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.SequenceOntologyFactory;
15 import jalview.io.gff.SequenceOntologyI;
16 import jalview.schemes.ResidueProperties;
17 import jalview.util.DBRefUtils;
18 import jalview.util.MapList;
19 import jalview.util.MappingUtils;
20 import jalview.util.StringUtils;
22 import java.io.IOException;
23 import java.net.MalformedURLException;
25 import java.util.ArrayList;
26 import java.util.Arrays;
27 import java.util.Collections;
28 import java.util.Comparator;
29 import java.util.LinkedHashMap;
30 import java.util.List;
31 import java.util.Map.Entry;
34 * Base class for Ensembl sequence fetchers
36 * @see http://rest.ensembl.org/documentation/info/sequence_id
39 public abstract class EnsemblSeqProxy extends EnsemblRestClient
41 private static final List<String> CROSS_REFERENCES = Arrays
42 .asList(new String[] { "CCDS", "Uniprot/SWISSPROT" });
44 protected static final String CONSEQUENCE_TYPE = "consequence_type";
46 protected static final String PARENT = "Parent";
48 protected static final String ID = "ID";
50 protected static final String NAME = "Name";
52 protected static final String DESCRIPTION = "description";
55 * enum for 'type' parameter to the /sequence REST service
57 public enum EnsemblSeqType
60 * type=genomic to fetch full dna including introns
65 * type=cdna to fetch dna including UTRs
70 * type=cds to fetch coding dna excluding UTRs
75 * type=protein to fetch peptide product sequence
80 * the value of the 'type' parameter to fetch this version of
85 EnsemblSeqType(String t)
90 public String getType()
98 * A comparator to sort ranges into ascending start position order
100 private class RangeSorter implements Comparator<int[]>
104 RangeSorter(boolean forward)
110 public int compare(int[] o1, int[] o2)
112 return (forwards ? 1 : -1) * Integer.compare(o1[0], o2[0]);
118 * Default constructor (to use rest.ensembl.org)
120 public EnsemblSeqProxy()
126 * Constructor given the target domain to fetch data from
128 public EnsemblSeqProxy(String d)
134 * Makes the sequence queries to Ensembl's REST service and returns an
135 * alignment consisting of the returned sequences.
138 public AlignmentI getSequenceRecords(String query) throws Exception
140 // TODO use a String... query vararg instead?
142 // danger: accession separator used as a regex here, a string elsewhere
143 // in this case it is ok (it is just a space), but (e.g.) '\' would not be
144 List<String> allIds = Arrays.asList(query
145 .split(getAccessionSeparator()));
146 AlignmentI alignment = null;
150 * execute queries, if necessary in batches of the
151 * maximum allowed number of ids
153 int maxQueryCount = getMaximumQueryCount();
154 for (int v = 0, vSize = allIds.size(); v < vSize; v += maxQueryCount)
156 int p = Math.min(vSize, v + maxQueryCount);
157 List<String> ids = allIds.subList(v, p);
160 alignment = fetchSequences(ids, alignment);
161 } catch (Throwable r)
164 String msg = "Aborting ID retrieval after " + v
165 + " chunks. Unexpected problem (" + r.getLocalizedMessage()
167 System.err.println(msg);
172 if (alignment == null)
178 * fetch and transfer genomic sequence features,
179 * fetch protein product and add as cross-reference
181 for (String accId : allIds)
183 addFeaturesAndProduct(accId, alignment);
186 for (SequenceI seq : alignment.getSequences())
188 getCrossReferences(seq);
195 * Fetches Ensembl features using the /overlap REST endpoint, and adds them to
196 * the sequence in the alignment. Also fetches the protein product, maps it
197 * from the CDS features of the sequence, and saves it as a cross-reference of
203 protected void addFeaturesAndProduct(String accId, AlignmentI alignment)
205 if (alignment == null)
213 * get 'dummy' genomic sequence with exon, cds and variation features
215 SequenceI genomicSequence = null;
216 EnsemblFeatures gffFetcher = new EnsemblFeatures(getDomain());
217 EnsemblFeatureType[] features = getFeaturesToFetch();
218 AlignmentI geneFeatures = gffFetcher.getSequenceRecords(accId,
220 if (geneFeatures.getHeight() > 0)
222 genomicSequence = geneFeatures.getSequenceAt(0);
224 if (genomicSequence != null)
227 * transfer features to the query sequence
229 SequenceI querySeq = alignment.findName(accId);
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 addProteinProduct(querySeq);
240 } catch (IOException e)
242 System.err.println("Error transferring Ensembl features: "
248 * Returns those sequence feature types to fetch from Ensembl. We may want
249 * features either because they are of interest to the user, or as means to
250 * identify the locations of the sequence on the genomic sequence (CDS
251 * features identify CDS, exon features identify cDNA etc).
255 protected abstract EnsemblFeatureType[] getFeaturesToFetch();
258 * Fetches and maps the protein product, and adds it as a cross-reference of
259 * the retrieved sequence
261 protected void addProteinProduct(SequenceI querySeq)
263 String accId = querySeq.getName();
266 AlignmentI protein = new EnsemblProtein(getDomain())
267 .getSequenceRecords(accId);
268 if (protein == null || protein.getHeight() == 0)
270 System.out.println("Failed to retrieve protein for " + accId);
273 SequenceI proteinSeq = protein.getSequenceAt(0);
276 * need dataset sequences (to be the subject of mappings)
278 proteinSeq.createDatasetSequence();
279 querySeq.createDatasetSequence();
281 MapList mapList = mapCdsToProtein(querySeq, proteinSeq);
284 // clunky: ensure Uniprot xref if we have one is on mapped sequence
285 SequenceI ds = proteinSeq.getDatasetSequence();
286 ds.setSourceDBRef(proteinSeq.getSourceDBRef());
287 Mapping map = new Mapping(ds, mapList);
288 DBRefEntry dbr = new DBRefEntry(getDbSource(), getDbVersion(),
290 querySeq.getDatasetSequence().addDBRef(dbr);
293 * compute peptide variants from dna variants and add as
294 * sequence features on the protein sequence ta-da
296 computeProteinFeatures(querySeq, proteinSeq, mapList);
298 } catch (Exception e)
301 .println(String.format("Error retrieving protein for %s: %s",
302 accId, e.getMessage()));
307 * Get database xrefs from Ensembl, and attach them to the sequence
311 protected void getCrossReferences(SequenceI seq)
313 while (seq.getDatasetSequence() != null)
315 seq = seq.getDatasetSequence();
318 EnsemblXref xrefFetcher = new EnsemblXref(getDomain());
319 List<DBRefEntry> xrefs = xrefFetcher.getCrossReferences(seq.getName(),
320 getCrossReferenceDatabases());
321 for (DBRefEntry xref : xrefs)
325 * Save any Uniprot xref to be the reference for SIFTS mapping
327 if (DBRefSource.UNIPROT.equals(xref.getSource()))
329 seq.setSourceDBRef(xref);
335 * Returns a list of database names to be used when fetching cross-references.
339 protected List<String> getCrossReferenceDatabases()
341 return CROSS_REFERENCES;
345 * Returns a mapping from dna to protein by inspecting sequence features of
346 * type "CDS" on the dna.
352 protected MapList mapCdsToProtein(SequenceI dnaSeq, SequenceI proteinSeq)
354 List<int[]> ranges = getCdsRanges(dnaSeq);
355 int mappedDnaLength = MappingUtils.getLength(ranges);
357 int proteinLength = proteinSeq.getLength();
358 int proteinEnd = proteinLength;
359 int proteinStart = 1;
362 * incomplete start codon may mean X at start of peptide
363 * we ignore both for mapping purposes
365 if (proteinSeq.getCharAt(0) == 'X')
370 List<int[]> proteinRange = new ArrayList<int[]>();
373 * dna length should map to protein (or protein plus stop codon)
375 int codesForResidues = mappedDnaLength / 3;
376 if (codesForResidues == (proteinLength + 1))
378 // assuming extra codon is for STOP and not in peptide
381 if (codesForResidues == proteinLength)
383 proteinRange.add(new int[] { proteinStart, proteinEnd });
384 return new MapList(ranges, proteinRange, 3, 1);
390 * Returns a list of CDS ranges found.
392 * No need to worry about reverse strand dna, here since the retrieved
393 * sequence is as transcribed (reverse complement for reverse strand), i.e in
394 * the same sense as the peptide.
399 protected List<int[]> getCdsRanges(SequenceI dnaSeq)
401 List<int[]> result = new ArrayList<int[]>();
402 SequenceFeature[] sfs = dnaSeq.getSequenceFeatures();
407 SequenceOntologyI so = SequenceOntologyFactory.getInstance();
408 for (SequenceFeature sf : sfs)
411 * process a CDS feature (or a sub-type of CDS)
413 if (so.isA(sf.getType(), SequenceOntologyI.CDS))
417 phase = Integer.parseInt(sf.getPhase());
418 } catch (NumberFormatException e)
423 * phase > 0 on first codon means 5' incomplete - skip to the start
424 * of the next codon; example ENST00000496384
426 int begin = sf.getBegin();
427 int end = sf.getEnd();
428 if (result.isEmpty())
433 continue; // shouldn't happen?
436 result.add(new int[] { begin, end });
443 * Fetches sequences for the list of accession ids and adds them to the
444 * alignment. Returns the extended (or created) alignment.
449 * @throws JalviewException
450 * @throws IOException
452 protected AlignmentI fetchSequences(List<String> ids, AlignmentI alignment)
453 throws JalviewException, IOException
455 if (!isEnsemblAvailable())
458 throw new JalviewException("ENSEMBL Rest API not available.");
460 FileParse fp = getSequenceReader(ids);
461 FastaFile fr = new FastaFile(fp);
462 if (fr.hasWarningMessage())
464 System.out.println(String.format(
465 "Warning when retrieving %d ids %s\n%s", ids.size(),
466 ids.toString(), fr.getWarningMessage()));
468 else if (fr.getSeqs().size() != ids.size())
470 System.out.println(String.format(
471 "Only retrieved %d sequences for %d query strings", fr
472 .getSeqs().size(), ids.size()));
475 if (fr.getSeqs().size() == 1 && fr.getSeqs().get(0).getLength() == 0)
478 * POST request has returned an empty FASTA file e.g. for invalid id
480 throw new IOException("No data returned for " + ids);
483 if (fr.getSeqs().size() > 0)
485 AlignmentI seqal = new Alignment(
486 fr.getSeqsAsArray());
487 for (SequenceI sq:seqal.getSequences())
489 if (sq.getDescription() == null)
491 sq.setDescription(getDbName());
493 String name = sq.getName();
494 if (ids.contains(name)
495 || ids.contains(name.replace("ENSP", "ENST")))
497 DBRefUtils.parseToDbRef(sq, DBRefSource.ENSEMBL, "0", name);
500 if (alignment == null)
506 alignment.append(seqal);
513 * Returns the URL for the REST call
516 * @throws MalformedURLException
519 protected URL getUrl(List<String> ids) throws MalformedURLException
522 * a single id is included in the URL path
523 * multiple ids go in the POST body instead
525 StringBuffer urlstring = new StringBuffer(128);
526 urlstring.append(getDomain() + "/sequence/id");
529 urlstring.append("/").append(ids.get(0));
531 // @see https://github.com/Ensembl/ensembl-rest/wiki/Output-formats
532 urlstring.append("?type=").append(getSourceEnsemblType().getType());
533 urlstring.append(("&Accept=text/x-fasta"));
535 URL url = new URL(urlstring.toString());
540 * A sequence/id POST request currently allows up to 50 queries
542 * @see http://rest.ensembl.org/documentation/info/sequence_id_post
545 public int getMaximumQueryCount()
551 protected boolean useGetRequest()
557 protected String getRequestMimeType(boolean multipleIds)
559 return multipleIds ? "application/json" : "text/x-fasta";
563 protected String getResponseMimeType()
565 return "text/x-fasta";
570 * @return the configured sequence return type for this source
572 protected abstract EnsemblSeqType getSourceEnsemblType();
575 * Returns a list of [start, end] genomic ranges corresponding to the sequence
578 * The correspondence between the frames of reference is made by locating
579 * those features on the genomic sequence which identify the retrieved
580 * sequence. Specifically
582 * <li>genomic sequence is identified by "transcript" features with
583 * ID=transcript:transcriptId</li>
584 * <li>cdna sequence is identified by "exon" features with
585 * Parent=transcript:transcriptId</li>
586 * <li>cds sequence is identified by "CDS" features with
587 * Parent=transcript:transcriptId</li>
590 * The returned ranges are sorted to run forwards (for positive strand) or
591 * backwards (for negative strand). Aborts and returns null if both positive
592 * and negative strand are found (this should not normally happen).
594 * @param sourceSequence
597 * the start position of the sequence we are mapping to
600 protected MapList getGenomicRangesFromFeatures(SequenceI sourceSequence,
601 String accId, int start)
603 SequenceFeature[] sfs = sourceSequence.getSequenceFeatures();
610 * generously initial size for number of cds regions
611 * (worst case titin Q8WZ42 has c. 313 exons)
613 List<int[]> regions = new ArrayList<int[]>(100);
614 int mappedLength = 0;
615 int direction = 1; // forward
616 boolean directionSet = false;
618 for (SequenceFeature sf : sfs)
621 * accept the target feature type or a specialisation of it
622 * (e.g. coding_exon for exon)
624 if (identifiesSequence(sf, accId))
626 int strand = sf.getStrand();
627 strand = strand == 0 ? 1 : strand; // treat unknown as forward
629 if (directionSet && strand != direction)
631 // abort - mix of forward and backward
632 System.err.println("Error: forward and backward strand for "
640 * add to CDS ranges, semi-sorted forwards/backwards
644 regions.add(0, new int[] { sf.getEnd(), sf.getBegin() });
648 regions.add(new int[] { sf.getBegin(), sf.getEnd() });
650 mappedLength += Math.abs(sf.getEnd() - sf.getBegin() + 1);
655 * 'gene' sequence is contiguous so we can stop as soon as its
656 * identifying feature has been found
663 if (regions.isEmpty())
665 System.out.println("Failed to identify target sequence for " + accId
666 + " from genomic features");
671 * a final sort is needed since Ensembl returns CDS sorted within source
672 * (havana / ensembl_havana)
674 Collections.sort(regions, new RangeSorter(direction == 1));
676 List<int[]> to = Arrays.asList(new int[] { start,
677 start + mappedLength - 1 });
679 return new MapList(regions, to, 1, 1);
683 * Answers true if the sequence being retrieved may occupy discontiguous
684 * regions on the genomic sequence.
686 protected boolean isSpliceable()
692 * Returns true if the sequence feature marks positions of the genomic
693 * sequence feature which are within the sequence being retrieved. For
694 * example, an 'exon' feature whose parent is the target transcript marks the
695 * cdna positions of the transcript.
701 protected abstract boolean identifiesSequence(SequenceFeature sf,
705 * Transfers the sequence feature to the target sequence, locating its start
706 * and end range based on the mapping. Features which do not overlap the
707 * target sequence are ignored.
710 * @param targetSequence
712 * mapping from the sequence feature's coordinates to the target
715 protected void transferFeature(SequenceFeature sf,
716 SequenceI targetSequence, MapList mapping)
718 int start = sf.getBegin();
719 int end = sf.getEnd();
720 int[] mappedRange = mapping.locateInTo(start, end);
722 if (mappedRange != null)
724 SequenceFeature copy = new SequenceFeature(sf);
725 copy.setBegin(Math.min(mappedRange[0], mappedRange[1]));
726 copy.setEnd(Math.max(mappedRange[0], mappedRange[1]));
727 targetSequence.addSequenceFeature(copy);
730 * for sequence_variant, make an additional feature with consequence
732 // if (SequenceOntologyFactory.getInstance().isA(sf.getType(),
733 // SequenceOntologyI.SEQUENCE_VARIANT))
735 // String consequence = (String) sf.getValue(CONSEQUENCE_TYPE);
736 // if (consequence != null)
738 // SequenceFeature sf2 = new SequenceFeature("consequence",
739 // consequence, copy.getBegin(), copy.getEnd(), 0f,
741 // targetSequence.addSequenceFeature(sf2);
748 * Transfers features from sourceSequence to targetSequence
751 * @param sourceSequence
752 * @param targetSequence
753 * @return true if any features were transferred, else false
755 protected boolean transferFeatures(String accessionId,
756 SequenceI sourceSequence, SequenceI targetSequence)
758 if (sourceSequence == null || targetSequence == null)
763 // long start = System.currentTimeMillis();
764 SequenceFeature[] sfs = sourceSequence.getSequenceFeatures();
765 MapList mapping = getGenomicRangesFromFeatures(sourceSequence, accessionId,
766 targetSequence.getStart());
772 boolean result = transferFeatures(sfs, targetSequence, mapping,
774 // System.out.println("transferFeatures (" + (sfs.length) + " --> "
775 // + targetSequence.getSequenceFeatures().length + ") to "
776 // + targetSequence.getName()
777 // + " took " + (System.currentTimeMillis() - start) + "ms");
782 * Transfer features to the target sequence. The start/end positions are
783 * converted using the mapping. Features which do not overlap are ignored.
784 * Features whose parent is not the specified identifier are also ignored.
787 * @param targetSequence
792 protected boolean transferFeatures(SequenceFeature[] features,
793 SequenceI targetSequence, MapList mapping, String parentId)
795 final boolean forwardStrand = mapping.isFromForwardStrand();
798 * sort features by start position (descending if reverse strand)
799 * before transferring (in forwards order) to the target sequence
801 Arrays.sort(features, new Comparator<SequenceFeature>()
804 public int compare(SequenceFeature o1, SequenceFeature o2)
806 int c = Integer.compare(o1.getBegin(), o2.getBegin());
807 return forwardStrand ? c : -c;
811 boolean transferred = false;
812 for (SequenceFeature sf : features)
814 if (retainFeature(sf, parentId))
816 transferFeature(sf, targetSequence, mapping);
824 * Answers true if the feature type is one we want to keep for the sequence.
825 * Some features are only retrieved in order to identify the sequence range,
826 * and may then be discarded as redundant information (e.g. "CDS" feature for
829 @SuppressWarnings("unused")
830 protected boolean retainFeature(SequenceFeature sf, String accessionId)
832 return true; // override as required
836 * Answers true if the feature has a Parent which refers to the given
837 * accession id, or if the feature has no parent. Answers false if the
838 * feature's Parent is for a different accession id.
844 protected boolean featureMayBelong(SequenceFeature sf, String identifier)
846 String parent = (String) sf.getValue(PARENT);
847 // using contains to allow for prefix "gene:", "transcript:" etc
848 if (parent != null && !parent.contains(identifier))
850 // this genomic feature belongs to a different transcript
857 public String getDescription()
859 return "Ensembl " + getSourceEnsemblType().getType()
860 + " sequence with variant features";
864 * Returns a (possibly empty) list of features on the sequence which have the
865 * specified sequence ontology type (or a sub-type of it), and the given
866 * identifier as parent
873 protected List<SequenceFeature> findFeatures(SequenceI sequence,
874 String type, String parentId)
876 List<SequenceFeature> result = new ArrayList<SequenceFeature>();
878 SequenceFeature[] sfs = sequence.getSequenceFeatures();
880 SequenceOntologyI so = SequenceOntologyFactory.getInstance();
881 for (SequenceFeature sf :sfs) {
882 if (so.isA(sf.getType(), type))
884 String parent = (String) sf.getValue(PARENT);
885 if (parent.equals(parentId))
896 * Maps exon features from dna to protein, and computes variants in peptide
897 * product generated by variants in dna, and adds them as sequence_variant
898 * features on the protein sequence. Returns the number of variant features
903 * @param dnaToProtein
905 static int computeProteinFeatures(SequenceI dnaSeq,
906 SequenceI peptide, MapList dnaToProtein)
908 while (dnaSeq.getDatasetSequence() != null)
910 dnaSeq = dnaSeq.getDatasetSequence();
912 while (peptide.getDatasetSequence() != null)
914 peptide = peptide.getDatasetSequence();
917 AlignmentUtils.transferFeatures(dnaSeq, peptide, dnaToProtein,
918 SequenceOntologyI.EXON);
920 LinkedHashMap<Integer, String[][]> variants = buildDnaVariantsMap(
921 dnaSeq, dnaToProtein);
924 * scan codon variations, compute peptide variants and add to peptide sequence
927 for (Entry<Integer, String[][]> variant : variants.entrySet())
929 int peptidePos = variant.getKey();
930 String[][] codonVariants = variant.getValue();
931 String residue = String.valueOf(peptide.getCharAt(peptidePos - 1)); // 0-based
932 List<String> peptideVariants = computePeptideVariants(codonVariants,
934 if (!peptideVariants.isEmpty())
936 String desc = StringUtils.listToDelimitedString(peptideVariants,
938 SequenceFeature sf = new SequenceFeature(
939 SequenceOntologyI.SEQUENCE_VARIANT, desc, peptidePos,
940 peptidePos, 0f, null);
941 peptide.addSequenceFeature(sf);
947 * ugly sort to get sequence features in start position order
948 * - would be better to store in Sequence as a TreeSet instead?
950 Arrays.sort(peptide.getSequenceFeatures(),
951 new Comparator<SequenceFeature>()
954 public int compare(SequenceFeature o1, SequenceFeature o2)
956 int c = Integer.compare(o1.getBegin(), o2.getBegin());
957 return c == 0 ? Integer.compare(o1.getEnd(), o2.getEnd())
965 * Builds a map whose key is position in the protein sequence, and value is an
966 * array of all variants for the coding codon positions
969 * @param dnaToProtein
972 static LinkedHashMap<Integer, String[][]> buildDnaVariantsMap(
973 SequenceI dnaSeq, MapList dnaToProtein)
976 * map from peptide position to all variant features of the codon for it
977 * LinkedHashMap ensures we add the peptide features in sequence order
979 LinkedHashMap<Integer, String[][]> variants = new LinkedHashMap<Integer, String[][]>();
980 SequenceOntologyI so = SequenceOntologyFactory.getInstance();
982 SequenceFeature[] dnaFeatures = dnaSeq.getSequenceFeatures();
983 if (dnaFeatures == null)
988 int dnaStart = dnaSeq.getStart();
989 int[] lastCodon = null;
990 int lastPeptidePostion = 0;
993 * build a map of codon variations for peptides
995 for (SequenceFeature sf : dnaFeatures)
997 int dnaCol = sf.getBegin();
998 if (dnaCol != sf.getEnd())
1000 // not handling multi-locus variant features
1003 if (so.isA(sf.getType(), SequenceOntologyI.SEQUENCE_VARIANT))
1005 int[] mapsTo = dnaToProtein.locateInTo(dnaCol, dnaCol);
1008 // feature doesn't lie within coding region
1011 int peptidePosition = mapsTo[0];
1012 String[][] codonVariants = variants.get(peptidePosition);
1013 if (codonVariants == null)
1015 codonVariants = new String[3][];
1016 variants.put(peptidePosition, codonVariants);
1020 * extract dna variants to a string array
1022 String alls = (String) sf.getValue("alleles");
1027 String[] alleles = alls.split(",");
1030 * get this peptides codon positions e.g. [3, 4, 5] or [4, 7, 10]
1032 int[] codon = peptidePosition == lastPeptidePostion ? lastCodon
1033 : MappingUtils.flattenRanges(dnaToProtein.locateInFrom(
1034 peptidePosition, peptidePosition));
1035 lastPeptidePostion = peptidePosition;
1039 * save nucleotide (and this variant) for each codon position
1041 for (int codonPos = 0; codonPos < 3; codonPos++)
1043 String nucleotide = String.valueOf(dnaSeq
1044 .getCharAt(codon[codonPos] - dnaStart));
1045 if (codon[codonPos] == dnaCol)
1048 * record current dna base and its alleles
1050 String[] dnaVariants = new String[alleles.length + 1];
1051 dnaVariants[0] = nucleotide;
1052 System.arraycopy(alleles, 0, dnaVariants, 1, alleles.length);
1053 codonVariants[codonPos] = dnaVariants;
1055 else if (codonVariants[codonPos] == null)
1058 * record current dna base only
1059 * (at least until we find any variation and overwrite it)
1061 codonVariants[codonPos] = new String[] { nucleotide };
1070 * Returns a sorted, non-redundant list of all peptide translations generated
1071 * by the given dna variants, excluding the current residue value
1073 * @param codonVariants
1074 * an array of base values (acgtACGT) for codon positions 1, 2, 3
1076 * the current residue translation
1079 static List<String> computePeptideVariants(
1080 String[][] codonVariants, String residue)
1082 List<String> result = new ArrayList<String>();
1083 for (String base1 : codonVariants[0])
1085 for (String base2 : codonVariants[1])
1087 for (String base3 : codonVariants[2])
1089 String codon = base1 + base2 + base3;
1090 // TODO: report frameshift/insertion/deletion
1091 // and multiple-base variants?!
1092 String peptide = codon.contains("-") ? "-" : ResidueProperties
1093 .codonTranslate(codon);
1094 if (peptide != null && !result.contains(peptide)
1095 && !peptide.equalsIgnoreCase(residue))
1097 result.add(peptide);
1104 * sort alphabetically with STOP at the end
1106 Collections.sort(result, new Comparator<String>()
1110 public int compare(String o1, String o2)
1112 if ("STOP".equals(o1))
1116 else if ("STOP".equals(o2))
1122 return o1.compareTo(o2);
1130 * Answers true if the feature type is either 'NMD_transcript_variant' or
1131 * 'transcript' or one of its sub-types in the Sequence Ontology. This is
1132 * needed because NMD_transcript_variant behaves like 'transcript' in Ensembl
1133 * although strictly speaking it is not (it is a sub-type of
1134 * sequence_variant).
1136 * @param featureType
1139 public static boolean isTranscript(String featureType)
1141 return SequenceOntologyI.NMD_TRANSCRIPT_VARIANT.equals(featureType)
1142 || SequenceOntologyFactory.getInstance().isA(featureType,
1143 SequenceOntologyI.TRANSCRIPT);