+
+ /**
+ * 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<DBRefEntry> selectDbRefs(boolean selectDna,
+ List<DBRefEntry> 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<DBRefEntry> searchRefsForSource(
+ List<DBRefEntry> dbRefs, String source)
+ {
+ List<DBRefEntry> matches = new ArrayList<DBRefEntry>();
+ 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
+ * <table>
+ * <tr>
+ * <th>Seq Type</th>
+ * <th>Primary DB</th>
+ * <th>Direct which will be promoted</th>
+ * </tr>
+ * <tr align=center>
+ * <td>peptides</td>
+ * <td>Ensembl</td>
+ * <td>Uniprot</td>
+ * </tr>
+ * <tr align=center>
+ * <td>peptides</td>
+ * <td>Ensembl</td>
+ * <td>Uniprot</td>
+ * </tr>
+ * <tr align=center>
+ * <td>dna</td>
+ * <td>Ensembl</td>
+ * <td>ENA</td>
+ * </tr>
+ * </table>
+ *
+ * @param sequence
+ */
+ public static void ensurePrimaries(SequenceI sequence,
+ List<DBRefEntry> 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<DBRefEntry> selfs = new ArrayList<DBRefEntry>();
+ // {
+
+ // List<DBRefEntry> selddfs = selectDbRefs(!isprot, sequence.getDBRefs());
+ // if (selfs == null || selfs.size() == 0)
+ // {
+ // // nothing to do
+ // return;
+ // }
+
+ List<DBRefEntry> dbrefs = sequence.getDBRefs();
+ bsSelect.set(0, dbrefs.size());
+
+ if (!selectRefsBS(dbrefs, isProtein ? DBRefSource.PROTEIN_MASK
+ : DBRefSource.DNA_CODING_MASK, bsSelect))
+ return;
+
+ // 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)
+ bsSelect.clear(i);
+ }
+ // while (selfs.contains(p))
+ // {
+ // selfs.remove(p);
+ // }
+ }
+ // List<DBRefEntry> toPromote = new ArrayList<DBRefEntry>();
+
+ 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))
+ return;
+ // 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(cand.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());
+ }
+ }
+ }
+ }
+ }
+