Merge pull request #12 from monarch-initiative/develop

Develop

Author: pnrobinson

docs/download.md

@@ -0,0 +1,4 @@
+# download
+
+download hp.json and several otherfiles needed by hpotools applications to a directory 
+called ``data`` that will be created if it does not already exist.

docs/index.md

@@ -1,3 +1,7 @@
 # hpotools
 
-This is a library of small applications that the HPO project regularly uses for internal purposes. Currently, we are not supporting this library for external users, but anybody is welcome to use it at their own risk. :-0.
+hpotools is a collection of various HPO-based applications.
+
+- [download](download.md)
+- [mondo](mondo.md): This application uses the mondo.json file to characterize collections of GA4GH phenopackets for use in ClintLR simulations.
+- [word](word.md)

docs/word.md

@@ -0,0 +1,5 @@
+# word
+
+Production and output of an RTF file that contains a table with all the terms
+that emanate from a certain term in the HPO. The RTF file can be opened in Word and is useful for
+those who want to revise an entire section of the HPO conveniently.

pom.xml

@@ -4,7 +4,7 @@
     <modelVersion>4.0.0</modelVersion>
     <groupId>org.monarchinitiative.hpotools</groupId>
     <artifactId>hpotools</artifactId>
-    <version>0.0.7</version>
+    <version>0.0.9</version>
     <packaging>jar</packaging>
     <name>hpotools</name>
 
@@ -51,12 +51,12 @@
             <dependency>
                 <groupId>info.picocli</groupId>
                 <artifactId>picocli</artifactId>
-                <version>4.7.5</version>
+                <version>4.7.6</version>
             </dependency>
             <dependency>
                 <groupId>org.phenopackets</groupId>
                 <artifactId>phenopacket-schema</artifactId>
-                <version>2.0.0</version>
+                <version>2.0.2</version>
             </dependency>
             <dependency>
                 <groupId>org.phenopackets.phenopackettools</groupId>
@@ -71,12 +71,12 @@
             <dependency>
                 <groupId>org.monarchinitiative.biodownload</groupId>
                 <artifactId>biodownload</artifactId>
-                <version>1.0.1</version>
+                <version>1.1.0</version>
             </dependency>
             <dependency>
                 <groupId>org.apache.poi</groupId>
                 <artifactId>poi-ooxml</artifactId>
-                <version>5.2.2</version>
+                <version>5.3.0</version>
             </dependency>
 
             <dependency>
@@ -159,7 +159,7 @@
         <dependency>
             <groupId>org.junit.jupiter</groupId>
             <artifactId>junit-jupiter</artifactId>
-            <version>5.9.0</version> <!-- Use the latest version available -->
+            <version>5.11.4</version> <!-- Use the latest version available -->
             <scope>test</scope>
         </dependency>
     </dependencies>
@@ -176,7 +176,7 @@
             <plugin>
                 <groupId>org.apache.maven.plugins</groupId>
                 <artifactId>maven-compiler-plugin</artifactId>
-                <version>3.8.0</version>
+                <version>3.13.0</version>
                 <configuration>
                     <source>${java.version}</source>
                     <target>${java.version}</target>
@@ -186,7 +186,7 @@
             <plugin>
                 <groupId>org.apache.maven.plugins</groupId>
                 <artifactId>maven-jar-plugin</artifactId>
-                <version>3.2.0</version>
+                <version>3.4.2</version>
                 <configuration>
                     <archive>
                         <index>true</index>
@@ -202,12 +202,12 @@
             <plugin>
                 <groupId>org.apache.maven.plugins</groupId>
                 <artifactId>maven-install-plugin</artifactId>
-                <version>3.0.0-M1</version>
+                <version>3.1.3</version>
             </plugin>
             <plugin>
                 <groupId>org.apache.maven.plugins</groupId>
                 <artifactId>maven-deploy-plugin</artifactId>
-                <version>3.0.0-M1</version>
+                <version>3.1.3</version>
             </plugin>
             <!-- site lifecycle, see https://maven.apache.org/ref/current/maven-core/lifecycles.html#site_Lifecycle -->
             <plugin>
@@ -218,12 +218,12 @@
             <plugin>
                 <groupId>org.apache.maven.plugins</groupId>
                 <artifactId>maven-surefire-plugin</artifactId>
-                <version>3.0.0-M5</version> <!-- Use the latest version -->
+                <version>3.5.2</version> <!-- Use the latest version -->
             </plugin>
             <plugin>
                 <groupId>org.apache.maven.plugins</groupId>
                 <artifactId>maven-shade-plugin</artifactId>
-                <version>3.2.2</version>
+                <version>3.6.0</version>
                 <configuration>
                     <createDependencyReducedPom>false</createDependencyReducedPom>
                 </configuration>
@@ -251,7 +251,7 @@
             <plugin>
                 <groupId>org.codehaus.mojo</groupId>
                 <artifactId>versions-maven-plugin</artifactId>
-                <version>2.16.2</version>
+                <version>2.18.0</version>
                 <configuration>
                     <generateBackupPoms>false</generateBackupPoms>
                 </configuration>

src/main/java/org/monarchinitiative/hpotools/Main.java

@@ -23,6 +23,9 @@ public static void main(String[] args) {
                     .addSubcommand("mondo", new MondoCommand())
                     .addSubcommand("onset", new OnsetCommand())
                     .addSubcommand("simhpo", new SimHpoCommand())
+                    .addSubcommand("stats", new StatsCommand())
+                    .addSubcommand("translate", new DiseaseTranslateCommand())
+                    .addSubcommand("tsv", new Hpo2TsvCommand())
                     .addSubcommand("word", new WordCommand())
                    ;
             cline.setToggleBooleanFlags(false);

src/main/java/org/monarchinitiative/hpotools/analysis/EncodingCheck.java

@@ -6,6 +6,7 @@
 
 import java.io.*;
 import java.nio.charset.StandardCharsets;
+import java.util.Set;
 import java.util.regex.Matcher;
 import java.util.regex.Pattern;
 
@@ -16,6 +17,9 @@ public class EncodingCheck {
     private final Pattern HPO_TERM_PATTERN = Pattern.compile(pattern);
     private final File hpoOwlFile;
 
+    private final static  Set<Character> ACCETABLE_CHARS = Set.of('ö','ü', 'ï', 'à', 'é');
+
+    private int badChars = 0;
 
     /**
      * @param hpoOwlFile file for hp-edit.owl. Presumed checked by client code
@@ -27,9 +31,11 @@ public EncodingCheck(File hpoOwlFile) {
     public void checkEncoding() {
 
         int i = 0;
+        int encodingErrors = 0;
         try (BufferedReader br = new BufferedReader(new FileReader(hpoOwlFile))){
             String line;
             String hp_term = "";
+
             while ((line = br.readLine()) != null) {
                 Matcher m = HPO_TERM_PATTERN.matcher(line);
                 if (m.find()) {
@@ -40,21 +46,33 @@ public void checkEncoding() {
         } catch (IOException e) {
             LOGGER.error("Could not read hpo owl file", e);
         }
+        if (badChars > 0) {
+            System.out.printf("Bad characters in HPO owl file: %d%n", badChars);
+        } else {
+            System.out.println("No bad encodings found in HPO owl file");
+        }
     }
 
-    public static void checkLine(String line, int lineno, String previous) {
+    public void checkLine(String line, int lineno, String previous) {
         byte[] bytes = line.getBytes(StandardCharsets.ISO_8859_1);
         String decodedLine = new String(bytes);
         if (! line.equals(decodedLine)) {
             for (int i=0; i< line.length();i++) {
                 if (line.charAt(i) != decodedLine.charAt(i)) {
+                    Character unequalChar = line.charAt(i);
+                    if (ACCETABLE_CHARS.contains(unequalChar)) {
+                        continue;
+                    } else {
+                        badChars++;
+                    }
                     System.out.println(previous);
                     System.out.printf("L.%d:Pos:%d: ", lineno, i);
                     int b = Math.max(0, i-20);
                     int e = Math.min(line.length(), i+20);
                     String ss1 = line.substring(b, i);
                     String ss2 = line.substring(i+1, e);
-                    System.out.printf("%s{%c}%sc\n\n", ss1, line.charAt(i), ss2);
+                    System.out.printf("%s{%c}%sc\n\n", ss1, unequalChar, ss2);
+                    return ;
                 }
             }
         }

src/main/java/org/monarchinitiative/hpotools/analysis/HpoStats.java

@@ -0,0 +1,69 @@
+package org.monarchinitiative.hpotools.analysis;
+
+import org.monarchinitiative.phenol.base.PhenolRuntimeException;
+import org.monarchinitiative.phenol.ontology.data.Ontology;
+import org.monarchinitiative.phenol.ontology.data.Term;
+import org.monarchinitiative.phenol.ontology.data.TermId;
+import org.monarchinitiative.phenol.ontology.data.TermSynonym;
+
+import java.util.HashSet;
+import java.util.List;
+import java.util.Optional;
+import java.util.Set;
+
+public class HpoStats {
+    private final Ontology ontology;
+
+    private final String HP_PREFIX = "HP";
+
+    private final Set<Term> nonObsoleteTerms;
+
+    public HpoStats(Ontology ontology) {
+        this.ontology = ontology;
+        nonObsoleteTerms = new HashSet<>();
+        for (TermId tid :ontology.nonObsoleteTermIds()) {
+            Optional<Term> opt = ontology.termForTermId(tid);
+            if (!opt.isPresent()) {
+                throw new PhenolRuntimeException("Could not find term " + tid + " in ontology");
+            }
+            nonObsoleteTerms.add(opt.get());
+        }
+    }
+
+    private int countTermsWithDefinition() {
+        int n_with_def = 0;
+        for (Term term: nonObsoleteTerms) {
+            if (!term.id().getPrefix().equals(HP_PREFIX))
+                continue;
+            String def = term.getDefinition();
+            if (!def.isEmpty()) n_with_def++;
+        }
+        return n_with_def;
+    }
+
+    private int countTermsWithSynonym() {
+        int n_with_synonym = 0;
+        for (Term term: nonObsoleteTerms) {
+            if (!term.id().getPrefix().equals(HP_PREFIX))
+                continue;
+            List<TermSynonym> synonyms = term.getSynonyms();
+            if (!synonyms.isEmpty()) n_with_synonym++;
+        }
+        return n_with_synonym;
+
+    }
+
+    public void printStats() {
+        System.out.println(ontology);
+        int n_non_obsolete = ontology.nonObsoleteTermIdCount();
+        int n_with_def = countTermsWithDefinition();
+        System.out.printf("Terms with definition: %d/%d (%.1f%%).\n",
+                n_with_def, n_non_obsolete, (100.0*n_with_def/n_non_obsolete));
+        int nWithSynonym = countTermsWithSynonym();
+        System.out.printf("Terms with synonyms: %d/%d (%.1f%%).\n",
+                nWithSynonym, n_non_obsolete, (100.0*nWithSynonym/n_non_obsolete));
+
+    }
+
+
+}

src/main/java/org/monarchinitiative/hpotools/analysis/mondo/MondoFromOmimMapper.java

@@ -0,0 +1,85 @@
+package org.monarchinitiative.hpotools.analysis.mondo;
+
+import org.monarchinitiative.phenol.ontology.data.Dbxref;
+import org.monarchinitiative.phenol.ontology.data.Ontology;
+import org.monarchinitiative.phenol.ontology.data.Term;
+import org.monarchinitiative.phenol.ontology.data.TermId;
+
+import java.util.*;
+
+public class MondoFromOmimMapper {
+
+
+    private final Ontology ontology;
+
+    private final Map<String, TermId> omimToMondoMap;
+
+    public MondoFromOmimMapper(Ontology mondo) {
+        ontology = mondo;
+        omimToMondoMap = fromOmimIds();
+    }
+
+
+    /**
+     * @param mondoId Mondo Term
+
+     * @return Omim term id corresponding to the MONDO id (if possible)
+     */
+    private Optional<TermId> getOmimIdIfPossible(TermId mondoId) {
+        if (! ontology.containsTerm(mondoId)) {
+            System.out.println("Obsolete");
+            return Optional.empty();
+        }
+        Optional<Term> opt = ontology.termForTermId(mondoId);
+        if (opt.isPresent()) {
+            Term term = opt.get();
+            Optional<Dbxref> dbxopt = term.getXrefs().stream()
+                    .filter(xr -> xr.getName().startsWith("OMIM:"))
+                    .findFirst();
+            if (dbxopt.isPresent()) {
+                Dbxref dbxref = dbxopt.get();
+                String name = dbxref.getName();
+                if (name.startsWith("OMIM:")) {
+                    return Optional.of(TermId.of(name));
+                }
+            }
+
+        }
+        return Optional.empty();
+    }
+
+    private Map<String, TermId> fromOmimIds() {
+
+        Map<String, TermId> omimToMondoMap = new HashMap<>();
+        for (TermId mondoId : ontology.allTermIds()) {
+            if (mondoId.getPrefix().equals("MONDO")) {
+                Optional<TermId> opt = getOmimIdIfPossible(mondoId);
+                opt.ifPresent(termId -> omimToMondoMap.put(termId.getValue(), mondoId));
+            }
+
+        }
+        return omimToMondoMap;
+    }
+
+
+   public List<String> getMappings(String geneSymbol, String omimId) {
+        List<String> items = new ArrayList<>();
+        if (omimToMondoMap.containsKey(omimId)) {
+            TermId mondoId = omimToMondoMap.get(omimId);
+            Optional<Term> opt = ontology.termForTermId(mondoId);
+            if (opt.isPresent()) {
+                Term term = opt.get();
+                String label = term.getName();
+                items.add(geneSymbol);
+                items.add(omimId);
+                items.add(mondoId.getValue());
+                items.add(label);
+                // TODO extract ORDO -- get Ordo if possible
+            }
+        } else {
+            System.err.println("Warning could not find "+ omimId);
+        }
+
+        return items;
+    }
+}