#450 stabilising sort order in VCF output

PapenfussLab · Feb 16, 2021 · 7b08669 · 7b08669
1 parent 9f0c15a
commit 7b08669
Show file tree

Hide file tree

Showing 2 changed files with 35 additions and 3 deletions.
diff --git a/src/main/java/au/edu/wehi/idsv/IdsvVariantContext.java b/src/main/java/au/edu/wehi/idsv/IdsvVariantContext.java
@@ -1,13 +1,17 @@
 package au.edu.wehi.idsv;
 
+import au.edu.wehi.idsv.debruijn.VariantEvidence;
 import au.edu.wehi.idsv.vcf.VcfFormatAttributes;
 import au.edu.wehi.idsv.vcf.VcfInfoAttributes;
 import com.google.common.collect.ComparisonChain;
 import com.google.common.collect.Ordering;
 import com.google.common.primitives.Doubles;
 import htsjdk.samtools.SAMSequenceDictionary;
+import htsjdk.variant.variantcontext.Allele;
 import htsjdk.variant.variantcontext.VariantContext;
 
+import java.util.List;
+
 /**
  * Generates variant context records from the underlying @link {@link VariantContext}
  * @author Daniel Cameron
@@ -81,6 +85,8 @@ public int compare(IdsvVariantContext o1, IdsvVariantContext o2) {
 			        .compare(o1.getReferenceIndex(), o2.getReferenceIndex())
 			        .compare(o1.getStart(), o2.getStart())
 			        .compare(o1.getEnd(), o2.getEnd())
+					.compare(o1.getReference(), o2.getReference())
+					.compare(getOrderingAllele(o1), getOrderingAllele(o2))
 			        .compare(o1.getID(), o2.getID())
 			        .result();
 		  }
@@ -91,6 +97,7 @@ public int compare(IdsvVariantContext o1, IdsvVariantContext o2) {
 			        .compare(o1.getReferenceIndex(), o2.getReferenceIndex())
 			        .compare(o1.getEnd(), o2.getEnd())
 			        .compare(o1.getStart(), o2.getStart())
+					.compare(getOrderingAllele(o1), getOrderingAllele(o2))
 			        .compare(o1.getID(), o2.getID())
 			        .result();
 		  }
@@ -114,9 +121,16 @@ public int compare(VariantContext o1, VariantContext o2) {
 				        .compare(dictionary.getSequenceIndex(o1.getContig()), dictionary.getSequenceIndex(o2.getContig()))
 				        .compare(o1.getEnd(), o2.getEnd())
 				        .compare(o1.getStart(), o2.getStart())
+						.compare(getOrderingAllele(o1), getOrderingAllele(o2))
 				        .compare(o1.getID(), o2.getID())
 				        .result();
 			  }
 		};
-	}	
+	}
+	private static Allele getOrderingAllele(VariantContext v) {
+		List<Allele> alt = v.getAlleles();
+		if (alt.size() >= 2) return alt.get(1);
+		if (alt.size() >= 1) return alt.get(0);
+		return null;
+	}
 }
diff --git a/src/test/java/au/edu/wehi/idsv/IdsvVariantContextTest.java b/src/test/java/au/edu/wehi/idsv/IdsvVariantContextTest.java
@@ -1,12 +1,14 @@
 package au.edu.wehi.idsv;
 
+import com.google.common.collect.ImmutableList;
+import com.google.common.collect.Ordering;
 import htsjdk.variant.variantcontext.VariantContext;
 import org.junit.Test;
 
+import java.util.ArrayList;
 import java.util.List;
 
-import static org.junit.Assert.assertEquals;
-import static org.junit.Assert.assertTrue;
+import static org.junit.Assert.*;
 
 public class IdsvVariantContextTest extends TestHelper {
 	@Test
@@ -91,4 +93,20 @@ public void getAttributeAsIntListOffset_should_default_should_get_ith_element()
 		TestIdsvVariantContext vc = new TestIdsvVariantContext(minimalVariant().attribute("intlist", L("1", "2")).make());
 		assertEquals(1, vc.getAttributeAsIntListOffset("intlist", 0, 7));
 	}
+	@Test
+	public void ByLocation_should_have_stable_sort_order() {
+		IdsvVariantContext v1 = (IdsvVariantContext)new IdsvVariantContextBuilder(getContext()).chr("polyA").start(1).stop(1).alleles("A", "C").make();
+		IdsvVariantContext v2 = (IdsvVariantContext)new IdsvVariantContextBuilder(getContext()).chr("polyA").start(1).stop(1).alleles("A", "T").make();
+		for (Ordering<IdsvVariantContext> order : ImmutableList.of(IdsvVariantContext.ByLocationStart, IdsvVariantContext.ByLocationEnd)) {
+			List<IdsvVariantContext> list1 = new ArrayList<>();
+			list1.add(v1);
+			list1.add(v2);
+			List<IdsvVariantContext> list2 = new ArrayList<>();
+			list2.add(v2);
+			list2.add(v1);
+			list1.sort(order);
+			list2.sort(order);
+			assertEquals(list1, list2);
+		}
+	}
 }