package jalview.io.gff;
+import java.io.BufferedInputStream;
+import java.io.BufferedReader;
+import java.io.IOException;
+import java.io.InputStream;
+import java.io.InputStreamReader;
+import java.text.ParseException;
+import java.util.ArrayList;
+import java.util.HashMap;
+import java.util.List;
+import java.util.Map;
+import java.util.NoSuchElementException;
+import java.util.zip.ZipEntry;
+import java.util.zip.ZipInputStream;
+
+import org.biojava.nbio.ontology.Ontology;
+import org.biojava.nbio.ontology.Term;
+import org.biojava.nbio.ontology.Term.Impl;
+import org.biojava.nbio.ontology.Triple;
+import org.biojava.nbio.ontology.io.OboParser;
+import org.biojava.nbio.ontology.utils.Annotation;
+
/**
* A wrapper class that parses the Sequence Ontology and exposes useful access
- * methods
+ * methods. This version uses the BioJava parser.
*/
public class SequenceOntology
{
- private static SequenceOntology instance = new SequenceOntology();
- public static SequenceOntology getInstance()
+ /*
+ * selected commonly used values for quick reference
+ */
+ // SO:0000316
+ public static final String CDS = "CDS";
+
+ // SO:0001060
+ public static final String SEQUENCE_VARIANT = "sequence_variant";
+
+ // SO:0000147
+ public static final String EXON = "exon";
+
+ // SO:0000673
+ public static final String TRANSCRIPT = "transcript";
+
+ /*
+ * singleton instance of this class
+ */
+ private static SequenceOntology instance;
+
+ /*
+ * the parsed Ontology data as modelled by BioJava
+ */
+ private Ontology ontology;
+
+ /*
+ * the ontology term for the isA relationship
+ */
+ private Term isA;
+
+ /*
+ * lookup of terms by user readable name (NB not guaranteed unique)
+ */
+ private Map<String, Term> termsByDescription;
+
+ /*
+ * Map where key is a Term and value is a (possibly empty) list of
+ * all Terms to which the key has an 'isA' relationship, either
+ * directly or indirectly (A isA B isA C)
+ */
+ private Map<Term, List<Term>> termIsA;
+
+ /**
+ * Returns singleton instance
+ *
+ * @return
+ */
+ public synchronized static SequenceOntology getInstance()
{
+ if (instance == null)
+ {
+ instance = new SequenceOntology();
+ }
return instance;
}
/**
- * Private constructor to enforce use of singleton.
+ * Private constructor to enforce use of singleton. Parses and caches the SO
+ * OBO data file.
*/
private SequenceOntology()
{
- // TODO: parse and cache so.obo data file e.g. using BioJava
+ termsByDescription = new HashMap<String, Term>();
+ termIsA = new HashMap<Term, List<Term>>();
+
+ loadOntologyZipFile("so-xp-simple.obo");
}
/**
- * Test whether the given Sequence Ontology term is nucleotide_match (either
- * directly or via is_a relationship)
+ * Loads the given ontology file from a zip file with ".zip" appended
*
- * @param soTerm
+ * @param ontologyFile
+ */
+ protected void loadOntologyZipFile(String ontologyFile)
+ {
+ ZipInputStream zipStream = null;
+ try
+ {
+ String zipFile = ontologyFile + ".zip";
+ System.out.println("Loading Sequence Ontology from " + zipFile);
+ InputStream inStream = this.getClass().getResourceAsStream(
+ "/" + zipFile);
+ zipStream = new ZipInputStream(new BufferedInputStream(inStream));
+ ZipEntry entry;
+ while ((entry = zipStream.getNextEntry()) != null)
+ {
+ if (entry.getName().equals(ontologyFile))
+ {
+ loadOboFile(zipStream);
+ }
+ }
+ } catch (Exception e)
+ {
+ e.printStackTrace();
+ } finally
+ {
+ closeStream(zipStream);
+ }
+ }
+
+ /**
+ * Closes the input stream, swallowing all exceptions
+ *
+ * @param is
+ */
+ protected void closeStream(InputStream is)
+ {
+ if (is != null)
+ {
+ try
+ {
+ is.close();
+ } catch (IOException e)
+ {
+ // ignore
+ }
+ }
+ }
+
+ /**
+ * Reads, parses and stores the OBO file data
+ *
+ * @param is
+ * @throws ParseException
+ * @throws IOException
+ */
+ protected void loadOboFile(InputStream is) throws ParseException,
+ IOException
+ {
+ BufferedReader oboFile = new BufferedReader(new InputStreamReader(is));
+ OboParser parser = new OboParser();
+ ontology = parser.parseOBO(oboFile, "SO", "the SO ontology");
+ isA = ontology.getTerm("is_a");
+ storeTermNames();
+ }
+
+ /**
+ * Stores a lookup table of terms by description. Note that description is not
+ * guaranteed unique. Where duplicate descriptions are found, try to discard
+ * the term that is flagged as obsolete. However we do store obsolete terms
+ * where there is no duplication of description.
+ */
+ protected void storeTermNames()
+ {
+ for (Term term : ontology.getTerms())
+ {
+ if (term instanceof Impl)
+ {
+ String description = term.getDescription();
+ if (description != null)
+ {
+ Term replaced = termsByDescription.get(description);
+ if (replaced != null)
+ {
+ boolean newTermIsObsolete = isObsolete(term);
+ boolean oldTermIsObsolete = isObsolete(replaced);
+ if (newTermIsObsolete && !oldTermIsObsolete)
+ {
+ System.err.println("Ignoring " + term.getName()
+ + " as obsolete and duplicated by "
+ + replaced.getName());
+ term = replaced;
+ }
+ else if (!newTermIsObsolete && oldTermIsObsolete)
+ {
+ System.err.println("Ignoring " + replaced.getName()
+ + " as obsolete and duplicated by " + term.getName());
+ }
+ else
+ {
+ System.err.println("Warning: " + term.getName()
+ + " has replaced " + replaced.getName()
+ + " for lookup of '" + description + "'");
+ }
+ }
+ termsByDescription.put(description, term);
+ }
+ }
+ }
+ }
+
+ /**
+ * Answers true if the term has property "is_obsolete" with value true, else
+ * false
+ *
+ * @param term
* @return
*/
- public boolean isNucleotideMatch(String soTerm)
+ public static boolean isObsolete(Term term)
{
- // temporary until OBO parser is in place!
- // (which should also match SO ids e.g. "SO:0000347")
- String[] nucMatch = { "nucleotide_match", "primer_match",
- "cross_genome_match", "expressed_sequence_match",
- "translated_nucleotide_match", "UST_match", "RSF_match",
- "cDNA_match", "EST_match" };
- for (int i = 0; i < nucMatch.length; i++)
+ Annotation ann = term.getAnnotation();
+ if (ann != null)
{
- if (nucMatch[i].equals(soTerm))
+ try
{
- return true;
+ if (Boolean.TRUE.equals(ann.getProperty("is_obsolete")))
+ {
+ return true;
+ }
+ } catch (NoSuchElementException e)
+ {
+ // fall through to false
}
}
return false;
}
/**
+ * Test whether the given Sequence Ontology term is nucleotide_match (either
+ * directly or via is_a relationship)
+ *
+ * @param soTerm
+ * SO name or description
+ * @return
+ */
+ public boolean isNucleotideMatch(String soTerm)
+ {
+ return isA(soTerm, "nucleotide_match");
+ }
+
+ /**
* Test whether the given Sequence Ontology term is protein_match (either
* directly or via is_a relationship)
*
* @param soTerm
+ * SO name or description
* @return
*/
public boolean isProteinMatch(String soTerm)
{
- // temporary until OBO parser is in place!
- return "protein_match".equals(soTerm)
- || "protein_hmm_match".equals(soTerm);
+ return isA(soTerm, "protein_match");
}
+ /**
+ * Test whether the given Sequence Ontology term is polypeptide (either
+ * directly or via is_a relationship)
+ *
+ * @param soTerm
+ * SO name or description
+ * @return
+ */
public boolean isPolypeptide(String soTerm)
{
- return "polypeptide".equals(soTerm);
+ return isA(soTerm, "polypeptide");
+ }
+
+ /**
+ * Returns true if the given term has a (direct or indirect) 'isA'
+ * relationship with the parent
+ *
+ * @param child
+ * @param parent
+ * @return
+ */
+ public boolean isA(String child, String parent)
+ {
+ /*
+ * optimise trivial checks like isA("CDS", "CDS")
+ */
+ if (child.equals(parent))
+ {
+ return true;
+ }
+
+ Term childTerm = getTerm(child);
+ Term parentTerm = getTerm(parent);
+
+ return termIsA(childTerm, parentTerm);
+ }
+
+ /**
+ * Returns true if the childTerm 'isA' parentTerm (directly or indirectly).
+ *
+ * @param childTerm
+ * @param parentTerm
+ * @return
+ */
+ protected synchronized boolean termIsA(Term childTerm, Term parentTerm)
+ {
+ /*
+ * null term could arise from a misspelled SO description
+ */
+ if (childTerm == null || parentTerm == null)
+ {
+ return false;
+ }
+
+ /*
+ * recursive search endpoint:
+ */
+ if (childTerm == parentTerm)
+ {
+ return true;
+ }
+
+ /*
+ * lazy initialisation - find all of a term's parents (recursively)
+ * the first time this is called, and save them in a map.
+ */
+ if (!termIsA.containsKey(childTerm))
+ {
+ findParents(childTerm);
+ }
+
+ List<Term> parents = termIsA.get(childTerm);
+ for (Term parent : parents)
+ {
+ if (termIsA(parent, parentTerm))
+ {
+ /*
+ * add (great-)grandparents to parents list as they are discovered,
+ * for faster lookup next time
+ */
+ if (!parents.contains(parentTerm))
+ {
+ parents.add(parentTerm);
+ }
+ return true;
+ }
+ }
+
+ return false;
+ }
+
+ /**
+ * Finds all the 'isA' parents of the childTerm and stores them as a (possibly
+ * empty) list.
+ *
+ * @param childTerm
+ */
+ protected synchronized void findParents(Term childTerm)
+ {
+ List<Term> result = new ArrayList<Term>();
+ for (Triple triple : ontology.getTriples(childTerm, null, isA))
+ {
+ Term parent = triple.getObject();
+ result.add(parent);
+
+ /*
+ * and search for the parent's parents recursively
+ */
+ findParents(parent);
+ }
+ termIsA.put(childTerm, result);
+ }
+
+ /**
+ * Returns the Term for a given name (e.g. "SO:0000735") or description (e.g.
+ * "sequence_location"), or null if not found.
+ *
+ * @param child
+ * @return
+ */
+ protected Term getTerm(String nameOrDescription)
+ {
+ Term t = termsByDescription.get(nameOrDescription);
+ if (t == null)
+ {
+ try
+ {
+ t = ontology.getTerm(nameOrDescription);
+ } catch (NoSuchElementException e)
+ {
+ // not found
+ }
+ }
+ return t;
+ }
+
+ public boolean isSequenceVariant(String soTerm)
+ {
+ return isA(soTerm, "sequence_variant");
}
}