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.datamodel.xdb.embl;
23 import jalview.analysis.SequenceIdMatcher;
24 import jalview.bin.Cache;
25 import jalview.datamodel.DBRefEntry;
26 import jalview.datamodel.DBRefSource;
27 import jalview.datamodel.FeatureProperties;
28 import jalview.datamodel.Mapping;
29 import jalview.datamodel.Sequence;
30 import jalview.datamodel.SequenceFeature;
31 import jalview.datamodel.SequenceI;
32 import jalview.util.DBRefUtils;
33 import jalview.util.DnaUtils;
34 import jalview.util.MapList;
35 import jalview.util.MappingUtils;
36 import jalview.util.StringUtils;
38 import java.text.ParseException;
39 import java.util.Arrays;
40 import java.util.Hashtable;
41 import java.util.List;
43 import java.util.Map.Entry;
44 import java.util.Vector;
45 import java.util.regex.Pattern;
48 * Data model for one entry returned from an EMBL query, as marshalled by a
51 * For example: http://www.ebi.ac.uk/ena/data/view/J03321&display=xml
53 * @see embl_mapping.xml
55 public class EmblEntry
57 private static final Pattern SPACE_PATTERN = Pattern.compile(" ");
63 String sequenceVersion;
71 String sequenceLength;
73 String taxonomicDivision;
77 String firstPublicDate;
79 String firstPublicRelease;
81 String lastUpdatedDate;
83 String lastUpdatedRelease;
85 Vector<String> keywords;
87 Vector<DBRefEntry> dbRefs;
89 Vector<EmblFeature> features;
91 EmblSequence sequence;
94 * @return the accession
96 public String getAccession()
103 * the accession to set
105 public void setAccession(String accession)
107 this.accession = accession;
113 public Vector<DBRefEntry> getDbRefs()
122 public void setDbRefs(Vector<DBRefEntry> dbRefs)
124 this.dbRefs = dbRefs;
128 * @return the features
130 public Vector<EmblFeature> getFeatures()
137 * the features to set
139 public void setFeatures(Vector<EmblFeature> features)
141 this.features = features;
145 * @return the keywords
147 public Vector<String> getKeywords()
154 * the keywords to set
156 public void setKeywords(Vector<String> keywords)
158 this.keywords = keywords;
162 * @return the sequence
164 public EmblSequence getSequence()
171 * the sequence to set
173 public void setSequence(EmblSequence sequence)
175 this.sequence = sequence;
179 * Recover annotated sequences from EMBL file
183 * a list of protein products found so far (to add to)
184 * @return dna dataset sequence with DBRefs and features
186 public SequenceI getSequence(String sourceDb, List<SequenceI> peptides)
188 SequenceI dna = makeSequence(sourceDb);
193 dna.setDescription(description);
194 DBRefEntry retrievedref = new DBRefEntry(sourceDb, getSequenceVersion(),
196 dna.addDBRef(retrievedref);
197 // add map to indicate the sequence is a valid coordinate frame for the
200 .setMap(new Mapping(null, new int[]
201 { 1, dna.getLength() }, new int[] { 1, dna.getLength() }, 1,
205 * transform EMBL Database refs to canonical form
209 for (DBRefEntry dbref : dbRefs)
211 dbref.setSource(DBRefUtils.getCanonicalName(dbref.getSource()));
216 SequenceIdMatcher matcher = new SequenceIdMatcher(peptides);
219 for (EmblFeature feature : features)
221 if (FeatureProperties.isCodingFeature(sourceDb, feature.getName()))
223 parseCodingFeature(feature, sourceDb, dna, peptides, matcher);
226 } catch (Exception e)
228 System.err.println("EMBL Record Features parsing error!");
230 .println("Please report the following to help@jalview.org :");
231 System.err.println("EMBL Record " + accession);
232 System.err.println("Resulted in exception: " + e.getMessage());
233 e.printStackTrace(System.err);
243 SequenceI makeSequence(String sourceDb)
245 if (sequence == null)
248 "No sequence was returned for ENA accession " + accession);
251 SequenceI dna = new Sequence(sourceDb + "|" + accession,
252 sequence.getSequence());
257 * Extracts coding region and product from a CDS feature and properly decorate
258 * it with annotations.
263 * source database for the EMBLXML
265 * parent dna sequence for this record
267 * list of protein product sequences for Embl entry
269 * helper to match xrefs in already retrieved sequences
271 void parseCodingFeature(EmblFeature feature, String sourceDb,
272 SequenceI dna, List<SequenceI> peptides,
273 SequenceIdMatcher matcher)
275 boolean isEmblCdna = sourceDb.equals(DBRefSource.EMBLCDS);
277 int[] exons = getCdsRanges(feature);
279 String translation = null;
280 String proteinName = "";
281 String proteinId = null;
282 Map<String, String> vals = new Hashtable<>();
285 * codon_start 1/2/3 in EMBL corresponds to phase 0/1/2 in CDS
286 * (phase is required for CDS features in GFF3 format)
291 * parse qualifiers, saving protein translation, protein id,
292 * codon start position, product (name), and 'other values'
294 if (feature.getQualifiers() != null)
296 for (Qualifier q : feature.getQualifiers())
298 String qname = q.getName();
299 if (qname.equals("translation"))
301 // remove all spaces (precompiled String.replaceAll(" ", ""))
302 translation = SPACE_PATTERN.matcher(q.getValues()[0])
305 else if (qname.equals("protein_id"))
307 proteinId = q.getValues()[0].trim();
309 else if (qname.equals("codon_start"))
313 codonStart = Integer.parseInt(q.getValues()[0].trim());
314 } catch (NumberFormatException e)
316 System.err.println("Invalid codon_start in XML for " + accession
317 + ": " + e.getMessage());
320 else if (qname.equals("product"))
322 // sometimes name is returned e.g. for V00488
323 proteinName = q.getValues()[0].trim();
327 // throw anything else into the additional properties hash
328 String[] qvals = q.getValues();
331 String commaSeparated = StringUtils.arrayToSeparatorList(qvals,
333 vals.put(qname, commaSeparated);
339 DBRefEntry proteinToEmblProteinRef = null;
340 exons = MappingUtils.removeStartPositions(codonStart - 1, exons);
342 SequenceI product = null;
343 Mapping dnaToProteinMapping = null;
344 if (translation != null && proteinName != null && proteinId != null)
346 int translationLength = translation.length();
349 * look for product in peptides list, if not found, add it
351 product = matcher.findIdMatch(proteinId);
354 product = new Sequence(proteinId, translation, 1,
356 product.setDescription(((proteinName.length() == 0)
357 ? "Protein Product from " + sourceDb
359 peptides.add(product);
360 matcher.add(product);
363 // we have everything - create the mapping and perhaps the protein
365 if (exons == null || exons.length == 0)
368 * workaround until we handle dna location for CDS sequence
369 * e.g. location="X53828.1:60..1058" correctly
372 "Implementation Notice: EMBLCDS records not properly supported yet - Making up the CDNA region of this sequence... may be incorrect ("
373 + sourceDb + ":" + getAccession() + ")");
374 if (translationLength
375 * 3 == (1 - codonStart + dna.getSequence().length))
378 "Not allowing for additional stop codon at end of cDNA fragment... !");
379 // this might occur for CDS sequences where no features are marked
380 exons = new int[] { dna.getStart() + (codonStart - 1),
382 dnaToProteinMapping = new Mapping(product, exons,
384 { 1, translationLength }, 3, 1);
386 if ((translationLength + 1)
387 * 3 == (1 - codonStart + dna.getSequence().length))
390 "Allowing for additional stop codon at end of cDNA fragment... will probably cause an error in VAMSAs!");
391 exons = new int[] { dna.getStart() + (codonStart - 1),
393 dnaToProteinMapping = new Mapping(product, exons,
395 { 1, translationLength }, 3, 1);
400 // Trim the exon mapping if necessary - the given product may only be a
401 // fragment of a larger protein. (EMBL:AY043181 is an example)
405 // TODO: Add a DbRef back to the parent EMBL sequence with the exon
407 // if given a dataset reference, search dataset for parent EMBL
408 // sequence if it exists and set its map
409 // make a new feature annotating the coding contig
413 // final product length truncation check
414 int[] cdsRanges = adjustForProteinLength(translationLength,
416 dnaToProteinMapping = new Mapping(product, cdsRanges,
418 { 1, translationLength }, 3, 1);
422 * make xref with mapping from protein to EMBL dna
424 DBRefEntry proteinToEmblRef = new DBRefEntry(DBRefSource.EMBL,
425 getSequenceVersion(), proteinId,
426 new Mapping(dnaToProteinMapping.getMap().getInverse()));
427 product.addDBRef(proteinToEmblRef);
430 * make xref from protein to EMBLCDS; we assume here that the
431 * CDS sequence version is same as dna sequence (?!)
433 MapList proteinToCdsMapList = new MapList(
435 { 1, translationLength },
437 { 1 + (codonStart - 1),
438 (codonStart - 1) + 3 * translationLength },
440 DBRefEntry proteinToEmblCdsRef = new DBRefEntry(
441 DBRefSource.EMBLCDS, getSequenceVersion(), proteinId,
442 new Mapping(proteinToCdsMapList));
443 product.addDBRef(proteinToEmblCdsRef);
446 * make 'direct' xref from protein to EMBLCDSPROTEIN
448 proteinToEmblProteinRef = new DBRefEntry(proteinToEmblCdsRef);
449 proteinToEmblProteinRef.setSource(DBRefSource.EMBLCDSProduct);
450 proteinToEmblProteinRef.setMap(null);
451 product.addDBRef(proteinToEmblProteinRef);
457 * add cds features to dna sequence
459 for (int xint = 0; exons != null && xint < exons.length; xint += 2)
461 SequenceFeature sf = makeCdsFeature(exons, xint, proteinName,
462 proteinId, vals, codonStart);
463 sf.setType(feature.getName()); // "CDS"
464 sf.setEnaLocation(feature.getLocation());
465 sf.setFeatureGroup(sourceDb);
466 dna.addSequenceFeature(sf);
471 * add feature dbRefs to sequence, and mappings for Uniprot xrefs
473 boolean hasUniprotDbref = false;
474 if (feature.dbRefs != null)
476 boolean mappingUsed = false;
477 for (DBRefEntry ref : feature.dbRefs)
480 * ensure UniProtKB/Swiss-Prot converted to UNIPROT
482 String source = DBRefUtils.getCanonicalName(ref.getSource());
483 ref.setSource(source);
484 DBRefEntry proteinDbRef = new DBRefEntry(ref.getSource(),
485 ref.getVersion(), ref.getAccessionId());
486 if (source.equals(DBRefSource.UNIPROT))
488 String proteinSeqName = DBRefSource.UNIPROT + "|"
489 + ref.getAccessionId();
490 if (dnaToProteinMapping != null
491 && dnaToProteinMapping.getTo() != null)
496 * two or more Uniprot xrefs for the same CDS -
497 * each needs a distinct Mapping (as to a different sequence)
499 dnaToProteinMapping = new Mapping(dnaToProteinMapping);
504 * try to locate the protein mapped to (possibly by a
505 * previous CDS feature); if not found, construct it from
506 * the EMBL translation
508 SequenceI proteinSeq = matcher.findIdMatch(proteinSeqName);
509 if (proteinSeq == null)
511 proteinSeq = new Sequence(proteinSeqName,
512 product.getSequenceAsString());
513 matcher.add(proteinSeq);
514 peptides.add(proteinSeq);
516 dnaToProteinMapping.setTo(proteinSeq);
517 dnaToProteinMapping.setMappedFromId(proteinId);
518 proteinSeq.addDBRef(proteinDbRef);
519 ref.setMap(dnaToProteinMapping);
521 hasUniprotDbref = true;
526 * copy feature dbref to our protein product
528 DBRefEntry pref = proteinDbRef;
529 pref.setMap(null); // reference is direct
530 product.addDBRef(pref);
531 // Add converse mapping reference
532 if (dnaToProteinMapping != null)
534 Mapping pmap = new Mapping(dna,
535 dnaToProteinMapping.getMap().getInverse());
536 pref = new DBRefEntry(sourceDb, getSequenceVersion(),
537 this.getAccession());
539 if (dnaToProteinMapping.getTo() != null)
541 dnaToProteinMapping.getTo().addDBRef(pref);
550 * if we have a product (translation) but no explicit Uniprot dbref
551 * (example: EMBL AAFI02000057 protein_id EAL65544.1)
552 * then construct mappings to an assumed EMBLCDSPROTEIN accession
554 if (!hasUniprotDbref && product != null)
556 if (proteinToEmblProteinRef == null)
558 // assuming CDSPROTEIN sequence version = dna version (?!)
559 proteinToEmblProteinRef = new DBRefEntry(DBRefSource.EMBLCDSProduct,
560 getSequenceVersion(), proteinId);
562 product.addDBRef(proteinToEmblProteinRef);
564 if (dnaToProteinMapping != null
565 && dnaToProteinMapping.getTo() != null)
567 DBRefEntry dnaToEmblProteinRef = new DBRefEntry(
568 DBRefSource.EMBLCDSProduct, getSequenceVersion(),
570 dnaToEmblProteinRef.setMap(dnaToProteinMapping);
571 dnaToProteinMapping.setMappedFromId(proteinId);
572 dna.addDBRef(dnaToEmblProteinRef);
578 * Helper method to construct a SequenceFeature for one cds range
581 * array of cds [start, end, ...] positions
582 * @param exonStartIndex
583 * offset into the exons array
585 * @param proteinAccessionId
587 * map of 'miscellaneous values' for feature
589 * codon start position for CDS (1/2/3, normally 1)
592 protected SequenceFeature makeCdsFeature(int[] exons, int exonStartIndex,
593 String proteinName, String proteinAccessionId,
594 Map<String, String> vals, int codonStart)
596 int exonNumber = exonStartIndex / 2 + 1;
597 SequenceFeature sf = new SequenceFeature();
598 sf.setBegin(Math.min(exons[exonStartIndex], exons[exonStartIndex + 1]));
599 sf.setEnd(Math.max(exons[exonStartIndex], exons[exonStartIndex + 1]));
600 sf.setDescription(String.format("Exon %d for protein '%s' EMBLCDS:%s",
601 exonNumber, proteinName, proteinAccessionId));
602 sf.setPhase(String.valueOf(codonStart - 1));
604 exons[exonStartIndex] <= exons[exonStartIndex + 1] ? "+" : "-");
605 sf.setValue(FeatureProperties.EXONPOS, exonNumber);
606 sf.setValue(FeatureProperties.EXONPRODUCT, proteinName);
609 StringBuilder sb = new StringBuilder();
610 boolean first = true;
611 for (Entry<String, String> val : vals.entrySet())
617 sb.append(val.getKey()).append("=").append(val.getValue());
619 sf.setValue(val.getKey(), val.getValue());
621 sf.setAttributes(sb.toString());
627 * Returns the CDS positions as a single array of [start, end, start, end...]
628 * positions. If on the reverse strand, these will be in descending order.
633 protected int[] getCdsRanges(EmblFeature feature)
635 if (feature.location == null)
642 List<int[]> ranges = DnaUtils.parseLocation(feature.location);
643 return listToArray(ranges);
644 } catch (ParseException e)
647 String.format("Not parsing inexact CDS location %s in ENA %s",
648 feature.location, this.accession));
654 * Converts a list of [start, end] ranges to a single array of [start, end,
660 int[] listToArray(List<int[]> ranges)
662 int[] result = new int[ranges.size() * 2];
664 for (int[] range : ranges)
666 result[i++] = range[0];
667 result[i++] = range[1];
673 * Truncates (if necessary) the exon intervals to match 3 times the length of
674 * the protein; also accepts 3 bases longer (for stop codon not included in
677 * @param proteinLength
679 * an array of [start, end, start, end...] intervals
680 * @return the same array (if unchanged) or a truncated copy
682 static int[] adjustForProteinLength(int proteinLength, int[] exon)
684 if (proteinLength <= 0 || exon == null)
688 int expectedCdsLength = proteinLength * 3;
689 int exonLength = MappingUtils.getLength(Arrays.asList(exon));
692 * if exon length matches protein, or is shorter, or longer by the
693 * length of a stop codon (3 bases), then leave it unchanged
695 if (expectedCdsLength >= exonLength
696 || expectedCdsLength == exonLength - 3)
704 origxon = new int[exon.length];
705 System.arraycopy(exon, 0, origxon, 0, exon.length);
707 for (int x = 0; x < exon.length; x += 2)
709 cdspos += Math.abs(exon[x + 1] - exon[x]) + 1;
710 if (expectedCdsLength <= cdspos)
712 // advanced beyond last codon.
714 if (expectedCdsLength != cdspos)
717 // .println("Truncating final exon interval on region by "
718 // + (cdspos - cdslength));
722 * shrink the final exon - reduce end position if forward
723 * strand, increase it if reverse
725 if (exon[x + 1] >= exon[x])
727 endxon = exon[x + 1] - cdspos + expectedCdsLength;
731 endxon = exon[x + 1] + cdspos - expectedCdsLength;
739 // and trim the exon interval set if necessary
740 int[] nxon = new int[sxpos + 2];
741 System.arraycopy(exon, 0, nxon, 0, sxpos + 2);
742 nxon[sxpos + 1] = endxon; // update the end boundary for the new exon
749 public String getSequenceVersion()
751 return sequenceVersion;
754 public void setSequenceVersion(String sequenceVersion)
756 this.sequenceVersion = sequenceVersion;
759 public String getSequenceLength()
761 return sequenceLength;
764 public void setSequenceLength(String sequenceLength)
766 this.sequenceLength = sequenceLength;
769 public String getEntryVersion()
774 public void setEntryVersion(String entryVersion)
776 this.entryVersion = entryVersion;
779 public String getMoleculeType()
784 public void setMoleculeType(String moleculeType)
786 this.moleculeType = moleculeType;
789 public String getTopology()
794 public void setTopology(String topology)
796 this.topology = topology;
799 public String getTaxonomicDivision()
801 return taxonomicDivision;
804 public void setTaxonomicDivision(String taxonomicDivision)
806 this.taxonomicDivision = taxonomicDivision;
809 public String getDescription()
814 public void setDescription(String description)
816 this.description = description;
819 public String getFirstPublicDate()
821 return firstPublicDate;
824 public void setFirstPublicDate(String firstPublicDate)
826 this.firstPublicDate = firstPublicDate;
829 public String getFirstPublicRelease()
831 return firstPublicRelease;
834 public void setFirstPublicRelease(String firstPublicRelease)
836 this.firstPublicRelease = firstPublicRelease;
839 public String getLastUpdatedDate()
841 return lastUpdatedDate;
844 public void setLastUpdatedDate(String lastUpdatedDate)
846 this.lastUpdatedDate = lastUpdatedDate;
849 public String getLastUpdatedRelease()
851 return lastUpdatedRelease;
854 public void setLastUpdatedRelease(String lastUpdatedRelease)
856 this.lastUpdatedRelease = lastUpdatedRelease;
859 public String getDataClass()
864 public void setDataClass(String dataClass)
866 this.dataClass = dataClass;