Merge pull request #459 from monarch-initiative/release-2.0.5

Release 2.0.5

CHANGELOG.rst

@@ -6,6 +6,14 @@ Changelog
 latest
 ------
 
+-----
+2.0.5
+-----
+
+- extend `OntologyGraph` with convenience methods
+- audit calculation of Resnik similarity scores
+- use Jackson annotations to describe serialization of `GeneIdentifier`
+
 -----
 2.0.4
 -----

README.md

@@ -14,7 +14,7 @@ associate phenotype annotation files.
 
 - **Language/Platform:** Java >=11
 - **License:** BSD 3-Clause Clear
-- **Version:** 2.0.4
+- **Version:** 2.0.5
 - **Authors:**
     - Sebastian Bauer
     - Peter N. Robinson
@@ -39,7 +39,7 @@ We recommend indicating the phenol version in the `properties` section of the po
 ```
 <properties>
   (...)
-  <phenol.version>2.0.4</phenol.version>
+  <phenol.version>2.0.5</phenol.version>
 </properties>
 ```
 

docs/conf.py

@@ -12,7 +12,7 @@
 # The short X.Y version.
 version = '2.0'
 # The full version, including alpha/beta/rc tags.
-release = '2.0.4'
+release = '2.0.5'
 
 # -- General configuration ---------------------------------------------------
 # https://www.sphinx-doc.org/en/master/usage/configuration.html#general-configuration

phenol-analysis/pom.xml

@@ -4,7 +4,7 @@
   <parent>
     <groupId>org.monarchinitiative.phenol</groupId>
     <artifactId>phenol</artifactId>
-    <version>2.0.4</version>
+    <version>2.0.5</version>
   </parent>
 
   <artifactId>phenol-analysis</artifactId>

phenol-annotations/pom.xml

@@ -4,7 +4,7 @@
   <parent>
     <groupId>org.monarchinitiative.phenol</groupId>
     <artifactId>phenol</artifactId>
-    <version>2.0.4</version>
+    <version>2.0.5</version>
   </parent>
 
   <artifactId>phenol-annotations</artifactId>
@@ -18,12 +18,34 @@
       <artifactId>phenol-core</artifactId>
       <version>${project.parent.version}</version>
     </dependency>
+    <dependency>
+      <groupId>com.fasterxml.jackson.core</groupId>
+      <artifactId>jackson-annotations</artifactId>
+      <scope>provided</scope>
+    </dependency>
     <dependency>
       <groupId>org.monarchinitiative.phenol</groupId>
       <artifactId>phenol-io</artifactId>
       <version>${project.parent.version}</version>
       <scope>test</scope>
     </dependency>
+    <dependency>
+      <groupId>com.fasterxml.jackson.core</groupId>
+      <artifactId>jackson-databind</artifactId>
+      <scope>test</scope>
+    </dependency>
   </dependencies>
 
+ <build>
+    <plugins>
+      <plugin>
+        <groupId>org.apache.maven.plugins</groupId>
+        <artifactId>maven-surefire-plugin</artifactId>
+        <configuration>
+          <argLine>--add-reads org.monarchinitiative.phenol.annotations=com.fasterxml.jackson.databind</argLine>
+        </configuration>
+      </plugin>
+    </plugins>
+  </build>
+
 </project>

phenol-annotations/src/main/java/module-info.java

@@ -6,6 +6,7 @@
 
   requires java.xml;
   requires org.slf4j;
+  requires com.fasterxml.jackson.annotation;
 
     /*                                             Exports                                                              */
   exports org.monarchinitiative.phenol.annotations.analysis;

phenol-annotations/src/main/java/org/monarchinitiative/phenol/annotations/formats/GeneIdentifier.java

@@ -1,5 +1,6 @@
 package org.monarchinitiative.phenol.annotations.formats;
 
+import com.fasterxml.jackson.annotation.JsonGetter;
 import org.monarchinitiative.phenol.ontology.data.Identified;
 import org.monarchinitiative.phenol.ontology.data.TermId;
 
@@ -22,13 +23,15 @@ private GeneIdentifier(TermId tid, String geneSymbol) {
 
   /** A CURIE for the gene, such as NCBIGene:3949 */
   @Override
+  @JsonGetter
   public TermId id() {
     return id;
   }
 
   /**
    * @return the symbol for the gene, e.g., LDLR
    */
+  @JsonGetter
   public String symbol() {
     return symbol;
   }

phenol-annotations/src/test/java/org/monarchinitiative/phenol/annotations/formats/GeneIdentifierTest.java

@@ -0,0 +1,22 @@
+package org.monarchinitiative.phenol.annotations.formats;
+
+import com.fasterxml.jackson.databind.ObjectMapper;
+import org.junit.jupiter.api.Test;
+import org.monarchinitiative.phenol.ontology.data.TermId;
+
+import static org.hamcrest.MatcherAssert.assertThat;
+import static org.hamcrest.Matchers.equalTo;
+
+public class GeneIdentifierTest {
+
+  private final ObjectMapper objectMapper = new ObjectMapper();
+
+  @Test
+  public void serializeToJson() throws Exception {
+    GeneIdentifier gi = GeneIdentifier.of(TermId.of("NCBIGene:6834"), "SURF1");
+
+    String actual = objectMapper.writeValueAsString(gi);
+
+    assertThat(actual, equalTo("{\"id\":\"NCBIGene:6834\",\"symbol\":\"SURF1\"}"));
+  }
+}

phenol-cli/pom.xml

@@ -4,7 +4,7 @@
   <parent>
     <groupId>org.monarchinitiative.phenol</groupId>
     <artifactId>phenol</artifactId>
-    <version>2.0.4</version>
+    <version>2.0.5</version>
   </parent>
 
   <artifactId>phenol-cli</artifactId>

phenol-cli/src/main/java/org/monarchinitiative/phenol/cli/Main.java

@@ -9,7 +9,7 @@
 
 @CommandLine.Command(name = "phenol demo",
   mixinStandardHelpOptions = true,
-  version = "2.0.4",
+  version = "2.0.5",
   description = "phenol demo programs")
 public class Main implements Callable<Integer> {
   private static final Logger LOGGER = LoggerFactory.getLogger(Main.class);

phenol-cli/src/main/java/org/monarchinitiative/phenol/cli/cmd/PrecomputeResnikMapCommand.java

@@ -7,10 +7,10 @@
 import org.monarchinitiative.phenol.annotations.io.hpo.HpoDiseaseLoaderOptions;
 import org.monarchinitiative.phenol.annotations.io.hpo.HpoDiseaseLoaders;
 import org.monarchinitiative.phenol.cli.demo.MicaCalculator;
-import org.monarchinitiative.phenol.io.OntologyLoader;
-import org.monarchinitiative.phenol.ontology.data.Ontology;
+import org.monarchinitiative.phenol.io.MinimalOntologyLoader;
+import org.monarchinitiative.phenol.ontology.data.MinimalOntology;
 import org.monarchinitiative.phenol.ontology.data.TermId;
-import org.monarchinitiative.phenol.ontology.similarity.HpoResnikSimilarity;
+import org.monarchinitiative.phenol.ontology.similarity.HpoResnikSimilarityPrecompute;
 import org.monarchinitiative.phenol.ontology.similarity.TermPair;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
@@ -59,7 +59,7 @@ public class PrecomputeResnikMapCommand implements Callable<Integer> {
   @Override
   public Integer call() throws Exception {
     LOGGER.info("Loading HPO from {}", hpoPath.toAbsolutePath());
-    Ontology hpo = OntologyLoader.loadOntology(hpoPath.toFile());
+    MinimalOntology hpo = MinimalOntologyLoader.loadOntology(hpoPath.toFile());
 
     LOGGER.info("Loading HPO annotations from {}", hpoaPath.toAbsolutePath());
     HpoDiseaseLoader loader = HpoDiseaseLoaders.defaultLoader(hpo, HpoDiseaseLoaderOptions.defaultOptions());
@@ -81,14 +81,13 @@ public Integer call() throws Exception {
     return 0;
   }
 
-  private Map<TermId, Double> calculateTermToIc(Ontology hpo, HpoDiseases diseases) {
+  private Map<TermId, Double> calculateTermToIc(MinimalOntology hpo, HpoDiseases diseases) {
     MicaCalculator micaCalculator = new MicaCalculator(hpo, assumeAnnotated);
     return micaCalculator.calculateMica(diseases).termToIc();
   }
 
-  private static Map<TermPair, Double> assignMicaToTermPairs(Ontology hpo, Map<TermId, Double> termToIc) {
-    HpoResnikSimilarity similarity = new HpoResnikSimilarity(hpo, termToIc);
-    return similarity.getTermPairResnikSimilarityMap();
+  private static Map<TermPair, Double> assignMicaToTermPairs(MinimalOntology hpo, Map<TermId, Double> termToIc) {
+    return HpoResnikSimilarityPrecompute.precomputeSimilaritiesForTermPairs(hpo, termToIc);
   }
 
   private void writeTermPairMap(Map<TermPair, Double> termPairResnikSimilarityMap,

phenol-cli/src/main/java/org/monarchinitiative/phenol/cli/demo/MicaCalculator.java

@@ -4,33 +4,35 @@
 import org.monarchinitiative.phenol.annotations.formats.hpo.AnnotatedItem;
 import org.monarchinitiative.phenol.annotations.formats.hpo.AnnotatedItemContainer;
 import org.monarchinitiative.phenol.ontology.data.Identified;
-import org.monarchinitiative.phenol.ontology.data.Ontology;
+import org.monarchinitiative.phenol.ontology.data.MinimalOntology;
 import org.monarchinitiative.phenol.ontology.data.TermId;
-import org.monarchinitiative.phenol.ontology.data.TermIds;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
 import java.time.Duration;
 import java.time.Instant;
 import java.util.*;
-import java.util.stream.Collectors;
 
 /**
- * Class for calculating information content of the <em>most informative common ancestor</em> (MICA)
- * for {@link Ontology} concepts using a corpus of annotated.
+ * Calculate information content of the <em>most informative common ancestor</em> (IC<sub>MICA</sub>) for the concepts
+ * in given {@link MinimalOntology} using a corpus of annotated items.
+ * <p>
+ * For instance, calculate IC<sub>MICA</sub> of HPO terms that annotate Mendelian diseases.
+ * <p>
+ * The IC<sub>MICA</sub> values are in <em>nats</em>.
  */
 public class MicaCalculator {
 
   private static final Logger LOGGER = LoggerFactory.getLogger(MicaCalculator.class);
 
-  private final Ontology hpo;
+  private final MinimalOntology hpo;
+  /**
+   * {@code true} if the count of annotations that are not used to annotate at least one corpus item should be set to 1
+   * to prevent ic being {@link Double#POSITIVE_INFINITY}.
+   */
   private final boolean assumeAnnotated;
 
-  public MicaCalculator(Ontology hpo) {
-    this(hpo, false);
-  }
-
-  public MicaCalculator(Ontology hpo, boolean assumeAnnotated) {
+  public MicaCalculator(MinimalOntology hpo, boolean assumeAnnotated) {
     this.assumeAnnotated = assumeAnnotated;
     this.hpo = Objects.requireNonNull(hpo);
   }
@@ -40,24 +42,25 @@ public MicaData calculateMica(AnnotatedItemContainer<? extends AnnotatedItem> it
     Map<TermId, Integer> phenotypeIdToDiseaseIds = new HashMap<>();
     Map<TermId, Collection<TermId>> diseaseIdToTermIds = new HashMap<>();
 
+    Set<TermId> termsAndAncestorsBuilder = new HashSet<>();
     for (AnnotatedItem item: itemContainer) {
-      List<TermId> annotationTermIds = item.annotations().stream()
-        .map(Identified::id)
-        .collect(Collectors.toList());
-      // add term ancestors
-      Set<TermId> inclAncestorTermIds = TermIds.augmentWithAncestors(hpo, annotationTermIds, true);
+      for (Identified annotation : item.annotations()) {
+        hpo.graph().extendWithAncestors(annotation.id(), true, termsAndAncestorsBuilder);
+      }
 
-      for (TermId tid : inclAncestorTermIds) {
+      for (TermId tid : termsAndAncestorsBuilder) {
+        // We can't do this within the hot loop above because there is no guarantee of uniqueness of the ancestors.
         phenotypeIdToDiseaseIds.compute(tid, (key, val) -> val == null ? 0 : val + 1);
         diseaseIdToTermIds.computeIfAbsent(item.id(), key -> new HashSet<>()).add(tid); // Note that this MUST be a Set
       }
+      termsAndAncestorsBuilder.clear();
     }
 
     Map<TermId, Double> termToIc = new HashMap<>();
-    int totalPopulationHpoTerms = phenotypeIdToDiseaseIds.get(HpoSubOntologyRootTermIds.PHENOTYPIC_ABNORMALITY);
+    double totalPopulationHpoTerms = phenotypeIdToDiseaseIds.get(HpoSubOntologyRootTermIds.PHENOTYPIC_ABNORMALITY);
     for (Map.Entry<TermId, Integer> e : phenotypeIdToDiseaseIds.entrySet()) {
       int annotatedCount = assumeAnnotated ? Math.max(e.getValue(), 1) : e.getValue();
-      double ic = -1 * Math.log((double)annotatedCount/totalPopulationHpoTerms);
+      double ic = Math.log(totalPopulationHpoTerms / annotatedCount);
       termToIc.put(e.getKey(), ic);
     }
 

phenol-core/pom.xml

@@ -4,7 +4,7 @@
   <parent>
     <groupId>org.monarchinitiative.phenol</groupId>
     <artifactId>phenol</artifactId>
-    <version>2.0.4</version>
+    <version>2.0.5</version>
   </parent>
 
   <artifactId>phenol-core</artifactId>