do not sort VCF samples alphabetically

pom.xml

@@ -61,7 +61,7 @@
 
         <amber.version>2.2</amber.version>
         <cobalt.version>1.5</cobalt.version>
-        <purple.version>2.22</purple.version>
+        <purple.version>2.23</purple.version>
 
         <immutables.version>2.4.4</immutables.version>
         <htsjdk.version>2.12.0</htsjdk.version>

purity-ploidy-estimator/README.md

@@ -55,8 +55,8 @@ Argument | Default | Description
 run_dir | None | If provided, parameters `amber`, `cobalt` and `output_dir` are no longer mandatory and will have default values of <run_dir>/amber, <run_dir>/cobalt and <run_dir>/purple respectively unless overridden.
 threads | 2 | Number of threads to use.
 somatic_vcf | None | Optional location of somatic variants vcf.  Sample name must match tumor parameter. GZ files supported.
-structural_vcf | None | Optional location of structural variants vcf. Sample name must match tumor parameter. GZ files supported.
-sv_recovery_vcf | None | Optional location of structural variants recovery vcf. Sample name must match tumor parameter. GZ files supported.
+structural_vcf | None | Optional location of structural variants vcf. GZ files supported.
+sv_recovery_vcf | None | Optional location of structural variants recovery vcf. GZ files supported.
 circos | None | Optional path to circos binary. When supplied, circos graphs will be written to <output_dir>/plot
 db_enabled | None | This parameter has no arguments. Optionally include if you wish to persist results to a database. Database initialization script can be found [here](https://github.com/hartwigmedical/hmftools/blob/master/patient-db/src/main/resources/generate_database.sql).
 db_user | None | Database username. Mandatory if db_enabled.
@@ -502,6 +502,8 @@ Threads | Elapsed Time| CPU Time | Peak Mem
 
 
 ## Version History
+- 2.23
+  - Fixed bug where SV VCF samples were being sorted into alphabetical order. Now they will be in same order as input VCF.
 - 2.22
   - Added new tool to annotate SNPs and INDELS in strelka output with AD field. Will **not** override existing AD values. 
   - Example Usage: `java -Xmx4G -cp purple.jar com.hartwig.hmftools.purple.tools.AnnotateStrelkaWithAllelicDepth -in strelka.vcf -out strelka.annotated.vcf`

purity-ploidy-estimator/src/main/java/com/hartwig/hmftools/purple/PurpleStructuralVariantSupplier.java

@@ -16,6 +16,7 @@
 import java.util.TreeSet;
 import java.util.function.Predicate;
 
+import com.google.common.annotations.VisibleForTesting;
 import com.google.common.collect.ListMultimap;
 import com.google.common.collect.Lists;
 import com.hartwig.hmftools.common.chromosome.Chromosome;
@@ -82,7 +83,6 @@ class PurpleStructuralVariantSupplier {
     private static final Allele INCREASING_ALLELE = Allele.create(".N", false);
     private static final Allele DECREASING_ALLELE = Allele.create("N.", false);
 
-    private final String purpleVersion;
     private final String outputVCF;
     private final Optional<VCFHeader> header;
     private final TreeSet<VariantContext> variantContexts;
@@ -95,14 +95,12 @@ class PurpleStructuralVariantSupplier {
         header = Optional.empty();
         variantContexts = new TreeSet<>();
         outputVCF = Strings.EMPTY;
-        purpleVersion = "0";
     }
 
     PurpleStructuralVariantSupplier(@NotNull final String version, @NotNull final String templateVCF, @NotNull final String outputVCF) {
         final VCFFileReader vcfReader = new VCFFileReader(new File(templateVCF), false);
         this.outputVCF = outputVCF;
-        purpleVersion = version;
-        header = Optional.of(generateOutputHeader(vcfReader.getFileHeader()));
+        header = Optional.of(generateOutputHeader(version, vcfReader.getFileHeader()));
         variantContexts = new TreeSet<>(new VCComparator(header.get().getSequenceDictionary()));
         for (VariantContext context : vcfReader) {
             variantContexts.add(context);
@@ -294,8 +292,9 @@ public List<StructuralVariant> variants() {
     }
 
     @NotNull
-    private VCFHeader generateOutputHeader(@NotNull final VCFHeader template) {
-        final VCFHeader outputVCFHeader = new VCFHeader(template.getMetaDataInInputOrder(), template.getSampleNamesInOrder());
+    @VisibleForTesting
+    static VCFHeader generateOutputHeader(@NotNull final String purpleVersion, @NotNull final VCFHeader template) {
+        final VCFHeader outputVCFHeader = new VCFHeader(template.getMetaDataInInputOrder(), template.getGenotypeSamples());
         outputVCFHeader.addMetaDataLine(new VCFHeaderLine("purpleVersion", purpleVersion));
 
         outputVCFHeader.addMetaDataLine(VCFStandardHeaderLines.getFormatLine("GT"));

purity-ploidy-estimator/src/test/java/com/hartwig/hmftools/purple/PurpleStructuralVariantSupplierTest.java

@@ -0,0 +1,24 @@
+package com.hartwig.hmftools.purple;
+
+import static org.junit.Assert.assertEquals;
+
+import java.util.Collections;
+
+import com.google.common.collect.Lists;
+
+import org.junit.Test;
+
+import htsjdk.variant.vcf.VCFHeader;
+
+public class PurpleStructuralVariantSupplierTest {
+
+    @Test
+    public void testHeaderSamplesAreNotSorted() {
+        final VCFHeader outOfOrderHeader = new VCFHeader(Collections.emptySet(), Lists.newArrayList("BBBBB", "AAAAA"));
+        final VCFHeader victim = PurpleStructuralVariantSupplier.generateOutputHeader("2.23", outOfOrderHeader);
+        assertEquals(2, victim.getGenotypeSamples().size());
+        assertEquals("BBBBB", victim.getGenotypeSamples().get(0));
+        assertEquals("AAAAA", victim.getGenotypeSamples().get(1));
+    }
+
+}