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 java.io.IOException;
24 import java.net.MalformedURLException;
26 import java.util.ArrayList;
27 import java.util.Arrays;
28 import java.util.Collections;
29 import java.util.List;
32 import org.json.simple.parser.ParseException;
34 import jalview.analysis.AlignmentUtils;
35 import jalview.analysis.Dna;
36 import jalview.bin.Cache;
37 import jalview.datamodel.Alignment;
38 import jalview.datamodel.AlignmentI;
39 import jalview.datamodel.DBRefEntry;
40 import jalview.datamodel.DBRefSource;
41 import jalview.datamodel.Mapping;
42 import jalview.datamodel.Sequence;
43 import jalview.datamodel.SequenceFeature;
44 import jalview.datamodel.SequenceI;
45 import jalview.datamodel.features.SequenceFeatures;
46 import jalview.exceptions.JalviewException;
47 import jalview.io.gff.Gff3Helper;
48 import jalview.io.gff.SequenceOntologyFactory;
49 import jalview.io.gff.SequenceOntologyI;
50 import jalview.util.Comparison;
51 import jalview.util.DBRefUtils;
52 import jalview.util.IntRangeComparator;
53 import jalview.util.MapList;
56 * Base class for Ensembl sequence fetchers
58 * @see http://rest.ensembl.org/documentation/info/sequence_id
61 public abstract class EnsemblSeqProxy extends EnsemblRestClient
63 protected static final String DESCRIPTION = "description";
66 * enum for 'type' parameter to the /sequence REST service
68 public enum EnsemblSeqType
71 * type=genomic to fetch full dna including introns
76 * type=cdna to fetch coding dna including UTRs
81 * type=cds to fetch coding dna excluding UTRs
86 * type=protein to fetch peptide product sequence
91 * the value of the 'type' parameter to fetch this version of
96 EnsemblSeqType(String t)
101 public String getType()
109 * Default constructor (to use rest.ensembl.org)
111 public EnsemblSeqProxy()
117 * Constructor given the target domain to fetch data from
119 public EnsemblSeqProxy(String d)
125 * Makes the sequence queries to Ensembl's REST service and returns an
126 * alignment consisting of the returned sequences.
129 public AlignmentI getSequenceRecords(String query) throws Exception
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
136 .asList(query.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);
164 if (alignment == null)
170 * fetch and transfer genomic sequence features,
171 * fetch protein product and add as cross-reference
173 for (int i = 0, n = allIds.size(); i < n; i++)
175 addFeaturesAndProduct(allIds.get(i), alignment);
178 List<SequenceI> seqs = alignment.getSequences();
179 for (int i = 0, n = seqs.size(); i < n; i++)
181 getCrossReferences(seqs.get(i));
188 * Fetches Ensembl features using the /overlap REST endpoint, and adds them to
189 * the sequence in the alignment. Also fetches the protein product, maps it
190 * from the CDS features of the sequence, and saves it as a cross-reference of
196 protected void addFeaturesAndProduct(String accId, AlignmentI alignment)
198 if (alignment == null)
206 * get 'dummy' genomic sequence with gene, transcript,
207 * exon, cds and variation features
209 SequenceI genomicSequence = null;
210 EnsemblFeatures gffFetcher = new EnsemblFeatures(getDomain());
211 EnsemblFeatureType[] features = getFeaturesToFetch();
213 // Platform.timeCheck("ESP.getsequencerec1", Platform.TIME_MARK);
215 AlignmentI geneFeatures = gffFetcher.getSequenceRecords(accId,
217 if (geneFeatures != null && geneFeatures.getHeight() > 0)
219 genomicSequence = geneFeatures.getSequenceAt(0);
222 // Platform.timeCheck("ESP.getsequencerec2", Platform.TIME_MARK);
224 if (genomicSequence != null)
227 * transfer features to the query sequence
229 SequenceI querySeq = alignment.findName(accId, true);
230 if (transferFeatures(accId, genomicSequence, querySeq))
234 * fetch and map protein product, and add it as a cross-reference
235 * of the retrieved sequence
237 // Platform.timeCheck("ESP.transferFeatures", Platform.TIME_MARK);
238 addProteinProduct(querySeq);
241 } catch (IOException e)
244 "Error transferring Ensembl features: " + e.getMessage());
246 // Platform.timeCheck("ESP.addfeat done", Platform.TIME_MARK);
250 * Returns those sequence feature types to fetch from Ensembl. We may want
251 * features either because they are of interest to the user, or as means to
252 * identify the locations of the sequence on the genomic sequence (CDS
253 * features identify CDS, exon features identify cDNA etc).
257 protected abstract EnsemblFeatureType[] getFeaturesToFetch();
260 * Fetches and maps the protein product, and adds it as a cross-reference of
261 * the retrieved sequence
263 protected void addProteinProduct(SequenceI querySeq)
265 String accId = querySeq.getName();
268 System.out.println("Adding protein product for " + accId);
269 AlignmentI protein = new EnsemblProtein(getDomain())
270 .getSequenceRecords(accId);
271 if (protein == null || protein.getHeight() == 0)
273 System.out.println("No protein product found for " + accId);
276 SequenceI proteinSeq = protein.getSequenceAt(0);
279 * need dataset sequences (to be the subject of mappings)
281 proteinSeq.createDatasetSequence();
282 querySeq.createDatasetSequence();
284 MapList mapList = AlignmentUtils.mapCdsToProtein(querySeq,
288 // clunky: ensure Uniprot xref if we have one is on mapped sequence
289 SequenceI ds = proteinSeq.getDatasetSequence();
290 // TODO: Verify ensp primary ref is on proteinSeq.getDatasetSequence()
291 Mapping map = new Mapping(ds, mapList);
292 DBRefEntry dbr = new DBRefEntry(getDbSource(),
293 getEnsemblDataVersion(), proteinSeq.getName(), map);
294 querySeq.getDatasetSequence().addDBRef(dbr);
295 List<DBRefEntry> uprots = DBRefUtils.selectRefs(ds.getDBRefs(),
297 { DBRefSource.UNIPROT });
298 List<DBRefEntry> upxrefs = DBRefUtils.selectRefs(querySeq.getDBRefs(),
300 { DBRefSource.UNIPROT });
303 for (DBRefEntry up : uprots)
305 // locate local uniprot ref and map
306 List<DBRefEntry> upx = DBRefUtils.searchRefs(upxrefs,
307 up.getAccessionId());
316 "Implementation issue - multiple uniprot acc on product sequence.");
321 upxref = new DBRefEntry(DBRefSource.UNIPROT,
322 getEnsemblDataVersion(), up.getAccessionId());
325 Mapping newMap = new Mapping(ds, mapList);
326 upxref.setVersion(getEnsemblDataVersion());
327 upxref.setMap(newMap);
330 // add the new uniprot ref
331 querySeq.getDatasetSequence().addDBRef(upxref);
338 * copy exon features to protein, compute peptide variants from dna
339 * variants and add as features on the protein sequence ta-da
341 // JAL-3187 render on the fly instead
342 // AlignmentUtils.computeProteinFeatures(querySeq, proteinSeq, mapList);
344 } catch (Exception e)
347 .println(String.format("Error retrieving protein for %s: %s",
348 accId, e.getMessage()));
353 * Get database xrefs from Ensembl, and attach them to the sequence
357 protected void getCrossReferences(SequenceI seq)
360 // Platform.timeCheck("ESP. getdataseq ", Platform.TIME_MARK);
363 while (seq.getDatasetSequence() != null)
365 seq = seq.getDatasetSequence();
368 // Platform.timeCheck("ESP. getxref ", Platform.TIME_MARK);
370 EnsemblXref xrefFetcher = new EnsemblXref(getDomain(), getDbSource(),
371 getEnsemblDataVersion());
372 List<DBRefEntry> xrefs = xrefFetcher.getCrossReferences(seq.getName());
374 for (int i = 0, n = xrefs.size(); i < n; i++)
376 // Platform.timeCheck("ESP. getxref + " + (i) + "/" + n, Platform.TIME_MARK);
377 // BH 2019.01.25 this next method was taking 174 ms PER addition for a 266-reference example.
378 // DBRefUtils.ensurePrimaries(seq)
379 // was at the end of seq.addDBRef, so executed after ever addition!
380 // This method was moved to seq.getPrimaryDBRefs()
381 seq.addDBRef(xrefs.get(i));
384 // System.out.println("primaries are " + seq.getPrimaryDBRefs().toString());
386 * and add a reference to itself
389 // Platform.timeCheck("ESP. getxref self ", Platform.TIME_MARK);
391 DBRefEntry self = new DBRefEntry(getDbSource(), getEnsemblDataVersion(),
394 // Platform.timeCheck("ESP. getxref self add ", Platform.TIME_MARK);
398 // Platform.timeCheck("ESP. seqprox done ", Platform.TIME_MARK);
402 * Fetches sequences for the list of accession ids and adds them to the
403 * alignment. Returns the extended (or created) alignment.
408 * @throws JalviewException
409 * @throws IOException
411 protected AlignmentI fetchSequences(List<String> ids,
412 AlignmentI alignment) throws JalviewException, IOException
414 if (!isEnsemblAvailable())
417 throw new JalviewException("ENSEMBL Rest API not available.");
419 // Platform.timeCheck("EnsemblSeqProx.fetchSeq ", Platform.TIME_MARK);
421 List<SequenceI> seqs = parseSequenceJson(ids);
427 throw new IOException("No data returned for " + ids);
430 if (seqs.size() != ids.size())
432 System.out.println(String.format(
433 "Only retrieved %d sequences for %d query strings",
434 seqs.size(), ids.size()));
439 AlignmentI seqal = new Alignment(
440 seqs.toArray(new SequenceI[seqs.size()]));
441 for (SequenceI seq : seqs)
443 if (seq.getDescription() == null)
445 seq.setDescription(getDbName());
447 String name = seq.getName();
448 if (ids.contains(name)
449 || ids.contains(name.replace("ENSP", "ENST")))
451 // TODO JAL-3077 use true accession version in dbref
452 DBRefEntry dbref = DBRefUtils.parseToDbRef(seq, getDbSource(),
453 getEnsemblDataVersion(), name);
457 if (alignment == null)
463 alignment.append(seqal);
470 * Parses a JSON response for a single sequence ID query
473 * @return a single jalview.datamodel.Sequence
474 * @see http://rest.ensembl.org/documentation/info/sequence_id
476 @SuppressWarnings("unchecked")
477 protected List<SequenceI> parseSequenceJson(List<String> ids)
479 List<SequenceI> result = new ArrayList<>();
483 * for now, assumes only one sequence returned; refactor if needed
484 * in future to handle a JSONArray with more than one
486 // Platform.timeCheck("ENS seqproxy", Platform.TIME_MARK);
487 Map<String, Object> val = (Map<String, Object>) getJSON(null, ids, -1, MODE_MAP, null);
490 Object s = val.get("desc");
491 String desc = s == null ? null : s.toString();
493 String id = s == null ? null : s.toString();
495 String seq = s == null ? null : s.toString();
496 Sequence sequence = new Sequence(id, seq);
499 sequence.setDescription(desc);
501 // todo JAL-3077 make a DBRefEntry with true accession version
502 // s = val.get("version");
503 // String version = s == null ? "0" : s.toString();
504 // DBRefEntry dbref = new DBRefEntry(getDbSource(), version, id);
505 // sequence.addDBRef(dbref);
506 result.add(sequence);
507 } catch (ParseException | IOException e)
509 System.err.println("Error processing JSON response: " + e.toString());
512 // Platform.timeCheck("ENS seqproxy2", Platform.TIME_MARK);
517 * Returns the URL for the REST call
520 * @throws MalformedURLException
523 protected URL getUrl(List<String> ids) throws MalformedURLException
526 * a single id is included in the URL path
527 * multiple ids go in the POST body instead
529 StringBuffer urlstring = new StringBuffer(128);
530 urlstring.append(getDomain() + "/sequence/id");
533 urlstring.append("/").append(ids.get(0));
535 // @see https://github.com/Ensembl/ensembl-rest/wiki/Output-formats
536 urlstring.append("?type=").append(getSourceEnsemblType().getType());
537 urlstring.append(("&Accept=application/json"));
538 urlstring.append(("&content-type=application/json"));
540 String objectType = getObjectType();
541 if (objectType != null)
543 urlstring.append("&").append(OBJECT_TYPE).append("=")
547 URL url = new URL(urlstring.toString());
552 * Override this method to specify object_type request parameter
556 protected String getObjectType()
562 * A sequence/id POST request currently allows up to 50 queries
564 * @see http://rest.ensembl.org/documentation/info/sequence_id_post
567 public int getMaximumQueryCount()
573 protected boolean useGetRequest()
580 * @return the configured sequence return type for this source
582 protected abstract EnsemblSeqType getSourceEnsemblType();
585 * Returns a list of [start, end] genomic ranges corresponding to the sequence
588 * The correspondence between the frames of reference is made by locating
589 * those features on the genomic sequence which identify the retrieved
590 * sequence. Specifically
592 * <li>genomic sequence is identified by "transcript" features with
593 * ID=transcript:transcriptId</li>
594 * <li>cdna sequence is identified by "exon" features with
595 * Parent=transcript:transcriptId</li>
596 * <li>cds sequence is identified by "CDS" features with
597 * Parent=transcript:transcriptId</li>
600 * The returned ranges are sorted to run forwards (for positive strand) or
601 * backwards (for negative strand). Aborts and returns null if both positive
602 * and negative strand are found (this should not normally happen).
604 * @param sourceSequence
607 * the start position of the sequence we are mapping to
610 protected MapList getGenomicRangesFromFeatures(SequenceI sourceSequence,
611 String accId, int start)
613 List<SequenceFeature> sfs = getIdentifyingFeatures(sourceSequence,
621 * generously initial size for number of cds regions
622 * (worst case titin Q8WZ42 has c. 313 exons)
624 List<int[]> regions = new ArrayList<>(100);
625 int mappedLength = 0;
626 int direction = 1; // forward
627 boolean directionSet = false;
629 for (SequenceFeature sf : sfs)
631 int strand = sf.getStrand();
632 strand = strand == 0 ? 1 : strand; // treat unknown as forward
634 if (directionSet && strand != direction)
636 // abort - mix of forward and backward
638 .println("Error: forward and backward strand for " + accId);
645 * add to CDS ranges, semi-sorted forwards/backwards
649 regions.add(0, new int[] { sf.getEnd(), sf.getBegin() });
653 regions.add(new int[] { sf.getBegin(), sf.getEnd() });
655 mappedLength += Math.abs(sf.getEnd() - sf.getBegin() + 1);
658 if (regions.isEmpty())
660 System.out.println("Failed to identify target sequence for " + accId
661 + " from genomic features");
666 * a final sort is needed since Ensembl returns CDS sorted within source
667 * (havana / ensembl_havana)
669 Collections.sort(regions, direction == 1 ? IntRangeComparator.ASCENDING
670 : IntRangeComparator.DESCENDING);
672 List<int[]> to = Arrays
674 { start, start + mappedLength - 1 });
676 return new MapList(regions, to, 1, 1);
680 * Answers a list of sequence features that mark positions of the genomic
681 * sequence feature which are within the sequence being retrieved. For
682 * example, an 'exon' feature whose parent is the target transcript marks the
683 * cdna positions of the transcript. For a gene sequence, this is trivially
684 * just the 'gene' feature with matching gene id.
690 protected abstract List<SequenceFeature> getIdentifyingFeatures(
691 SequenceI seq, String accId);
696 * Transfers the sequence feature to the target sequence, locating its start
697 * and end range based on the mapping. Features which do not overlap the
698 * target sequence are ignored.
701 * @param targetSequence
703 * mapping from the sequence feature's coordinates to the target
705 * @param forwardStrand
707 protected void transferFeature(SequenceFeature sf,
708 SequenceI targetSequence, MapList mapping, boolean forwardStrand)
710 int start = sf.getBegin();
711 int end = sf.getEnd();
712 int[] mappedRange = mapping.locateInTo(start, end);
714 if (mappedRange != null)
716 // Platform.timeCheck(null, Platform.TIME_SET);
717 String group = sf.getFeatureGroup();
718 if (".".equals(group))
720 group = getDbSource();
722 int newBegin = Math.min(mappedRange[0], mappedRange[1]);
723 int newEnd = Math.max(mappedRange[0], mappedRange[1]);
724 // Platform.timeCheck(null, Platform.TIME_MARK);
727 SequenceFeature copy = new SequenceFeature(sf, newBegin, newEnd, group, sf.getScore());
729 targetSequence.addSequenceFeature(copy);
732 * for sequence_variant on reverse strand, have to convert the allele
733 * values to their complements
735 if (!forwardStrand && SequenceOntologyFactory.getSequenceOntology()
736 .isA(sf.getType(), SequenceOntologyI.SEQUENCE_VARIANT))
738 reverseComplementAlleles(copy);
744 * Change the 'alleles' value of a feature by converting to complementary
745 * bases, and also update the feature description to match
749 static void reverseComplementAlleles(SequenceFeature sf)
751 final String alleles = (String) sf.getValue(Gff3Helper.ALLELES);
756 StringBuilder complement = new StringBuilder(alleles.length());
757 for (String allele : alleles.split(","))
759 reverseComplementAllele(complement, allele);
761 String comp = complement.toString();
762 sf.setValue(Gff3Helper.ALLELES, comp);
763 sf.setDescription(comp);
767 * Makes the 'reverse complement' of the given allele and appends it to the
768 * buffer, after a comma separator if not the first
773 static void reverseComplementAllele(StringBuilder complement,
776 if (complement.length() > 0)
778 complement.append(",");
782 * some 'alleles' are actually descriptive terms
783 * e.g. HGMD_MUTATION, PhenCode_variation
784 * - we don't want to 'reverse complement' these
786 if (!Comparison.isNucleotideSequence(allele, true))
788 complement.append(allele);
792 for (int i = allele.length() - 1; i >= 0; i--)
794 complement.append(Dna.getComplement(allele.charAt(i)));
800 * Transfers features from sourceSequence to targetSequence
803 * @param sourceSequence
804 * @param targetSequence
805 * @return true if any features were transferred, else false
807 protected boolean transferFeatures(String accessionId,
808 SequenceI sourceSequence, SequenceI targetSequence)
810 if (sourceSequence == null || targetSequence == null)
815 // long start = System.currentTimeMillis();
816 List<SequenceFeature> sfs = sourceSequence.getFeatures()
817 .getPositionalFeatures();
818 MapList mapping = getGenomicRangesFromFeatures(sourceSequence,
819 accessionId, targetSequence.getStart());
825 // Platform.timeCheck("ESP. xfer " + sfs.size(), Platform.TIME_MARK);
827 boolean result = transferFeatures(sfs, targetSequence, mapping,
829 // System.out.println("transferFeatures (" + (sfs.size()) + " --> "
830 // + targetSequence.getFeatures().getFeatureCount(true) + ") to "
831 // + targetSequence.getName() + " took "
832 // + (System.currentTimeMillis() - start) + "ms");
837 * Transfer features to the target sequence. The start/end positions are
838 * converted using the mapping. Features which do not overlap are ignored.
839 * Features whose parent is not the specified identifier are also ignored.
842 * @param targetSequence
847 protected boolean transferFeatures(List<SequenceFeature> sfs,
848 SequenceI targetSequence, MapList mapping, String parentId)
850 final boolean forwardStrand = mapping.isFromForwardStrand();
853 * sort features by start position (which corresponds to end
854 * position descending if reverse strand) so as to add them in
855 * 'forwards' order to the target sequence
857 SequenceFeatures.sortFeatures(sfs, forwardStrand);
859 boolean transferred = false;
861 for (int i = 0, n = sfs.size(); i < n; i++)
864 // if ((i%1000) == 0) {
865 //// Platform.timeCheck("Feature " + bhtest, Platform.TIME_GET);
866 // Platform.timeCheck("ESP. xferFeature + " + (i) + "/" + n, Platform.TIME_MARK);
869 SequenceFeature sf = sfs.get(i);
870 if (retainFeature(sf, parentId))
872 transferFeature(sf, targetSequence, mapping, forwardStrand);
881 * Answers true if the feature type is one we want to keep for the sequence.
882 * Some features are only retrieved in order to identify the sequence range,
883 * and may then be discarded as redundant information (e.g. "CDS" feature for
886 @SuppressWarnings("unused")
887 protected boolean retainFeature(SequenceFeature sf, String accessionId)
889 return true; // override as required
893 * Answers true if the feature has a Parent which refers to the given
894 * accession id, or if the feature has no parent. Answers false if the
895 * feature's Parent is for a different accession id.
901 protected boolean featureMayBelong(SequenceFeature sf, String identifier)
903 String parent = (String) sf.getValue(PARENT);
905 && !parent.equalsIgnoreCase(identifier))
907 // this genomic feature belongs to a different transcript
914 * Answers a short description of the sequence fetcher
917 public String getDescription()
919 return "Ensembl " + getSourceEnsemblType().getType()
920 + " sequence with variant features";
924 * Returns a (possibly empty) list of features on the sequence which have the
925 * specified sequence ontology term (or a sub-type of it), and the given
926 * identifier as parent
933 protected List<SequenceFeature> findFeatures(SequenceI sequence,
934 String term, String parentId)
936 List<SequenceFeature> result = new ArrayList<>();
938 List<SequenceFeature> sfs = sequence.getFeatures()
939 .getFeaturesByOntology(term);
940 for (SequenceFeature sf : sfs)
942 String parent = (String) sf.getValue(PARENT);
943 if (parent != null && parent.equalsIgnoreCase(parentId))
953 * Answers true if the feature type is either 'NMD_transcript_variant' or
954 * 'transcript' (or one of its sub-types in the Sequence Ontology). This is
955 * because NMD_transcript_variant behaves like 'transcript' in Ensembl
956 * although strictly speaking it is not (it is a sub-type of
959 * (This test was needed when fetching transcript features as GFF. As we are
960 * now fetching as JSON, all features have type 'transcript' so the check for
961 * NMD_transcript_variant is redundant. Left in for any future case arising.)
966 public static boolean isTranscript(String featureType)
968 return SequenceOntologyI.NMD_TRANSCRIPT_VARIANT.equals(featureType)
969 || SequenceOntologyFactory.getSequenceOntology().isA(featureType,
970 SequenceOntologyI.TRANSCRIPT);