matches = new ArrayList<>();
+ if (dbRefs != null && source != null)
{
- if (refa.getSource() != null && refb.getSource() != null
- && refb.getSource().equals(refa.getSource()))
+ for (DBRefEntry dbref : dbRefs)
{
- // We dont care about version
- // if ((refa.getVersion()==null || refb.getVersion()==null)
- // || refb.getVersion().equals(refa.getVersion()))
- // {
- if (refa.getAccessionId() != null && refb.getAccessionId() != null
- || refb.getAccessionId().equals(refa.getAccessionId()))
+ if (source.equalsIgnoreCase(dbref.getSource()))
{
- if (refa.getMap() == null && refb.getMap() == null)
- {
- return true;
- }
- if (refa.getMap() != null
- && refb.getMap() != null
- && ((refb.getMap().getMap() == null && refa.getMap()
- .getMap() == null) || (refb.getMap().getMap() != null
- && refa.getMap().getMap() != null && refb
- .getMap().getMap().equals(refa.getMap().getMap()))))
- {
- return true;
- }
+ matches.add(dbref);
}
}
- return false;
}
- };
+ return matches;
+ }
- /**
- * accession ID and DB must be identical. Version is ignored. No map on either
- * or map but no maplist on either or maplist of map on a is equivalent to the
- * maplist of map on b.
- */
- public static DbRefComp matchDbAndIdAndEitherMapOrEquivalentMapList = new DbRefComp()
+ private static Regex PARSE_REGEX;
+
+ private static Regex getParseRegex()
{
- public boolean matches(DBRefEntry refa, DBRefEntry refb)
+ return (PARSE_REGEX == null ? PARSE_REGEX = Platform.newRegex(
+ "([0-9][0-9A-Za-z]{3})\\s*(.?)\\s*;\\s*([0-9]+)-([0-9]+)", null)
+ : PARSE_REGEX);
+ }
+ /**
+ * Parses a DBRefEntry and adds it to the sequence, also a PDBEntry if the
+ * database is PDB.
+ *
+ * Used by file parsers to generate DBRefs from annotation within file (eg
+ * Stockholm)
+ *
+ * @param dbname
+ * @param version
+ * @param acn
+ * @param seq where to annotate with reference
+ * @return parsed version of entry that was added to seq (if any)
+ */
+ public static DBRefEntry parseToDbRef(SequenceI seq, String dbname, String version, String acn) {
+ DBRefEntry ref = null;
+ if (dbname != null) {
+ String locsrc = DBRefUtils.getCanonicalName(dbname);
+ if (locsrc.equals(DBRefSource.PDB)) {
+ /*
+ * Check for PFAM style stockhom PDB accession id citation e.g. "1WRI A; 7-80;"
+ */
+ Regex r = getParseRegex();
+ if (r.search(acn.trim())) {
+ String pdbid = r.stringMatched(1);
+ String chaincode = r.stringMatched(2);
+ if (chaincode == null) {
+ chaincode = " ";
+ }
+ // String mapstart = r.stringMatched(3);
+ // String mapend = r.stringMatched(4);
+ if (chaincode.equals(" ")) {
+ chaincode = "_";
+ }
+ // construct pdb ref.
+ ref = new DBRefEntry(locsrc, version, pdbid + chaincode);
+ PDBEntry pdbr = new PDBEntry();
+ pdbr.setId(pdbid);
+ pdbr.setType(PDBEntry.Type.PDB);
+ pdbr.setChainCode(chaincode);
+ seq.addPDBId(pdbr);
+ } else {
+ System.err.println("Malformed PDB DR line:" + acn);
+ }
+ } else {
+ // default:
+ ref = new DBRefEntry(locsrc, version, acn);
+ }
+ }
+ if (ref != null) {
+ seq.addDBRef(ref);
+ }
+ return ref;
+ }
+
+ /**
+ * Returns true if either object is null, or they are equal
+ *
+ * @param o1
+ * @param o2
+ * @return
+ */
+ public static boolean nullOrEqual(Object o1, Object o2) {
+ if (o1 == null || o2 == null) {
+ return true;
+ }
+ return o1.equals(o2);
+ }
+
+ /**
+ * canonicalise source string before comparing. null is always wildcard
+ *
+ * @param o1 - null or source string to compare
+ * @param o2 - null or source string to compare
+ * @return true if either o1 or o2 are null, or o1 equals o2 under
+ * DBRefUtils.getCanonicalName
+ * (o1).equals(DBRefUtils.getCanonicalName(o2))
+ */
+ public static boolean nullOrEqualSource(String o1, String o2) {
+ if (o1 == null || o2 == null) {
+ return true;
+ }
+ return DBRefUtils.getCanonicalName(o1).equals(DBRefUtils.getCanonicalName(o2));
+ }
+
+ /**
+ * Selects just the DNA or protein references from a set of references
+ *
+ * @param selectDna if true, select references to 'standard' DNA databases, else
+ * to 'standard' peptide databases
+ * @param refs a set of references to select from
+ * @return
+ */
+ public static List selectDbRefs(boolean selectDna, List refs) {
+ return selectRefs(refs, selectDna ? DBRefSource.DNACODINGDBS : DBRefSource.PROTEINDBS);
+ // could attempt to find other cross
+ // refs here - ie PDB xrefs
+ // (not dna, not protein seq)
+ }
+
+ /**
+ * Returns the (possibly empty) list of those supplied dbrefs which have the
+ * specified source database, with a case-insensitive match of source name
+ *
+ * @param dbRefs
+ * @param source
+ * @return
+ */
+ public static List searchRefsForSource(List dbRefs, String source) {
+ List matches = new ArrayList<>();
+ if (dbRefs != null && source != null) {
+ for (DBRefEntry dbref : dbRefs) {
+ if (source.equalsIgnoreCase(dbref.getSource())) {
+ matches.add(dbref);
+ }
+ }
+ }
+ return matches;
+ }
+
+ /**
+ * promote direct database references to primary for nucleotide or protein
+ * sequences if they have an appropriate primary ref
+ *
+ *
+ * Seq Type |
+ * Primary DB |
+ * Direct which will be promoted |
+ *
+ *
+ * peptides |
+ * Ensembl |
+ * Uniprot |
+ *
+ *
+ * peptides |
+ * Ensembl |
+ * Uniprot |
+ *
+ *
+ * dna |
+ * Ensembl |
+ * ENA |
+ *
+ *
+ *
+ * @param sequence
+ */
+ public static void ensurePrimaries(SequenceI sequence, List pr) {
+ if (pr.size() == 0) {
+ // nothing to do
+ return;
+ }
+ int sstart = sequence.getStart();
+ int send = sequence.getEnd();
+ boolean isProtein = sequence.isProtein();
+ BitSet bsSelect = new BitSet();
+
+// List selfs = new ArrayList();
+// {
+
+// List selddfs = selectDbRefs(!isprot, sequence.getDBRefs());
+// if (selfs == null || selfs.size() == 0)
+// {
+// // nothing to do
+// return;
+// }
+
+ List dbrefs = sequence.getDBRefs();
+ bsSelect.set(0, dbrefs.size());
+
+ if (!selectRefsBS(dbrefs, isProtein ? DBRefSource.PROTEIN_MASK : DBRefSource.DNA_CODING_MASK, bsSelect))
{
-// System.err.println("Comparing A: "+refa.getSrcAccString()+(refa.hasMap()?" has map.":"."));
-// System.err.println("Comparing B: "+refb.getSrcAccString()+(refb.hasMap()?" has map.":"."));
- if (refa.getSource() != null && refb.getSource() != null
- && refb.getSource().equals(refa.getSource()))
- {
- // We dont care about version
- // if ((refa.getVersion()==null || refb.getVersion()==null)
- // || refb.getVersion().equals(refa.getVersion()))
- // {
- if (refa.getAccessionId() != null && refb.getAccessionId() != null
- && refb.getAccessionId().equals(refa.getAccessionId()))
- {
- if (refa.getMap() == null || refb.getMap() == null)
- {
- return true;
- }
- if ((refa.getMap() != null && refb.getMap() != null)
- && (refb.getMap().getMap() == null && refa.getMap()
- .getMap() == null)
- || (refb.getMap().getMap() != null
- && refa.getMap().getMap() != null &&
- (refb
- .getMap().getMap().equals(refa.getMap().getMap())))) { // getMap().getMap().containsEither(false,refa.getMap().getMap())
- return true;
- }
- }
- }
- return false;
+ return;
}
- };
- /**
- * used by file parsers to generate DBRefs from annotation within file (eg
- * stockholm)
- *
- * @param dbname
- * @param version
- * @param acn
- * @param seq
- * where to anotate with reference
- * @return parsed version of entry that was added to seq (if any)
- */
- public static DBRefEntry parseToDbRef(SequenceI seq, String dbname,
- String version, String acn)
- {
- DBRefEntry ref = null;
- if (dbname != null)
- {
- String locsrc = jalview.util.DBRefUtils.getCanonicalName(dbname);
- if (locsrc.equals(jalview.datamodel.DBRefSource.PDB))
- {
- // check for chaincode and mapping
- // PFAM style stockhom PDB citation
- com.stevesoft.pat.Regex r = new com.stevesoft.pat.Regex(
- "([0-9][0-9A-Za-z]{3})\\s*(.?)\\s*;([0-9]+)-([0-9]+)");
- if (r.search(acn.trim()))
+// selfs.addAll(selfArray);
+// }
+
+ // filter non-primary refs
+ for (int ip = pr.size(); --ip >= 0;) {
+ DBRefEntry p = pr.get(ip);
+ for (int i = bsSelect.nextSetBit(0); i >= 0; i = bsSelect.nextSetBit(i + 1)) {
+ if (dbrefs.get(i) == p)
{
- String pdbid = r.stringMatched(1);
- String chaincode = r.stringMatched(2);
- String mapstart = r.stringMatched(3);
- String mapend = r.stringMatched(4);
- if (chaincode.equals(" "))
- {
- chaincode = "_";
- }
- // construct pdb ref.
- ref = new DBRefEntry(locsrc, version, pdbid + chaincode);
- PDBEntry pdbr = new PDBEntry();
- pdbr.setId(pdbid);
- seq.addPDBId(pdbr);
+ bsSelect.clear(i);
}
- }
- else
+ }
+// while (selfs.contains(p))
+// {
+// selfs.remove(p);
+// }
+ }
+// List toPromote = new ArrayList();
+
+ for (int ip = pr.size(), keys = 0; --ip >= 0 && keys != DBRefSource.PRIMARY_MASK;) {
+ DBRefEntry p = pr.get(ip);
+ if (isProtein) {
+ switch (getCanonicalName(p.getSource())) {
+ case DBRefSource.UNIPROT:
+ keys |= DBRefSource.UNIPROT_MASK;
+ break;
+ case DBRefSource.ENSEMBL:
+ keys |= DBRefSource.ENSEMBL_MASK;
+ break;
+ }
+ } else {
+ // TODO: promote transcript refs ??
+ }
+ if (keys == 0 || !selectRefsBS(dbrefs, keys, bsSelect))
{
- // default:
- ref = new DBRefEntry(locsrc, version, acn);
+ return;
}
- }
- if (ref != null)
- {
- seq.addDBRef(ref);
- }
- return ref;
- }
+// if (candidates != null)
+ {
+ for (int ic = bsSelect.nextSetBit(0); ic >= 0; ic = bsSelect.nextSetBit(ic + 1))
+// for (int ic = 0, n = candidates.size(); ic < n; ic++)
+ {
+ DBRefEntry cand = dbrefs.get(ic);// candidates.get(ic);
+ if (cand.hasMap()) {
+ Mapping map = cand.getMap();
+ SequenceI cto = map.getTo();
+ if (cto != null && cto != sequence) {
+ // can't promote refs with mappings to other sequences
+ continue;
+ }
+ MapList mlist = map.getMap();
+ if (mlist.getFromLowest() != sstart && mlist.getFromHighest() != send) {
+ // can't promote refs with mappings from a region of this sequence
+ // - eg CDS
+ continue;
+ }
+ }
+ // and promote - not that version must be non-null here,
+ // as p must have passed isPrimaryCandidate()
+ cand.setVersion(p.getVersion() + " (promoted)");
+ bsSelect.clear(ic);
+ // selfs.remove(cand);
+// toPromote.add(cand);
+ if (!cand.isPrimaryCandidate()) {
+ System.out.println("Warning: Couldn't promote dbref " + cand.toString() + " for sequence "
+ + sequence.toString());
+ }
+ }
+ }
+ }
+ }
}