fix(data): get all allele definitions straight from PharmGKB

src/main/java/org/pharmgkb/pharmcat/definition/model/DefinitionFile.java

@@ -228,6 +228,15 @@ public void removeIgnoredNamedAlleles(DefinitionExemption exemption) {
         .collect(Collectors.toCollection(TreeSet::new));
   }
 
+  /**
+   * Filter out structural variant alleles, they have no definition to match against
+   */
+  public void removeStructuralVariants() {
+    m_namedAlleles = m_namedAlleles.stream()
+        .filter((a) -> !a.isStructuralVariant())
+        .collect(Collectors.toCollection(TreeSet::new));
+  }
+
 
   /**
    * Removes any unused positions and ignored positions specified in {@link DefinitionExemption}.
@@ -324,7 +333,7 @@ public void removeIgnoredPositions(SortedSet<VariantLocus> ignoredPositions, boo
             " because it has no alleles after removing unused/ignored positions");
       } else {
         updatedNamedAlleles.add(new NamedAllele(namedAllele.getId(), namedAllele.getName(), null, cpicAlleles,
-            namedAllele.isReference()));
+            namedAllele.isReference(), namedAllele.isStructuralVariant()));
       }
     }
     m_namedAlleles = updatedNamedAlleles;
@@ -380,7 +389,7 @@ public void doVcfTranslation(VcfHelper vcfHelper) throws IOException {
       if (mustResort) {
         updated = reorderHaplotypeAlleles(na, m_variants, sortedVariants, fixedAlleles);
       } else {
-        updated = new NamedAllele(na.getId(), na.getName(), fixedAlleles, na.getCpicAlleles(), na.isReference());
+        updated = new NamedAllele(na.getId(), na.getName(), fixedAlleles, na.getCpicAlleles(), na.isReference(), na.isStructuralVariant());
       }
       updatedNamedAlleles.add(updated);
     }
@@ -641,6 +650,6 @@ private NamedAllele reorderHaplotypeAlleles(NamedAllele hap, VariantLocus[] oldP
     }
 
     return new NamedAllele(hap.getId(), hap.getName(), alleles, cpicAlleles, hap.getMissingPositions(),
-        hap.isReference(), hap.getNumCombinations(), hap.getNumPartials());
+        hap.isReference(), hap.getNumCombinations(), hap.getNumPartials(), hap.isStructuralVariant());
   }
 }

src/main/java/org/pharmgkb/pharmcat/definition/model/NamedAllele.java

@@ -41,6 +41,9 @@ public class NamedAllele implements Comparable<NamedAllele> {
   @Expose
   @SerializedName(value = "reference", alternate = {"matchesreferencesequence"})
   private final boolean m_isReference;
+  @Expose
+  @SerializedName("structuralVariant")
+  private final boolean m_isStructuralVariant;
   //-- variables after this point are used by NamedAlleleMatcher --//
   /** The set of positions that are missing from this copy of the NamedAllele **/
   private final SortedSet<VariantLocus> m_missingPositions;
@@ -66,20 +69,20 @@ public class NamedAllele implements Comparable<NamedAllele> {
    * Primary constructor.
    * Use this when reading in allele definitions.
    */
-  public NamedAllele(String id, String name, String[] alleles, String[] cpicAlleles, boolean isReference) {
-    this(id, name, alleles, cpicAlleles, Collections.emptySortedSet(), isReference, 0, 0);
+  public NamedAllele(String id, String name, String[] alleles, String[] cpicAlleles, boolean isReference, boolean isStructuralVariant) {
+    this(id, name, alleles, cpicAlleles, Collections.emptySortedSet(), isReference, 0, 0, isStructuralVariant);
   }
 
   public NamedAllele(String id, String name, String[] alleles, String[] cpicAlleles,
-      SortedSet<VariantLocus> missingPositions, boolean isReference) {
-    this(id, name, alleles, cpicAlleles, missingPositions, isReference, 0, 0);
+      SortedSet<VariantLocus> missingPositions, boolean isReference, boolean isStructuralVariant) {
+    this(id, name, alleles, cpicAlleles, missingPositions, isReference, 0, 0, isStructuralVariant);
   }
 
   /**
    * Constructor for duplicating/modifying a {@link NamedAllele}.
    */
   public NamedAllele(String id, String name, String[] alleles, String[] cpicAlleles,
-      SortedSet<VariantLocus> missingPositions, boolean isReference, int numCombinations, int numPartials) {
+      SortedSet<VariantLocus> missingPositions, boolean isReference, int numCombinations, int numPartials, boolean isStructuralVariant) {
     Preconditions.checkNotNull(id);
     Preconditions.checkNotNull(name);
     Preconditions.checkNotNull(cpicAlleles);
@@ -90,6 +93,7 @@ public NamedAllele(String id, String name, String[] alleles, String[] cpicAllele
     m_cpicAlleles = cpicAlleles;
     m_missingPositions = missingPositions;
     m_isReference = isReference;
+    m_isStructuralVariant = isStructuralVariant;
     m_numCombinations = numCombinations;
     m_numPartials = numPartials;
 
@@ -219,6 +223,10 @@ public boolean isReference() {
     return m_isReference;
   }
 
+  public boolean isStructuralVariant() {
+    return m_isStructuralVariant;
+  }
+
 
   public boolean isCombination() {
     return m_numCombinations > 1;

src/main/java/org/pharmgkb/pharmcat/haplotype/MatchData.java

@@ -162,7 +162,7 @@ void marshallHaplotypes(String gene, SortedSet<NamedAllele> allHaplotypes, boole
             .collect(Collectors.toCollection(TreeSet::new));
 
         NamedAllele newHap = new NamedAllele(hap.getId(), hap.getName(), availableAlleles, cpicAlleles,
-            missingPositions, hap.isReference());
+            missingPositions, hap.isReference(), hap.isStructuralVariant());
         newHap.initialize(m_positions);
         if (newHap.getScore() > 0) {
           m_haplotypes.add(newHap);
@@ -216,7 +216,7 @@ void defaultMissingAllelesToReference() {
       }
 
       NamedAllele fixedHap = new NamedAllele(hap.getId(), hap.getName(), newAlleles, cpicAlleles,
-          hap.getMissingPositions(), hap.isReference());
+          hap.getMissingPositions(), hap.isReference(), hap.isStructuralVariant());
       fixedHap.initialize(m_positions, hap.getScore());
       updatedHaplotypes.add(fixedHap);
     }

src/main/java/org/pharmgkb/pharmcat/haplotype/model/BaseMatch.java

@@ -113,13 +113,13 @@ public void finalizeCombinationHaplotype(MatchData matchData, boolean findPartia
       setName("[" + builder + "]");
 
       NamedAllele partialHap = new NamedAllele(hap.getId(), builder.toString(), hap.getAlleles(),
-          hap.getCpicAlleles(), hap.getMissingPositions(), false, hap.getNumCombinations(), partials.size());
+          hap.getCpicAlleles(), hap.getMissingPositions(), false, hap.getNumCombinations(), partials.size(), hap.isStructuralVariant());
       partialHap.initialize(refVariants);
       m_haplotype = partialHap;
 
     } else {
       NamedAllele newHap = new NamedAllele(hap.getId(), hap.getName(), hap.getAlleles(),
-          hap.getCpicAlleles(), hap.getMissingPositions(), hap.isReference(), hap.getNumCombinations(), 0);
+          hap.getCpicAlleles(), hap.getMissingPositions(), hap.isReference(), hap.getNumCombinations(), 0, hap.isStructuralVariant());
       newHap.initialize(refVariants);
       m_haplotype = newHap;
     }

src/main/java/org/pharmgkb/pharmcat/haplotype/model/CombinationMatch.java

@@ -143,7 +143,7 @@ private NamedAllele buildHaplotype(int numPartials, boolean isOffReferencePartia
       }
     }
     NamedAllele na = new NamedAllele(idBuilder.toString(), getName(), alleles, cpicAlleles, missingPositions,
-        false, m_componentHaplotypes.size(), numPartials);
+        false, m_componentHaplotypes.size(), numPartials, false);
     if (isOffReferencePartial) {
       na.initialize(m_refVariants, 0);
     } else {

src/main/java/org/pharmgkb/pharmcat/util/DataManager.java

@@ -48,7 +48,6 @@ public class DataManager {
   public static final Path DEFAULT_EXEMPTIONS_FILE = DEFAULT_DEFINITION_DIR.resolve(EXEMPTIONS_JSON_FILE_NAME);
   public static final String POSITIONS_VCF = "pharmcat_positions.vcf";
   private static final String ALLELES_FILE_NAME = "allele_translations.json";
-  private static final String CPIC_ALLELES_FILE_NAME = "allele_definitions.json";
   private static final String PRESCRIBING_GUIDANCE_FILE_NAME = "prescribing_guidance.json";
   private static final String sf_zipFileName = "pharmcat.zip";
   private static final String sf_googleDocUrlFmt = "https://docs.google.com/spreadsheets/d/%s/export?format=tsv";
@@ -176,13 +175,6 @@ public static void main(String[] args) {
           Path exemptionsJson = allelesDir.resolve(EXEMPTIONS_JSON_FILE_NAME);
           Map<String, DefinitionExemption> exemptionsMap = manager.transformExemptions(exemptionsTsv, exemptionsJson);
 
-          if (!skipDownload) {
-            // download allele definitions
-            // use S3 link to avoid caching problems
-            FileUtils.copyURLToFile(
-                new URL("http://files.cpicpgx.org.s3-us-west-2.amazonaws.com/data/report/current/allele_definitions.json"),
-                downloadDir.resolve(CPIC_ALLELES_FILE_NAME).toFile());
-          }
           // transform allele definitions
           definitionReader = manager.transformAlleleDefinitions(downloadDir, allelesDir, exemptionsMap);
 
@@ -218,8 +210,8 @@ public static void main(String[] args) {
   }
 
 
-  private DefinitionFile[] parseDefinitionFiles(Path downloadDir, String fileName) throws IOException {
-    Path definitionsFile = downloadDir.resolve(fileName);
+  private DefinitionFile[] parseDefinitionFiles(Path downloadDir) throws IOException {
+    Path definitionsFile = downloadDir.resolve(ALLELES_FILE_NAME);
     if (!Files.exists(definitionsFile)) {
       throw new IOException("Cannot find alleles definitions (" + definitionsFile + ")");
     }
@@ -236,12 +228,7 @@ private DefinitionReader transformAlleleDefinitions(Path downloadDir, Path defin
 
     System.out.println("Generating allele definitions...");
     List<DefinitionFile> definitionFiles = new ArrayList<>();
-    for (DefinitionFile df : parseDefinitionFiles(downloadDir, CPIC_ALLELES_FILE_NAME)) {
-      df.setSource(DataSource.CPIC);
-      df.validateAlleleNames();
-      definitionFiles.add(df);
-    }
-    for (DefinitionFile df : parseDefinitionFiles(downloadDir, ALLELES_FILE_NAME)) {
+    for (DefinitionFile df : parseDefinitionFiles(downloadDir)) {
       df.setSource(DataSource.PHARMGKB);
       df.validateAlleleNames();
       definitionFiles.add(df);
@@ -251,13 +238,10 @@ private DefinitionReader transformAlleleDefinitions(Path downloadDir, Path defin
     try (VcfHelper vcfHelper = new VcfHelper()) {
       for (DefinitionFile df : definitionFiles) {
         String gene = df.getGeneSymbol();
-        if (definitionFileMap.containsKey(gene)) {
-          // this will prefer CPIC allele definitions over PharmGKB ones
-          continue;
-        }
-        if (gene.equals("MT-RNR1")) {
-          continue;
-        }
+
+        // always strip structural variants since they are unmatchable
+        df.removeStructuralVariants();
+
         DefinitionExemption exemption = exemptionsMap.get(gene);
         if (exemption != null) {
           if (!exemption.getIgnoredAlleles().isEmpty()) {

src/main/resources/org/pharmgkb/pharmcat/definition/alleles/ABCG2_translation.json

@@ -1,8 +1,8 @@
 {
   "formatVersion": "2",
-  "source": "CPIC",
-  "version": "v1.40.1-1-gf95ca46",
-  "modificationDate": "2024-07-11T23:06:42.472Z",
+  "source": "PHARMGKB",
+  "version": "2024-07-15-21-58",
+  "modificationDate": "2023-02-03T16:20:52.044Z",
   "gene": "ABCG2",
   "orientation": null,
   "chromosome": "chr4",
@@ -32,27 +32,29 @@
   "namedAlleles": [
     {
       "name": "rs2231142 reference (G)",
-      "id": "1508724",
+      "id": "PA166287823",
       "alleles": [
         "G"
       ],
       "cpicAlleles": [
         "G"
       ],
       "reference": true,
+      "structuralVariant": false,
       "numCombinations": 0,
       "numPartials": 0
     },
     {
       "name": "rs2231142 variant (T)",
-      "id": "1508726",
+      "id": "PA166287824",
       "alleles": [
         "T"
       ],
       "cpicAlleles": [
         "T"
       ],
       "reference": false,
+      "structuralVariant": false,
       "numCombinations": 0,
       "numPartials": 0
     }

src/main/resources/org/pharmgkb/pharmcat/definition/alleles/CACNA1S_translation.json

@@ -1,8 +1,8 @@
 {
   "formatVersion": "2",
-  "source": "CPIC",
-  "version": "v1.40.1-1-gf95ca46",
-  "modificationDate": "2024-07-11T23:06:42.624Z",
+  "source": "PHARMGKB",
+  "version": "2024-07-15-21-58",
+  "modificationDate": "2021-10-22T20:00:04.974Z",
   "gene": "CACNA1S",
   "orientation": null,
   "chromosome": "chr1",
@@ -51,7 +51,7 @@
   "namedAlleles": [
     {
       "name": "Reference",
-      "id": "777262",
+      "id": "PA166180429",
       "alleles": [
         "C",
         "G"
@@ -61,12 +61,13 @@
         "G"
       ],
       "reference": true,
+      "structuralVariant": false,
       "numCombinations": 0,
       "numPartials": 0
     },
     {
       "name": "c.520C>T",
-      "id": "777264",
+      "id": "PA166180430",
       "alleles": [
         null,
         "A"
@@ -76,12 +77,13 @@
         "A"
       ],
       "reference": false,
+      "structuralVariant": false,
       "numCombinations": 0,
       "numPartials": 0
     },
     {
       "name": "c.3257G>A",
-      "id": "777266",
+      "id": "PA166180431",
       "alleles": [
         "T",
         null
@@ -91,6 +93,7 @@
         null
       ],
       "reference": false,
+      "structuralVariant": false,
       "numCombinations": 0,
       "numPartials": 0
     }