New tools - modular implementations

src/main/java/org/broadinstitute/hellbender/tools/walkers/sv/SVCleanPt4.java

@@ -1,7 +1,15 @@
 package org.broadinstitute.hellbender.tools.walkers.sv;
 
-import htsjdk.variant.variantcontext.*;
+import htsjdk.variant.variantcontext.Allele;
+
+import htsjdk.variant.variantcontext.Genotype;
+import htsjdk.variant.variantcontext.GenotypeBuilder;
+import htsjdk.variant.variantcontext.VariantContext;
+import htsjdk.variant.variantcontext.VariantContextBuilder;
 import htsjdk.variant.variantcontext.writer.VariantContextWriter;
+import htsjdk.variant.vcf.VCFHeader;
+import htsjdk.variant.vcf.VCFHeaderLineType;
+import htsjdk.variant.vcf.VCFInfoHeaderLine;
 
 import htsjdk.variant.vcf.*;
 import org.broadinstitute.barclay.argparser.Argument;

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

@@ -0,0 +1,106 @@
+package org.broadinstitute.hellbender.tools.walkers.sv;
+
+import htsjdk.variant.variantcontext.*;
+import htsjdk.variant.variantcontext.writer.VariantContextWriter;
+import org.broadinstitute.barclay.argparser.Argument;
+import org.broadinstitute.barclay.argparser.BetaFeature;
+import org.broadinstitute.barclay.argparser.CommandLineProgramProperties;
+import org.broadinstitute.barclay.help.DocumentedFeature;
+import org.broadinstitute.hellbender.cmdline.StandardArgumentDefinitions;
+import org.broadinstitute.hellbender.cmdline.programgroups.StructuralVariantDiscoveryProgramGroup;
+import org.broadinstitute.hellbender.engine.*;
+import org.broadinstitute.hellbender.tools.spark.sv.utils.GATKSVVCFConstants;
+
+import java.util.List;
+import java.util.ArrayList;
+
+/**
+ * Completes an initial series of cleaning steps for a VCF produced by the GATK-SV pipeline.
+ *
+ * <h3>Inputs</h3>
+ * <ul>
+ *     <li>
+ *         VCF containing structural variant (SV) records from the GATK-SV pipeline.
+ *     </li>
+ *     <li>
+ *         TODO
+ *     </li>
+ * </ul>
+ *
+ * <h3>Output</h3>
+ * <ul>
+ *     <li>
+ *         Cleansed VCF.
+ *     </li>
+ * </ul>
+ *
+ * <h3>Usage Example</h3>
+ * <pre>
+ *     TODO
+ * </pre>
+ *
+ * <h3>Processing Steps</h3>
+ * <ol>
+ *     <li>
+ *         TODO
+ *     </li>
+ * </ol>
+ */
+@CommandLineProgramProperties(
+        summary = "Clean and format SV VCF",
+        oneLineSummary = "Clean and format SV VCF",
+        programGroup = StructuralVariantDiscoveryProgramGroup.class
+)
+@BetaFeature
+@DocumentedFeature
+public class SVReviseAbnormalAllosomes extends VariantWalker {
+    @Argument(
+            fullName = StandardArgumentDefinitions.OUTPUT_LONG_NAME,
+            shortName = StandardArgumentDefinitions.OUTPUT_SHORT_NAME,
+            doc = "Output VCF name"
+    )
+    private GATKPath outputVcf;
+
+    private VariantContextWriter vcfWriter;
+
+    @Override
+    public void onTraversalStart() {
+        vcfWriter = createVCFWriter(outputVcf);
+        vcfWriter.writeHeader(getHeaderForVariants());
+    }
+
+    @Override
+    public void closeTool() {
+        if (vcfWriter != null) {
+            vcfWriter.close();
+        }
+    }
+
+    @Override
+    public void apply(final VariantContext variant, final ReadsContext readsContext, final ReferenceContext referenceContext, final FeatureContext featureContext) {
+        VariantContextBuilder builder = new VariantContextBuilder(variant);
+        if (!variant.getAttributeAsBoolean(GATKSVVCFConstants.REVISED_EVENT, false)) {
+            processRevisedSex(variant, builder);
+        }
+        vcfWriter.add(builder.make());
+    }
+
+    private void processRevisedSex(final VariantContext variant, final VariantContextBuilder builder) {
+        List<Genotype> genotypes = variant.getGenotypes();
+        List<Genotype> updatedGenotypes = new ArrayList<>(genotypes.size());
+        for (Genotype genotype : genotypes) {
+            if (Integer.parseInt(genotype.getExtendedAttribute(GATKSVVCFConstants.RD_CN, 0).toString()) > 0) {
+                int newRdCn = Integer.parseInt(genotype.getExtendedAttribute(GATKSVVCFConstants.RD_CN).toString()) - 1;
+                GenotypeBuilder gb = new GenotypeBuilder(genotype);
+                gb.attribute(GATKSVVCFConstants.RD_CN, newRdCn);
+                if (genotype.hasExtendedAttribute(GATKSVVCFConstants.COPY_NUMBER_FORMAT)) {
+                    gb.attribute(GATKSVVCFConstants.COPY_NUMBER_FORMAT, newRdCn);
+                }
+                updatedGenotypes.add(gb.make());
+            } else {
+                updatedGenotypes.add(genotype);
+            }
+        }
+        builder.genotypes(genotypes);
+    }
+}