Do not strip ECN fields from truth records in SVConcordance (#9161)

Author: mwalker174

src/main/java/org/broadinstitute/hellbender/tools/walkers/sv/SVConcordance.java

@@ -62,6 +62,11 @@
  * of the specific fields. For multi-allelic CNVs, only a copy state concordance metric is
  * annotated. Allele frequencies will be recalculated automatically if unavailable in the provided VCFs.
  *
+ * Minimum matching criteria can be specified as in {@link SVCluster} (e.g. reciprocal overlap). These serve to
+ * improve computational efficiency, but may be set to relaxed values that can be iterated on post hoc. Note that
+ * this set of parameters includes minimum sample overlap (Jaccard index of carrier samples), which should generally
+ * be set to 0 for concordance analysis.
+ *
  * This tool also allows supports stratification of the SVs into groups with specified matching criteria including SV type,
  * size range, and interval overlap. Please see the {@link GroupedSVCluster} tool documentation for further details
  * on how to specify stratification groups. Stratification only affects the criteria applied to each "eval" SV. In
@@ -307,6 +312,9 @@ private static Genotype stripTruthGenotype(final Genotype genotype) {
         if (genotype.hasExtendedAttribute(GATKSVVCFConstants.COPY_NUMBER_FORMAT)) {
             builder.attribute(GATKSVVCFConstants.COPY_NUMBER_FORMAT, genotype.getExtendedAttribute(GATKSVVCFConstants.COPY_NUMBER_FORMAT));
         }
+        if (genotype.hasExtendedAttribute(GATKSVVCFConstants.EXPECTED_COPY_NUMBER_FORMAT)) {
+            builder.attribute(GATKSVVCFConstants.EXPECTED_COPY_NUMBER_FORMAT, genotype.getExtendedAttribute(GATKSVVCFConstants.EXPECTED_COPY_NUMBER_FORMAT));
+        }
         return builder.make();
     }
 

src/test/java/org/broadinstitute/hellbender/tools/walkers/sv/SVConcordanceIntegrationTest.java

@@ -154,6 +154,30 @@ public void testRefPanel() {
         Assert.assertTrue(checkedVariantsSet.contains("ref_panel_1kg.chr22.final_cleanup_INS_chr22_100"));
     }
 
+    // Regression test with non-zero sample overlap criteria, which require the ECN FORMAT field to be present after
+    //   stripping truth record genotypes of extra fields.
+    @Test
+    public void testRefPanelWithSampleOverlap() {
+        final File output = createTempFile("concord", ".vcf.gz");
+        final String evalVcfPath = getToolTestDataDir() + "ref_panel_1kg.cleaned.gatk.chr22_chrY.vcf.gz";
+        final String truthVcfPath = getToolTestDataDir() + "ref_panel_1kg.raw_calls.chr22_chrY.vcf.gz";
+        final ArgumentsBuilder args = new ArgumentsBuilder()
+                .addOutput(output)
+                .add(StandardArgumentDefinitions.SEQUENCE_DICTIONARY_NAME, GATKBaseTest.FULL_HG38_DICT)
+                .add(SVClusterEngineArgumentsCollection.DEPTH_SAMPLE_OVERLAP_FRACTION_NAME, 0.1)
+                .add(SVClusterEngineArgumentsCollection.DEPTH_INTERVAL_OVERLAP_FRACTION_NAME, 0.5)
+                .add(SVClusterEngineArgumentsCollection.DEPTH_BREAKEND_WINDOW_NAME, 10000000)
+                .add(SVClusterEngineArgumentsCollection.MIXED_SAMPLE_OVERLAP_FRACTION_NAME, 0.1)
+                .add(SVClusterEngineArgumentsCollection.MIXED_INTERVAL_OVERLAP_FRACTION_NAME, 0.1)
+                .add(SVClusterEngineArgumentsCollection.MIXED_BREAKEND_WINDOW_NAME, 2000)
+                .add(SVClusterEngineArgumentsCollection.PESR_SAMPLE_OVERLAP_FRACTION_NAME, 0.1)
+                .add(SVClusterEngineArgumentsCollection.PESR_INTERVAL_OVERLAP_FRACTION_NAME, 0.1)
+                .add(SVClusterEngineArgumentsCollection.PESR_BREAKEND_WINDOW_NAME, 500)
+                .add(AbstractConcordanceWalker.TRUTH_VARIANTS_LONG_NAME, truthVcfPath)
+                .add(AbstractConcordanceWalker.EVAL_VARIANTS_SHORT_NAME, evalVcfPath);
+        runCommandLine(args, SVConcordance.class.getSimpleName());
+    }
+
     @Test
     public void testRefPanelStratified() {
         final File output = createTempFile("concord", ".vcf.gz");