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.
23 import java.util.Locale;
25 import java.util.ArrayList;
26 import java.util.BitSet;
27 import java.util.HashMap;
28 import java.util.HashSet;
29 import java.util.List;
32 import com.stevesoft.pat.Regex;
34 import jalview.datamodel.DBRefEntry;
35 import jalview.datamodel.DBRefSource;
36 import jalview.datamodel.Mapping;
37 import jalview.datamodel.PDBEntry;
38 import jalview.datamodel.SequenceI;
41 * Utilities for handling DBRef objects and their collections.
43 public class DBRefUtils
46 * lookup from lower-case form of a name to its canonical (standardised) form
48 private static Map<String, String> canonicalSourceNameLookup = new HashMap<>();
50 public final static int DB_SOURCE = 1;
52 public final static int DB_VERSION = 2;
54 public final static int DB_ID = 4;
56 public final static int DB_MAP = 8;
58 public final static int SEARCH_MODE_NO_MAP_NO_VERSION = DB_SOURCE | DB_ID;
60 public final static int SEARCH_MODE_FULL = DB_SOURCE | DB_VERSION | DB_ID
65 // TODO load these from a resource file?
66 canonicalSourceNameLookup.put("uniprotkb/swiss-prot",
68 canonicalSourceNameLookup.put("uniprotkb/trembl", DBRefSource.UNIPROT);
70 // Ensembl values for dbname in xref REST service:
71 canonicalSourceNameLookup.put("uniprot/sptrembl", DBRefSource.UNIPROT);
72 canonicalSourceNameLookup.put("uniprot/swissprot", DBRefSource.UNIPROT);
74 canonicalSourceNameLookup.put("pdb", DBRefSource.PDB);
75 canonicalSourceNameLookup.put("ensembl", DBRefSource.ENSEMBL);
76 // Ensembl Gn and Tr are for Ensembl genomic and transcript IDs as served
78 canonicalSourceNameLookup.put("ensembl-tr", DBRefSource.ENSEMBL);
79 canonicalSourceNameLookup.put("ensembl-gn", DBRefSource.ENSEMBL);
81 // guarantee we always have lowercase entries for canonical string lookups
82 for (String k : canonicalSourceNameLookup.keySet())
84 canonicalSourceNameLookup.put(k.toLowerCase(Locale.ROOT),
85 canonicalSourceNameLookup.get(k));
90 * Returns those DBRefEntry objects whose source identifier (once converted to
91 * Jalview's canonical form) is in the list of sources to search for. Returns
92 * null if no matches found.
95 * DBRefEntry objects to search
97 * array of sources to select
100 public static List<DBRefEntry> selectRefs(List<DBRefEntry> dbrefs,
103 if (dbrefs == null || sources == null)
109 HashSet<String> srcs = new HashSet<String>();
110 for (String src : sources)
112 srcs.add(src.toUpperCase(Locale.ROOT));
115 int nrefs = dbrefs.size();
116 List<DBRefEntry> res = new ArrayList<DBRefEntry>();
117 for (int ib = 0; ib < nrefs; ib++)
119 DBRefEntry dbr = dbrefs.get(ib);
120 String source = getCanonicalName(dbr.getSource());
121 if (srcs.contains(source.toUpperCase(Locale.ROOT)))
128 // List<DBRefEntry> reply = new DBRefEntry[res.size()];
129 return res;// .toArray(reply);
134 private static boolean selectRefsBS(List<DBRefEntry> dbrefs,
135 int sourceKeys, BitSet bsSelect)
137 if (dbrefs == null || sourceKeys == 0)
141 for (int i = 0, n = dbrefs.size(); i < n; i++)
143 DBRefEntry dbr = dbrefs.get(i);
144 if ((dbr.getSourceKey() & sourceKeys) != 0)
149 return !bsSelect.isEmpty();
153 * Returns a (possibly empty) list of those references that match the given
154 * entry, according to the given comparator.
157 * an array of database references to search
159 * an entry to compare against
163 static List<DBRefEntry> searchRefs(DBRefEntry[] refs, DBRefEntry entry,
164 DbRefComp comparator)
166 List<DBRefEntry> rfs = new ArrayList<>();
167 if (refs == null || entry == null)
171 for (int i = 0; i < refs.length; i++)
173 if (comparator.matches(entry, refs[i]))
182 * look up source in an internal list of database reference sources and return
183 * the canonical jalview name for the source, or the original string if it has
187 * @return canonical jalview source (one of jalview.datamodel.DBRefSource.*)
190 public static String getCanonicalName(String source)
196 String canonical = canonicalSourceNameLookup
197 .get(source.toLowerCase(Locale.ROOT));
198 return canonical == null ? source : canonical;
202 * Returns a (possibly empty) list of those references that match the given
203 * entry. Currently uses a comparator which matches if
205 * <li>database sources are the same</li>
206 * <li>accession ids are the same</li>
207 * <li>both have no mapping, or the mappings are the same</li>
211 * Set of references to search
215 * SEARCH_MODE_FULL for all; SEARCH_MODE_NO_MAP_NO_VERSION optional
218 public static List<DBRefEntry> searchRefs(List<DBRefEntry> ref,
219 DBRefEntry entry, int mode)
221 return searchRefs(ref, entry,
222 matchDbAndIdAndEitherMapOrEquivalentMapList, mode);
226 * Returns a list of those references that match the given accession id
228 * <li>database sources are the same</li>
229 * <li>accession ids are the same</li>
230 * <li>both have no mapping, or the mappings are the same</li>
234 * Set of references to search
236 * accession id to match
239 public static List<DBRefEntry> searchRefs(List<DBRefEntry> refs,
242 List<DBRefEntry> rfs = new ArrayList<DBRefEntry>();
243 if (refs == null || accId == null)
247 for (int i = 0, n = refs.size(); i < n; i++)
249 DBRefEntry e = refs.get(i);
250 if (accId.equals(e.getAccessionId()))
256 // return searchRefs(refs, new DBRefEntry("", "", accId), matchId,
257 // SEARCH_MODE_FULL);
261 * Returns a (possibly empty) list of those references that match the given
262 * entry, according to the given comparator.
265 * an array of database references to search
267 * an entry to compare against
270 * SEARCH_MODE_FULL for all; SEARCH_MODE_NO_MAP_NO_VERSION optional
273 static List<DBRefEntry> searchRefs(List<DBRefEntry> refs,
274 DBRefEntry entry, DbRefComp comparator, int mode)
276 List<DBRefEntry> rfs = new ArrayList<DBRefEntry>();
277 if (refs == null || entry == null)
281 for (int i = 0, n = refs.size(); i < n; i++)
283 DBRefEntry e = refs.get(i);
284 if (comparator.matches(entry, e, SEARCH_MODE_FULL))
294 default public boolean matches(DBRefEntry refa, DBRefEntry refb)
296 return matches(refa, refb, SEARCH_MODE_FULL);
299 public boolean matches(DBRefEntry refa, DBRefEntry refb, int mode);
303 * match on all non-null fields in refa
305 // TODO unused - remove? would be broken by equating "" with null
306 public static DbRefComp matchNonNullonA = new DbRefComp()
309 public boolean matches(DBRefEntry refa, DBRefEntry refb, int mode)
311 if ((mode & DB_SOURCE) != 0 && (refa.getSource() == null
312 || DBRefUtils.getCanonicalName(refb.getSource()).equals(
313 DBRefUtils.getCanonicalName(refa.getSource()))))
315 if ((mode & DB_VERSION) != 0 && (refa.getVersion() == null
316 || refb.getVersion().equals(refa.getVersion())))
318 if ((mode & DB_ID) != 0 && (refa.getAccessionId() == null
319 || refb.getAccessionId().equals(refa.getAccessionId())))
321 if ((mode & DB_MAP) != 0
322 && (refa.getMap() == null || (refb.getMap() != null
323 && refb.getMap().equals(refa.getMap()))))
335 * either field is null or field matches for all of source, version, accession
338 // TODO unused - remove?
339 public static DbRefComp matchEitherNonNull = new DbRefComp()
342 public boolean matches(DBRefEntry refa, DBRefEntry refb, int mode)
344 if (nullOrEqualSource(refa.getSource(), refb.getSource())
345 && nullOrEqual(refa.getVersion(), refb.getVersion())
346 && nullOrEqual(refa.getAccessionId(), refb.getAccessionId())
347 && nullOrEqual(refa.getMap(), refb.getMap()))
356 private static Regex PARSE_REGEX;
358 private static Regex getParseRegex()
360 return (PARSE_REGEX == null ? PARSE_REGEX = Platform.newRegex(
361 "([0-9][0-9A-Za-z]{3})\\s*(.?)\\s*;\\s*([0-9]+)-([0-9]+)")
366 * Parses a DBRefEntry and adds it to the sequence, also a PDBEntry if the
369 * Used by file parsers to generate DBRefs from annotation within file (eg
376 * where to annotate with reference
377 * @return parsed version of entry that was added to seq (if any)
379 public static DBRefEntry parseToDbRef(SequenceI seq, String dbname,
380 String version, String acn)
382 DBRefEntry ref = null;
385 String locsrc = DBRefUtils.getCanonicalName(dbname);
386 if (locsrc.equals(DBRefSource.PDB))
389 * Check for PFAM style stockhom PDB accession id citation e.g.
392 Regex r = getParseRegex();
393 if (r.search(acn.trim()))
395 String pdbid = r.stringMatched(1);
396 String chaincode = r.stringMatched(2);
397 if (chaincode == null)
401 // String mapstart = r.stringMatched(3);
402 // String mapend = r.stringMatched(4);
403 if (chaincode.equals(" "))
407 // construct pdb ref.
408 ref = new DBRefEntry(locsrc, version, pdbid + chaincode);
409 PDBEntry pdbr = new PDBEntry();
411 pdbr.setType(PDBEntry.Type.PDB);
412 pdbr.setChainCode(chaincode);
417 System.err.println("Malformed PDB DR line:" + acn);
423 ref = new DBRefEntry(locsrc, version, acn.trim());
434 * accession ID and DB must be identical. Version is ignored. Map is either
435 * not defined or is a match (or is compatible?)
437 // TODO unused - remove?
438 public static DbRefComp matchDbAndIdAndEitherMap = new DbRefComp()
441 public boolean matches(DBRefEntry refa, DBRefEntry refb, int mode)
443 if (refa.getSource() != null && refb.getSource() != null
444 && DBRefUtils.getCanonicalName(refb.getSource()).equals(
445 DBRefUtils.getCanonicalName(refa.getSource())))
447 // We dont care about version
448 if (refa.getAccessionId() != null && refb.getAccessionId() != null
449 // FIXME should be && not || here?
450 || refb.getAccessionId().equals(refa.getAccessionId()))
452 if ((refa.getMap() == null || refb.getMap() == null)
453 || (refa.getMap() != null && refb.getMap() != null
454 && refb.getMap().equals(refa.getMap())))
465 * accession ID and DB must be identical. Version is ignored. No map on either
466 * or map but no maplist on either or maplist of map on a is the complement of
467 * maplist of map on b.
469 // TODO unused - remove?
470 public static DbRefComp matchDbAndIdAndComplementaryMapList = new DbRefComp()
473 public boolean matches(DBRefEntry refa, DBRefEntry refb, int mode)
475 if (refa.getSource() != null && refb.getSource() != null
476 && DBRefUtils.getCanonicalName(refb.getSource()).equals(
477 DBRefUtils.getCanonicalName(refa.getSource())))
479 // We dont care about version
480 if (refa.getAccessionId() != null && refb.getAccessionId() != null
481 || refb.getAccessionId().equals(refa.getAccessionId()))
483 if ((refa.getMap() == null && refb.getMap() == null)
484 || (refa.getMap() != null && refb.getMap() != null))
486 if ((refb.getMap().getMap() == null
487 && refa.getMap().getMap() == null)
488 || (refb.getMap().getMap() != null
489 && refa.getMap().getMap() != null
490 && refb.getMap().getMap().getInverse()
491 .equals(refa.getMap().getMap())))
503 * accession ID and DB must be identical. Version is ignored. No map on both
504 * or or map but no maplist on either or maplist of map on a is equivalent to
505 * the maplist of map on b.
507 // TODO unused - remove?
508 public static DbRefComp matchDbAndIdAndEquivalentMapList = new DbRefComp()
511 public boolean matches(DBRefEntry refa, DBRefEntry refb, int mode)
513 if (refa.getSource() != null && refb.getSource() != null
514 && DBRefUtils.getCanonicalName(refb.getSource()).equals(
515 DBRefUtils.getCanonicalName(refa.getSource())))
517 // We dont care about version
518 // if ((refa.getVersion()==null || refb.getVersion()==null)
519 // || refb.getVersion().equals(refa.getVersion()))
521 if (refa.getAccessionId() != null && refb.getAccessionId() != null
522 || refb.getAccessionId().equals(refa.getAccessionId()))
524 if (refa.getMap() == null && refb.getMap() == null)
528 if (refa.getMap() != null && refb.getMap() != null
529 && ((refb.getMap().getMap() == null
530 && refa.getMap().getMap() == null)
531 || (refb.getMap().getMap() != null
532 && refa.getMap().getMap() != null
533 && refb.getMap().getMap()
534 .equals(refa.getMap().getMap()))))
545 * accession ID and DB must be identical, or null on a. Version is ignored. No
546 * map on either or map but no maplist on either or maplist of map on a is
547 * equivalent to the maplist of map on b.
549 public static DbRefComp matchDbAndIdAndEitherMapOrEquivalentMapList = new DbRefComp()
552 public boolean matches(DBRefEntry refa, DBRefEntry refb, int mode)
554 if (refa.getSource() != null && refb.getSource() != null
555 && DBRefUtils.getCanonicalName(refb.getSource()).equals(
556 DBRefUtils.getCanonicalName(refa.getSource())))
558 // We dont care about version
559 if (refa.getAccessionId() == null
560 || refa.getAccessionId().equals(refb.getAccessionId()))
562 if (refa.getMap() == null || refb.getMap() == null)
566 if ((refa.getMap() != null && refb.getMap() != null)
567 && (refb.getMap().getMap() == null
568 && refa.getMap().getMap() == null)
569 || (refb.getMap().getMap() != null
570 && refa.getMap().getMap() != null
571 && (refb.getMap().getMap()
572 .equals(refa.getMap().getMap()))))
583 * Returns the (possibly empty) list of those supplied dbrefs which have the
584 * specified source database, with a case-insensitive match of source name
590 public static List<DBRefEntry> searchRefsForSource(DBRefEntry[] dbRefs,
593 List<DBRefEntry> matches = new ArrayList<>();
594 if (dbRefs != null && source != null)
596 for (DBRefEntry dbref : dbRefs)
598 if (source.equalsIgnoreCase(dbref.getSource()))
608 * Returns true if either object is null, or they are equal
614 public static boolean nullOrEqual(Object o1, Object o2)
616 if (o1 == null || o2 == null)
620 return o1.equals(o2);
624 * canonicalise source string before comparing. null is always wildcard
627 * - null or source string to compare
629 * - null or source string to compare
630 * @return true if either o1 or o2 are null, or o1 equals o2 under
631 * DBRefUtils.getCanonicalName
632 * (o1).equals(DBRefUtils.getCanonicalName(o2))
634 public static boolean nullOrEqualSource(String o1, String o2)
636 if (o1 == null || o2 == null)
640 return DBRefUtils.getCanonicalName(o1)
641 .equals(DBRefUtils.getCanonicalName(o2));
645 * Selects just the DNA or protein references from a set of references
648 * if true, select references to 'standard' DNA databases, else to
649 * 'standard' peptide databases
651 * a set of references to select from
654 public static List<DBRefEntry> selectDbRefs(boolean selectDna,
655 List<DBRefEntry> refs)
657 return selectRefs(refs,
658 selectDna ? DBRefSource.DNACODINGDBS : DBRefSource.PROTEINDBS);
659 // could attempt to find other cross
660 // refs here - ie PDB xrefs
661 // (not dna, not protein seq)
665 * Returns the (possibly empty) list of those supplied dbrefs which have the
666 * specified source database, with a case-insensitive match of source name
672 public static List<DBRefEntry> searchRefsForSource(
673 List<DBRefEntry> dbRefs, String source)
675 List<DBRefEntry> matches = new ArrayList<DBRefEntry>();
676 if (dbRefs != null && source != null)
678 for (DBRefEntry dbref : dbRefs)
680 if (source.equalsIgnoreCase(dbref.getSource()))
690 * promote direct database references to primary for nucleotide or protein
691 * sequences if they have an appropriate primary ref
695 * <th>Primary DB</th>
696 * <th>Direct which will be promoted</th>
717 public static void ensurePrimaries(SequenceI sequence,
725 int sstart = sequence.getStart();
726 int send = sequence.getEnd();
727 boolean isProtein = sequence.isProtein();
728 BitSet bsSelect = new BitSet();
730 // List<DBRefEntry> selfs = new ArrayList<DBRefEntry>();
733 // List<DBRefEntry> selddfs = selectDbRefs(!isprot, sequence.getDBRefs());
734 // if (selfs == null || selfs.size() == 0)
740 List<DBRefEntry> dbrefs = sequence.getDBRefs();
741 bsSelect.set(0, dbrefs.size());
743 if (!selectRefsBS(dbrefs, isProtein ? DBRefSource.PROTEIN_MASK
744 : DBRefSource.DNA_CODING_MASK, bsSelect))
747 // selfs.addAll(selfArray);
750 // filter non-primary refs
751 for (int ip = pr.size(); --ip >= 0;)
753 DBRefEntry p = pr.get(ip);
754 for (int i = bsSelect.nextSetBit(0); i >= 0; i = bsSelect
757 if (dbrefs.get(i) == p)
760 // while (selfs.contains(p))
765 // List<DBRefEntry> toPromote = new ArrayList<DBRefEntry>();
767 for (int ip = pr.size(), keys = 0; --ip >= 0
768 && keys != DBRefSource.PRIMARY_MASK;)
770 DBRefEntry p = pr.get(ip);
773 switch (getCanonicalName(p.getSource()))
775 case DBRefSource.UNIPROT:
776 keys |= DBRefSource.UNIPROT_MASK;
778 case DBRefSource.ENSEMBL:
779 keys |= DBRefSource.ENSEMBL_MASK;
785 // TODO: promote transcript refs ??
787 if (keys == 0 || !selectRefsBS(dbrefs, keys, bsSelect))
789 // if (candidates != null)
791 for (int ic = bsSelect.nextSetBit(0); ic >= 0; ic = bsSelect
793 // for (int ic = 0, n = candidates.size(); ic < n; ic++)
795 DBRefEntry cand = dbrefs.get(ic);// candidates.get(ic);
798 Mapping map = cand.getMap();
799 SequenceI cto = map.getTo();
800 if (cto != null && cto != sequence)
802 // can't promote refs with mappings to other sequences
805 MapList mlist = map.getMap();
806 if (mlist.getFromLowest() != sstart
807 && mlist.getFromHighest() != send)
809 // can't promote refs with mappings from a region of this sequence
814 // and promote - not that version must be non-null here,
815 // as p must have passed isPrimaryCandidate()
816 cand.setVersion(p.getVersion() + " (promoted)");
818 // selfs.remove(cand);
819 // toPromote.add(cand);
820 if (!cand.isPrimaryCandidate())
823 "Warning: Couldn't promote dbref " + cand.toString()
824 + " for sequence " + sequence.toString());