From 91bdf069d3d6d67f03bedd757e4b9b80a75b296d Mon Sep 17 00:00:00 2001
From: Eric Banks <ebanks@broadinstitute.org>
Date: Mon, 13 Jan 2014 22:13:53 -0500
Subject: [PATCH] Some updates to CRCV.

1. Throw a user error when the input data for a given genotype does not contain PLs.
2. Add VCF header line for --dbsnp input
3. Need to check that the UG result is not null
4. Don't error out at positions with no gVCFs (which is possible when using a dbSNP rod)
---
 .../variantutils/CombineReferenceCalculationVariants.java  | 7 +++++--
 .../sting/utils/variant/GATKVariantContextUtils.java       | 7 +++++--
 2 files changed, 10 insertions(+), 4 deletions(-)

diff --git a/protected/gatk-protected/src/main/java/org/broadinstitute/sting/gatk/walkers/variantutils/CombineReferenceCalculationVariants.java b/protected/gatk-protected/src/main/java/org/broadinstitute/sting/gatk/walkers/variantutils/CombineReferenceCalculationVariants.java
index 2a004aaca..66bf562d7 100644
--- a/protected/gatk-protected/src/main/java/org/broadinstitute/sting/gatk/walkers/variantutils/CombineReferenceCalculationVariants.java
+++ b/protected/gatk-protected/src/main/java/org/broadinstitute/sting/gatk/walkers/variantutils/CombineReferenceCalculationVariants.java
@@ -153,9 +153,12 @@ public class CombineReferenceCalculationVariants extends RodWalker<VariantContex
     public void initialize() {
         // take care of the VCF headers
         final Map<String, VCFHeader> vcfRods = GATKVCFUtils.getVCFHeadersFromRods(getToolkit());
-        final Set<String> samples = SampleUtils.getSampleList(vcfRods, GATKVariantContextUtils.GenotypeMergeType.REQUIRE_UNIQUE);
         final Set<VCFHeaderLine> headerLines = VCFUtils.smartMergeHeaders(vcfRods.values(), true);
         headerLines.addAll(Arrays.asList(ChromosomeCountConstants.descriptions));
+        if ( dbsnp != null && dbsnp.dbsnp.isBound() )
+            VCFStandardHeaderLines.addStandardInfoLines(headerLines, true, VCFConstants.DBSNP_KEY);
+
+        final Set<String> samples = SampleUtils.getSampleList(vcfRods, GATKVariantContextUtils.GenotypeMergeType.REQUIRE_UNIQUE);
         final VCFHeader vcfHeader = new VCFHeader(headerLines, samples);
         vcfWriter.writeHeader(vcfHeader);
 
@@ -204,7 +207,7 @@ public class CombineReferenceCalculationVariants extends RodWalker<VariantContex
             result = genotypingEngine.calculateGenotypes(result);
 
         // if it turned monomorphic and we don't want such sites, quit
-        if ( !INCLUDE_NON_VARIANTS && result.isMonomorphicInSamples() )
+        if ( result == null || (!INCLUDE_NON_VARIANTS && result.isMonomorphicInSamples()) )
             return null;
 
         // re-annotate it
diff --git a/public/gatk-framework/src/main/java/org/broadinstitute/sting/utils/variant/GATKVariantContextUtils.java b/public/gatk-framework/src/main/java/org/broadinstitute/sting/utils/variant/GATKVariantContextUtils.java
index e8ce80d07..e1ed63bc0 100644
--- a/public/gatk-framework/src/main/java/org/broadinstitute/sting/utils/variant/GATKVariantContextUtils.java
+++ b/public/gatk-framework/src/main/java/org/broadinstitute/sting/utils/variant/GATKVariantContextUtils.java
@@ -1045,8 +1045,9 @@ public class GATKVariantContextUtils {
      * @return new VariantContext representing the merge of all VCs or null if it not relevant
      */
     public static VariantContext referenceConfidenceMerge(final List<VariantContext> VCs, final GenomeLoc loc, final Byte refBase) {
-
-        if ( VCs == null || VCs.size() == 0 ) throw new IllegalArgumentException("VCs cannot be null or empty");
+        // this can happen if e.g. you are using a dbSNP file that spans a region with no gVCFs
+        if ( VCs == null || VCs.size() == 0 )
+            return null;
 
         // establish the baseline info (sometimes from the first VC)
         final VariantContext first = VCs.get(0);
@@ -1536,6 +1537,8 @@ public class GATKVariantContextUtils {
                                                     final List<Allele> remappedAlleles,
                                                     final List<Allele> targetAlleles) {
         for ( final Genotype g : VC.getGenotypes() ) {
+	    if ( !g.hasPL() )
+		throw new UserException("cannot merge genotypes from samples without PLs; sample " + g.getSampleName() + " does not have likelihoods at position " + VC.getChr() + ":" + VC.getStart());
 
             // only add if the name is new
             final String name = g.getSampleName();