequence,
543 String accId, int start)
545 SequenceFeature[] sfs = sourceSequence.getSequenceFeatures();
552 * generously initial size for number of cds regions
553 * (worst case titin Q8WZ42 has c. 313 exons)
555 List<int[]> regions = new ArrayList<int[]>(100);
556 int mappedLength = 0;
557 int direction = 1; // forward
558 boolean directionSet = false;
560 for (SequenceFeature sf : sfs)
563 * accept the target feature type or a specialisation of it
564 * (e.g. coding_exon for exon)
566 if (identifiesSequence(sf, accId))
568 int strand = sf.getStrand();
570 if (directionSet && strand != direction)
572 // abort - mix of forward and backward
573 System.err.println("Error: forward and backward strand for "
581 * add to CDS ranges, semi-sorted forwards/backwards
585 regions.add(0, new int[] { sf.getEnd(), sf.getBegin() });
589 regions.add(new int[] { sf.getBegin(), sf.getEnd() });
591 mappedLength += Math.abs(sf.getEnd() - sf.getBegin() + 1);
596 * 'gene' sequence is contiguous so we can stop as soon as its
597 * identifying feature has been found
604 if (regions.isEmpty())
606 System.out.println("Failed to identify target sequence for " + accId
607 + " from genomic features");
612 * a final sort is needed since Ensembl returns CDS sorted within source
613 * (havana / ensembl_havana)
615 Collections.sort(regions, new RangeSorter(direction == 1));
617 List<int[]> to = new ArrayList<int[]>();
618 to.add(new int[] { start, start + mappedLength - 1 });
620 return new MapList(regions, to, 1, 1);
624 * Answers true if the sequence being retrieved may occupy discontiguous
625 * regions on the genomic sequence.
627 protected boolean isSpliceable()
633 * Returns true if the sequence feature marks positions of the genomic
634 * sequence feature which are within the sequence being retrieved. For
635 * example, an 'exon' feature whose parent is the target transcript marks the
636 * cdna positions of the transcript.
642 protected abstract boolean identifiesSequence(SequenceFeature sf,
646 * Transfers the sequence feature to the target sequence, locating its start
647 * and end range based on the mapping. Features which do not overlap the
648 * target sequence are ignored.
651 * @param targetSequence
653 * mapping from the sequence feature's coordinates to the target
656 protected void transferFeature(SequenceFeature sf,
657 SequenceI targetSequence, MapList mapping)
659 int start = sf.getBegin();
660 int end = sf.getEnd();
661 int[] mappedRange = mapping.locateInTo(start, end);
663 if (mappedRange != null)
665 SequenceFeature copy = new SequenceFeature(sf);
666 copy.setBegin(Math.min(mappedRange[0], mappedRange[1]));
667 copy.setEnd(Math.max(mappedRange[0], mappedRange[1]));
668 targetSequence.addSequenceFeature(copy);
671 * for sequence_variant, make an additional feature with consequence
673 if (SequenceOntology.getInstance().isSequenceVariant(sf.getType()))
675 String consequence = (String) sf.getValue(CONSEQUENCE_TYPE);
676 if (consequence != null)
678 SequenceFeature sf2 = new SequenceFeature("consequence",
679 consequence, copy.getBegin(), copy.getEnd(), 0f,
681 targetSequence.addSequenceFeature(sf2);
688 * Transfers features from sourceSequence to targetSequence
691 * @param sourceSequence
692 * @param targetSequence
693 * @return true if any features were transferred, else false
695 protected boolean transferFeatures(String accessionId,
696 SequenceI sourceSequence, SequenceI targetSequence)
698 if (sourceSequence == null || targetSequence == null)
703 SequenceFeature[] sfs = sourceSequence.getSequenceFeatures();
704 MapList mapping = getGenomicRanges(sourceSequence, accessionId,
705 targetSequence.getStart());
711 return transferFeatures(sfs, targetSequence, mapping, accessionId);
715 * Transfer features to the target sequence. The start/end positions are
716 * converted using the mapping. Features which do not overlap are ignored.
717 * Features whose parent is not the specified identifier are also ignored.
720 * @param targetSequence
725 protected boolean transferFeatures(SequenceFeature[] features,
726 SequenceI targetSequence, MapList mapping, String parentId)
728 final boolean forwardStrand = mapping.isFromForwardStrand();
731 * sort features by start position (descending if reverse strand)
732 * before transferring (in forwards order) to the target sequence
734 Arrays.sort(features, new Comparator<SequenceFeature>()
737 public int compare(SequenceFeature o1, SequenceFeature o2)
739 int c = Integer.compare(o1.getBegin(), o2.getBegin());
740 return forwardStrand ? c : -c;
744 boolean transferred = false;
745 for (SequenceFeature sf : features)
747 if (retainFeature(sf, parentId))
749 transferFeature(sf, targetSequence, mapping);
757 * Answers true if the feature type is one we want to keep for the sequence.
758 * Some features are only retrieved in order to identify the sequence range,
759 * and may then be discarded as redundant information (e.g. "CDS" feature for
762 @SuppressWarnings("unused")
763 protected boolean retainFeature(SequenceFeature sf, String accessionId)
765 return true; // override as required
769 * Answers true if the feature has a Parent which refers to the given
770 * accession id, or if the feature has no parent. Answers false if the
771 * feature's Parent is for a different accession id.
777 protected boolean featureMayBelong(SequenceFeature sf, String identifier)
779 String parent = (String) sf.getValue(PARENT);
780 // using contains to allow for prefix "gene:", "transcript:" etc
781 if (parent != null && !parent.contains(identifier))
783 // this genomic feature belongs to a different transcript
790 public String getDescription()
792 return "Ensembl " + getSourceEnsemblType().getType()
793 + " sequence with variant features";
797 * Returns a (possibly empty) list of features on the sequence which have the
798 * specified sequence ontology type (or a sub-type of it), and the given
799 * identifier as parent
806 protected List<SequenceFeature> findFeatures(SequenceI sequence,
807 String type, String parentId)
809 List<SequenceFeature> result = new ArrayList<SequenceFeature>();
811 SequenceFeature[] sfs = sequence.getSequenceFeatures();
813 SequenceOntology so = SequenceOntology.getInstance();
814 for (SequenceFeature sf :sfs) {
815 if (so.isA(sf.getType(), type))
817 String parent = (String) sf.getValue(PARENT);
818 if (parent.equals(parentId))
829 * Maps exon features from dna to protein, and computes variants in peptide
830 * product generated by variants in dna, and adds them as sequence_variant
831 * features on the protein sequence. Returns the number of variant features
836 * @param dnaToProtein
838 static int computeProteinFeatures(SequenceI dnaSeq,
839 SequenceI peptide, MapList dnaToProtein)
841 while (dnaSeq.getDatasetSequence() != null)
843 dnaSeq = dnaSeq.getDatasetSequence();
845 while (peptide.getDatasetSequence() != null)
847 peptide = peptide.getDatasetSequence();
850 mapExonFeaturesToProtein(dnaSeq, peptide, dnaToProtein);
852 LinkedHashMap<Integer, String[][]> variants = buildDnaVariantsMap(
853 dnaSeq, dnaToProtein);
856 * scan codon variations, compute peptide variants and add to peptide sequence
859 for (Entry<Integer, String[][]> variant : variants.entrySet())
861 int peptidePos = variant.getKey();
862 String[][] codonVariants = variant.getValue();
863 String residue = String.valueOf(peptide.getCharAt(peptidePos - 1)); // 0-based
864 List<String> peptideVariants = computePeptideVariants(codonVariants,
866 if (!peptideVariants.isEmpty())
868 String desc = StringUtils.listToDelimitedString(peptideVariants,
870 SequenceFeature sf = new SequenceFeature(
871 SequenceOntology.SEQUENCE_VARIANT, desc, peptidePos,
872 peptidePos, 0f, null);
873 peptide.addSequenceFeature(sf);
881 * Transfers exon features to the corresponding mapped regions of the protein
882 * sequence. This is useful because it allows visualisation of exon boundaries
883 * on the peptide (using 'colour by label' for the exon name). Returns the
884 * number of features written.
888 * @param dnaToProtein
890 static int mapExonFeaturesToProtein(SequenceI dnaSeq, SequenceI peptide,
891 MapList dnaToProtein)
893 SequenceFeature[] sfs = dnaSeq.getSequenceFeatures();
899 SequenceOntology so = SequenceOntology.getInstance();
902 for (SequenceFeature sf : sfs)
904 if (so.isA(sf.getType(), SequenceOntology.EXON))
906 int start = sf.getBegin();
907 int end = sf.getEnd();
908 int[] mapsTo = dnaToProtein.locateInTo(start, end);
911 SequenceFeature copy = new SequenceFeature(SequenceOntology.EXON,
912 sf.getDescription(), mapsTo[0], mapsTo[1], 0f, null);
913 peptide.addSequenceFeature(copy);
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);