X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Futil%2FDBRefUtils.java;fp=src%2Fjalview%2Futil%2FDBRefUtils.java;h=ae0243e9ce2de227f76c812b580a5f4f8e3aa944;hb=4f30214e8098748469c6a4269ac2ed6c5750e4b0;hp=4d5a0250b24c403591ae80f5b6b6e6bddb1924f0;hpb=9dabc02511e3a334a5749a504f57f69d6c9017bd;p=jalview.git diff --git a/src/jalview/util/DBRefUtils.java b/src/jalview/util/DBRefUtils.java index 4d5a025..ae0243e 100755 --- a/src/jalview/util/DBRefUtils.java +++ b/src/jalview/util/DBRefUtils.java @@ -20,21 +20,21 @@ */ package jalview.util; -import jalview.datamodel.DBRefEntry; -import jalview.datamodel.DBRefSource; -import jalview.datamodel.PDBEntry; -import jalview.datamodel.SequenceI; - import java.util.ArrayList; -import java.util.Arrays; +import java.util.BitSet; import java.util.HashMap; import java.util.HashSet; import java.util.List; import java.util.Map; -import java.util.Set; import com.stevesoft.pat.Regex; +import jalview.datamodel.DBRefEntry; +import jalview.datamodel.DBRefSource; +import jalview.datamodel.Mapping; +import jalview.datamodel.PDBEntry; +import jalview.datamodel.SequenceI; + /** * Utilities for handling DBRef objects and their collections. */ @@ -45,135 +45,96 @@ public class DBRefUtils */ private static Map canonicalSourceNameLookup = new HashMap<>(); - - static - { - // TODO load these from a resource file? - canonicalSourceNameLookup.put("uniprotkb/swiss-prot", - DBRefSource.UNIPROT); - canonicalSourceNameLookup.put("uniprotkb/trembl", DBRefSource.UNIPROT); - - // Ensembl values for dbname in xref REST service: - canonicalSourceNameLookup.put("uniprot/sptrembl", DBRefSource.UNIPROT); - canonicalSourceNameLookup.put("uniprot/swissprot", DBRefSource.UNIPROT); - - canonicalSourceNameLookup.put("pdb", DBRefSource.PDB); - canonicalSourceNameLookup.put("ensembl", DBRefSource.ENSEMBL); - // Ensembl Gn and Tr are for Ensembl genomic and transcript IDs as served - // from ENA. - canonicalSourceNameLookup.put("ensembl-tr", DBRefSource.ENSEMBL); - canonicalSourceNameLookup.put("ensembl-gn", DBRefSource.ENSEMBL); - - // Make sure we have lowercase entries for all canonical string lookups - Set keys = canonicalSourceNameLookup.keySet(); - for (String k : keys) - { - canonicalSourceNameLookup.put(k.toLowerCase(), - canonicalSourceNameLookup.get(k)); - } - - } + public final static int DB_SOURCE = 1; + public final static int DB_VERSION = 2; + public final static int DB_ID = 4; + public final static int DB_MAP = 8; + + public final static int SEARCH_MODE_NO_MAP_NO_VERSION = DB_SOURCE | DB_ID; + public final static int SEARCH_MODE_FULL = DB_SOURCE | DB_VERSION | DB_ID | DB_MAP; + + static + { + // TODO load these from a resource file? + canonicalSourceNameLookup.put("uniprotkb/swiss-prot", DBRefSource.UNIPROT); + canonicalSourceNameLookup.put("uniprotkb/trembl", DBRefSource.UNIPROT); + + // Ensembl values for dbname in xref REST service: + canonicalSourceNameLookup.put("uniprot/sptrembl", DBRefSource.UNIPROT); + canonicalSourceNameLookup.put("uniprot/swissprot", DBRefSource.UNIPROT); + + canonicalSourceNameLookup.put("pdb", DBRefSource.PDB); + canonicalSourceNameLookup.put("ensembl", DBRefSource.ENSEMBL); + // Ensembl Gn and Tr are for Ensembl genomic and transcript IDs as served + // from ENA. + canonicalSourceNameLookup.put("ensembl-tr", DBRefSource.ENSEMBL); + canonicalSourceNameLookup.put("ensembl-gn", DBRefSource.ENSEMBL); + + // guarantee we always have lowercase entries for canonical string lookups + for (String k : canonicalSourceNameLookup.keySet()) + { + canonicalSourceNameLookup.put(k.toLowerCase(), + canonicalSourceNameLookup.get(k)); + } + } /** * Returns those DBRefEntry objects whose source identifier (once converted to * Jalview's canonical form) is in the list of sources to search for. Returns * null if no matches found. * - * @param dbrefs - * DBRefEntry objects to search - * @param sources - * array of sources to select + * @param dbrefs DBRefEntry objects to search + * @param sources array of sources to select * @return */ - public static DBRefEntry[] selectRefs(DBRefEntry[] dbrefs, - String[] sources) + public static List selectRefs(List dbrefs, String[] sources) { - if (dbrefs == null || sources == null) - { - return dbrefs; - } - HashSet srcs = new HashSet<>(); - for (String src : sources) - { - srcs.add(src.toUpperCase()); - } - - List res = new ArrayList<>(); - for (DBRefEntry dbr : dbrefs) - { - String source = getCanonicalName(dbr.getSource()); - if (srcs.contains(source.toUpperCase())) - { - res.add(dbr); - } - } - - if (res.size() > 0) - { - DBRefEntry[] reply = new DBRefEntry[res.size()]; - return res.toArray(reply); - } - return null; - } - - /** - * look up source in an internal list of database reference sources and return - * the canonical jalview name for the source, or the original string if it has - * no canonical form. - * - * @param source - * @return canonical jalview source (one of jalview.datamodel.DBRefSource.*) - * or original source - */ - public static String getCanonicalName(String source) - { - if (source == null) - { - return null; - } - String canonical = canonicalSourceNameLookup.get(source.toLowerCase()); - return canonical == null ? source : canonical; + if (dbrefs == null || sources == null) + { + return dbrefs; + } + + // BH TODO (what?) + HashSet srcs = new HashSet(); + for (String src : sources) + { + srcs.add(src.toUpperCase()); + } + + int nrefs = dbrefs.size(); + List res = new ArrayList(); + for (int ib = 0; ib < nrefs; ib++) + { + DBRefEntry dbr = dbrefs.get(ib); + String source = getCanonicalName(dbr.getSource()); + if (srcs.contains(source.toUpperCase())) + { + res.add(dbr); + } + } + if (res.size() > 0) + { + // List reply = new DBRefEntry[res.size()]; + return res;// .toArray(reply); + } + return null; } - /** - * Returns a (possibly empty) list of those references that match the given - * entry. Currently uses a comparator which matches if - *
    - *
  • database sources are the same
  • - *
  • accession ids are the same
  • - *
  • both have no mapping, or the mappings are the same
  • - *
- * - * @param ref - * Set of references to search - * @param entry - * pattern to match - * @return - */ - public static List searchRefs(DBRefEntry[] ref, - DBRefEntry entry) - { - return searchRefs(ref, entry, - matchDbAndIdAndEitherMapOrEquivalentMapList); - } - - /** - * Returns a list of those references that match the given accession id - *
    - *
  • database sources are the same
  • - *
  • accession ids are the same
  • - *
  • both have no mapping, or the mappings are the same
  • - *
- * - * @param refs - * Set of references to search - * @param accId - * accession id to match - * @return - */ - public static List searchRefs(DBRefEntry[] refs, String accId) + private static boolean selectRefsBS(List dbrefs, int sourceKeys, BitSet bsSelect) { - return searchRefs(refs, new DBRefEntry("", "", accId), matchId); + if (dbrefs == null || sourceKeys == 0) + { + return false; + } + for (int i = 0, n = dbrefs.size(); i < n; i++) + { + DBRefEntry dbr = dbrefs.get(i); + if ((dbr.getSourceKey() & sourceKeys) != 0) + { + bsSelect.clear(i); + } + } + return !bsSelect.isEmpty(); } /** @@ -205,229 +166,146 @@ public class DBRefUtils return rfs; } - interface DbRefComp - { - public boolean matches(DBRefEntry refa, DBRefEntry refb); - } - - /** - * match on all non-null fields in refa - */ - // TODO unused - remove? - public static DbRefComp matchNonNullonA = new DbRefComp() - { - @Override - public boolean matches(DBRefEntry refa, DBRefEntry refb) - { - if (refa.getSource() == null - || DBRefUtils.getCanonicalName(refb.getSource()).equals( - DBRefUtils.getCanonicalName(refa.getSource()))) - { - if (refa.getVersion() == null - || refb.getVersion().equals(refa.getVersion())) - { - if (refa.getAccessionId() == null - || refb.getAccessionId().equals(refa.getAccessionId())) - { - if (refa.getMap() == null || (refb.getMap() != null - && refb.getMap().equals(refa.getMap()))) - { - return true; - } - } - } - } - return false; - } - }; - - /** - * either field is null or field matches for all of source, version, accession - * id and map. - */ - // TODO unused - remove? - public static DbRefComp matchEitherNonNull = new DbRefComp() - { - @Override - public boolean matches(DBRefEntry refa, DBRefEntry refb) - { - if (nullOrEqualSource(refa.getSource(), refb.getSource()) - && nullOrEqual(refa.getVersion(), refb.getVersion()) - && nullOrEqual(refa.getAccessionId(), refb.getAccessionId()) - && nullOrEqual(refa.getMap(), refb.getMap())) - { - return true; - } - return false; - } - }; - - /** - * accession ID and DB must be identical. Version is ignored. Map is either - * not defined or is a match (or is compatible?) - */ - // TODO unused - remove? - public static DbRefComp matchDbAndIdAndEitherMap = new DbRefComp() - { - @Override - public boolean matches(DBRefEntry refa, DBRefEntry refb) - { - if (refa.getSource() != null && refb.getSource() != null - && DBRefUtils.getCanonicalName(refb.getSource()).equals( - DBRefUtils.getCanonicalName(refa.getSource()))) - { - // We dont care about version - if (refa.getAccessionId() != null && refb.getAccessionId() != null - // FIXME should be && not || here? - || refb.getAccessionId().equals(refa.getAccessionId())) - { - if ((refa.getMap() == null || refb.getMap() == null) - || (refa.getMap() != null && refb.getMap() != null - && refb.getMap().equals(refa.getMap()))) - { - return true; - } - } - } - return false; - } - }; - - /** - * 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 the complement of - * maplist of map on b. - */ - // TODO unused - remove? - public static DbRefComp matchDbAndIdAndComplementaryMapList = new DbRefComp() - { - @Override - public boolean matches(DBRefEntry refa, DBRefEntry refb) - { - if (refa.getSource() != null && refb.getSource() != null - && DBRefUtils.getCanonicalName(refb.getSource()).equals( - DBRefUtils.getCanonicalName(refa.getSource()))) - { - // We dont care about version - if (refa.getAccessionId() != null && refb.getAccessionId() != null - || refb.getAccessionId().equals(refa.getAccessionId())) - { - if ((refa.getMap() == null && refb.getMap() == null) - || (refa.getMap() != null && refb.getMap() != null)) - { - if ((refb.getMap().getMap() == null - && refa.getMap().getMap() == null) - || (refb.getMap().getMap() != null - && refa.getMap().getMap() != null - && refb.getMap().getMap().getInverse() - .equals(refa.getMap().getMap()))) - { - return true; - } - } - } - } - return false; - } - }; - - /** - * accession ID and DB must be identical. Version is ignored. No map on both - * or or map but no maplist on either or maplist of map on a is equivalent to - * the maplist of map on b. - */ - // TODO unused - remove? - public static DbRefComp matchDbAndIdAndEquivalentMapList = new DbRefComp() - { - @Override - public boolean matches(DBRefEntry refa, DBRefEntry refb) - { - if (refa.getSource() != null && refb.getSource() != null - && DBRefUtils.getCanonicalName(refb.getSource()).equals( - DBRefUtils.getCanonicalName(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())))) - { - return true; - } - } - } - return false; - } - }; - - /** - * accession ID and DB must be identical, or null on a. 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() - { - @Override - public boolean matches(DBRefEntry refa, DBRefEntry refb) - { - if (refa.getSource() != null && refb.getSource() != null - && DBRefUtils.getCanonicalName(refb.getSource()).equals( - DBRefUtils.getCanonicalName(refa.getSource()))) - { - // We dont care about version - - if (refa.getAccessionId() == null - || refa.getAccessionId().equals(refb.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())))) - { - return true; - } - } - } - return false; - } - }; - - /** - * accession ID only must be identical. - */ - public static DbRefComp matchId = new DbRefComp() - { - @Override - public boolean matches(DBRefEntry refa, DBRefEntry refb) - { - if (refa.getAccessionId() != null && refb.getAccessionId() != null - && refb.getAccessionId().equals(refa.getAccessionId())) - { - return true; - } - return false; - } - }; + /** + * look up source in an internal list of database reference sources and return + * the canonical jalview name for the source, or the original string if it has + * no canonical form. + * + * @param source + * @return canonical jalview source (one of jalview.datamodel.DBRefSource.*) or + * original source + */ + public static String getCanonicalName(String source) + { + if (source == null) + { + return null; + } + String canonical = canonicalSourceNameLookup.get(source.toLowerCase()); + return canonical == null ? source : canonical; + } + + /** + * Returns a (possibly empty) list of those references that match the given + * entry. Currently uses a comparator which matches if + *
    + *
  • database sources are the same
  • + *
  • accession ids are the same
  • + *
  • both have no mapping, or the mappings are the same
  • + *
+ * + * @param ref Set of references to search + * @param entry pattern to match + * @param mode SEARCH_MODE_FULL for all; SEARCH_MODE_NO_MAP_NO_VERSION optional + * @return + */ + public static List searchRefs(List ref, DBRefEntry entry, int mode) { + return searchRefs(ref, entry, matchDbAndIdAndEitherMapOrEquivalentMapList, mode); + } + + /** + * Returns a list of those references that match the given accession id + *
    + *
  • database sources are the same
  • + *
  • accession ids are the same
  • + *
  • both have no mapping, or the mappings are the same
  • + *
+ * + * @param refs Set of references to search + * @param accId accession id to match + * @return + */ + public static List searchRefs(List refs, String accId) { + List rfs = new ArrayList(); + if (refs == null || accId == null) { + return rfs; + } + for (int i = 0, n = refs.size(); i < n; i++) { + DBRefEntry e = refs.get(i); + if (accId.equals(e.getAccessionId())) { + rfs.add(e); + } + } + return rfs; +// return searchRefs(refs, new DBRefEntry("", "", accId), matchId, SEARCH_MODE_FULL); + } + + /** + * Returns a (possibly empty) list of those references that match the given + * entry, according to the given comparator. + * + * @param refs an array of database references to search + * @param entry an entry to compare against + * @param comparator + * @param mode SEARCH_MODE_FULL for all; SEARCH_MODE_NO_MAP_NO_VERSION + * optional + * @return + */ + static List searchRefs(List refs, DBRefEntry entry, DbRefComp comparator, int mode) { + List rfs = new ArrayList(); + if (refs == null || entry == null) { + return rfs; + } + for (int i = 0, n = refs.size(); i < n; i++) { + DBRefEntry e = refs.get(i); + if (comparator.matches(entry, e, SEARCH_MODE_FULL)) { + rfs.add(e); + } + } + return rfs; + } + + interface DbRefComp { + default public boolean matches(DBRefEntry refa, DBRefEntry refb) { + return matches(refa, refb, SEARCH_MODE_FULL); + }; + + public boolean matches(DBRefEntry refa, DBRefEntry refb, int mode); + } + + /** + * match on all non-null fields in refa + */ + // TODO unused - remove? would be broken by equating "" with null + public static DbRefComp matchNonNullonA = new DbRefComp() { + @Override + public boolean matches(DBRefEntry refa, DBRefEntry refb, int mode) { + if ((mode & DB_SOURCE) != 0 && + (refa.getSource() == null || DBRefUtils.getCanonicalName(refb.getSource()) + .equals(DBRefUtils.getCanonicalName(refa.getSource())))) { + if ((mode & DB_VERSION) != 0 && + (refa.getVersion() == null || refb.getVersion().equals(refa.getVersion()))) { + if ((mode & DB_ID) != 0 && + (refa.getAccessionId() == null || refb.getAccessionId().equals(refa.getAccessionId()))) { + if ((mode & DB_MAP) != 0 && + (refa.getMap() == null || (refb.getMap() != null && refb.getMap().equals(refa.getMap())))) { + return true; + } + } + } + } + return false; + } + }; + + /** + * either field is null or field matches for all of source, version, accession + * id and map. + */ + // TODO unused - remove? + public static DbRefComp matchEitherNonNull = new DbRefComp() { + @Override + public boolean matches(DBRefEntry refa, DBRefEntry refb, int mode) { + if (nullOrEqualSource(refa.getSource(), refb.getSource()) + && nullOrEqual(refa.getVersion(), refb.getVersion()) + && nullOrEqual(refa.getAccessionId(), refb.getAccessionId()) + && nullOrEqual(refa.getMap(), refb.getMap())) { + return true; + } + return false; + } + + }; /** * Parses a DBRefEntry and adds it to the sequence, also a PDBEntry if the @@ -498,64 +376,124 @@ public class DBRefUtils 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 DBRefEntry[] selectDbRefs(boolean selectDna, - 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) - } - - /** + /** + * accession ID and DB must be identical. Version is ignored. Map is either not + * defined or is a match (or is compatible?) + */ + // TODO unused - remove? + public static DbRefComp matchDbAndIdAndEitherMap = new DbRefComp() { + @Override + public boolean matches(DBRefEntry refa, DBRefEntry refb, int mode) { + if (refa.getSource() != null && refb.getSource() != null && DBRefUtils.getCanonicalName(refb.getSource()) + .equals(DBRefUtils.getCanonicalName(refa.getSource()))) { + // We dont care about version + if (refa.getAccessionId() != null && refb.getAccessionId() != null + // FIXME should be && not || here? + || refb.getAccessionId().equals(refa.getAccessionId())) { + if ((refa.getMap() == null || refb.getMap() == null) || (refa.getMap() != null + && refb.getMap() != null && refb.getMap().equals(refa.getMap()))) { + return true; + } + } + } + return false; + } + }; + + /** + * 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 the complement of + * maplist of map on b. + */ + // TODO unused - remove? + public static DbRefComp matchDbAndIdAndComplementaryMapList = new DbRefComp() { + @Override + public boolean matches(DBRefEntry refa, DBRefEntry refb, int mode) { + if (refa.getSource() != null && refb.getSource() != null && DBRefUtils.getCanonicalName(refb.getSource()) + .equals(DBRefUtils.getCanonicalName(refa.getSource()))) { + // We dont care about version + if (refa.getAccessionId() != null && refb.getAccessionId() != null + || refb.getAccessionId().equals(refa.getAccessionId())) { + if ((refa.getMap() == null && refb.getMap() == null) + || (refa.getMap() != null && refb.getMap() != null)) { + if ((refb.getMap().getMap() == null && refa.getMap().getMap() == null) + || (refb.getMap().getMap() != null && refa.getMap().getMap() != null + && refb.getMap().getMap().getInverse().equals(refa.getMap().getMap()))) { + return true; + } + } + } + } + return false; + } + }; + + /** + * accession ID and DB must be identical. Version is ignored. No map on both or + * or map but no maplist on either or maplist of map on a is equivalent to the + * maplist of map on b. + */ + // TODO unused - remove? + public static DbRefComp matchDbAndIdAndEquivalentMapList = new DbRefComp() { + @Override + public boolean matches(DBRefEntry refa, DBRefEntry refb, int mode) { + if (refa.getSource() != null && refb.getSource() != null && DBRefUtils.getCanonicalName(refb.getSource()) + .equals(DBRefUtils.getCanonicalName(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())))) { + return true; + } + } + } + return false; + } + }; + + /** + * accession ID and DB must be identical, or null on a. 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() + { + @Override + public boolean matches(DBRefEntry refa, DBRefEntry refb, int mode) + { + if (refa.getSource() != null && refb.getSource() != null && DBRefUtils.getCanonicalName(refb.getSource()) + .equals(DBRefUtils.getCanonicalName(refa.getSource()))) + { + // We dont care about version + if (refa.getAccessionId() == null || refa.getAccessionId().equals(refb.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())))) + { + return true; + } + } + } + return false; + } + }; + + /** * Returns the (possibly empty) list of those supplied dbrefs which have the * specified source database, with a case-insensitive match of source name * @@ -580,126 +518,191 @@ public class DBRefUtils return matches; } - /** - * promote direct database references to primary for nucleotide or protein - * sequences if they have an appropriate primary ref - * - * - * - * - * - * - * - * - * - * - * - * - * - * - * - * - * - * - * - * - * - *
Seq TypePrimary DBDirect which will be promoted
peptidesEnsemblUniprot
peptidesEnsemblUniprot
dnaEnsemblENA
- * - * @param sequence - */ - public static void ensurePrimaries(SequenceI sequence) - { - List pr = sequence.getPrimaryDBRefs(); - if (pr.size() == 0) - { - // nothing to do - return; - } - List selfs = new ArrayList<>(); - { - DBRefEntry[] selfArray = selectDbRefs(!sequence.isProtein(), - sequence.getDBRefs()); - if (selfArray == null || selfArray.length == 0) - { - // nothing to do - return; - } - selfs.addAll(Arrays.asList(selfArray)); - } - - // filter non-primary refs - for (DBRefEntry p : pr) - { - while (selfs.contains(p)) - { - selfs.remove(p); - } - } - List toPromote = new ArrayList<>(); - - for (DBRefEntry p : pr) - { - List promType = new ArrayList<>(); - if (sequence.isProtein()) - { - switch (getCanonicalName(p.getSource())) - { - case DBRefSource.UNIPROT: - // case DBRefSource.UNIPROTKB: - // case DBRefSource.UP_NAME: - // search for and promote ensembl - promType.add(DBRefSource.ENSEMBL); - break; - case DBRefSource.ENSEMBL: - // search for and promote Uniprot - promType.add(DBRefSource.UNIPROT); - break; - } - } - else - { - // TODO: promote transcript refs - } - - // collate candidates and promote them - DBRefEntry[] candidates = selectRefs(selfs.toArray(new DBRefEntry[0]), - promType.toArray(new String[0])); - if (candidates != null) - { - for (DBRefEntry cand : candidates) - { - if (cand.hasMap()) - { - if (cand.getMap().getTo() != null - && cand.getMap().getTo() != sequence) - { - // can't promote refs with mappings to other sequences - continue; - } - if (cand.getMap().getMap().getFromLowest() != sequence - .getStart() - && cand.getMap().getMap().getFromHighest() != sequence - .getEnd()) - { - // can't promote refs with mappings from a region of this sequence - // - eg CDS - continue; - } - } - // and promote - cand.setVersion(p.getVersion() + " (promoted)"); - selfs.remove(cand); - toPromote.add(cand); - if (!cand.isPrimaryCandidate()) - { - System.out.println( - "Warning: Couldn't promote dbref " + cand.toString() - + " for sequence " + sequence.toString()); - } - } - } - } - } + /** + * 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 TypePrimary DBDirect which will be promoted
peptidesEnsemblUniprot
peptidesEnsemblUniprot
dnaEnsemblENA
+ * + * @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)) + 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 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)) + 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(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()); + } + } + } + } + } }