BQSR: avoid throwing an error when read group is missing in the recal table, and some refactoring.

src/main/java/org/broadinstitute/hellbender/tools/ApplyBQSRArgumentCollection.java

@@ -8,6 +8,7 @@
  */
 public class ApplyBQSRArgumentCollection extends ApplyBQSRUniqueArgumentCollection {
     private static final long serialVersionUID = 1L;
+    public static final String USE_ORIGINAL_QUALITIES_LONG_NAME = "use-original-qualities";
 
     /**
      * This flag tells GATK not to modify quality scores less than this value. Instead they will be written out
@@ -26,6 +27,6 @@ public class ApplyBQSRArgumentCollection extends ApplyBQSRUniqueArgumentCollecti
      * are stored in the OQ tag, if they are present, rather than use the post-recalibration quality scores. If no OQ
      * tag is present for a read, the standard quality score will be used.
      */
-    @Argument(fullName="use-original-qualities", shortName = "OQ", doc = "Use the base quality scores from the OQ tag", optional = true)
+    @Argument(fullName=USE_ORIGINAL_QUALITIES_LONG_NAME, shortName = "OQ", doc = "Use the base quality scores from the OQ tag", optional = true)
     public Boolean useOriginalBaseQualities = false;
 }

src/main/java/org/broadinstitute/hellbender/tools/ApplyBQSRUniqueArgumentCollection.java

@@ -17,6 +17,9 @@
 public class ApplyBQSRUniqueArgumentCollection implements Serializable {
     private static final long serialVersionUID = 1L;
 
+    public static final String STATIC_QUANTIZED_QUALS_LONG_NAME = "static-quantized-quals";
+    public static final String ALLOW_MISSING_READ_GROUPS_LONG_NAME = "allow-missing-read-group";
+
     /**
      * Turns on the base quantization module. It requires a recalibration report.
      *
@@ -31,9 +34,11 @@ public class ApplyBQSRUniqueArgumentCollection implements Serializable {
     /**
      * Static quantized quals are entirely separate from the quantize_qual option which uses dynamic binning.
      * The two types of binning should not be used together.
+     *
+     * For example, the Warp germline pipeline uses the static bins { 10, 20, 30, 40 }
      */
     @Advanced
-    @Argument(fullName="static-quantized-quals", doc = "Use static quantized quality scores to a given number of levels (with -"+ StandardArgumentDefinitions.BQSR_TABLE_SHORT_NAME+ ")", optional=true, mutex = "quantize-quals")
+    @Argument(fullName=STATIC_QUANTIZED_QUALS_LONG_NAME, doc = "Use static quantized quality scores to a given number of levels (with -"+ StandardArgumentDefinitions.BQSR_TABLE_SHORT_NAME+ ")", optional=true, mutex = "quantize-quals")
     public List<Integer> staticQuantizationQuals = new ArrayList<>();
 
     /**
@@ -61,6 +66,13 @@ public class ApplyBQSRUniqueArgumentCollection implements Serializable {
     @Argument(fullName = "global-qscore-prior", doc = "Global Qscore Bayesian prior to use for BQSR", optional = true)
     public double globalQScorePrior = -1.0;
 
+    /**
+     * If set to true, do not throw an error upon encountering a read with a read group that's not in the recalibration table.
+     * Instead, simply set the quantized original base qualities as the recalibrated base qualities.
+     */
+    @Argument(fullName = ALLOW_MISSING_READ_GROUPS_LONG_NAME, doc = "Do not throw an error when encountering a read group not in the recal table", optional = true)
+    public boolean allowMissingReadGroups = false;
+
     /**
      * Combine the extra arguments in {@link ApplyBQSRArgumentCollection} that are not in this {@link ApplyBQSRUniqueArgumentCollection}
      * from the given {@link RecalibrationArgumentCollection} to create a {@link ApplyBQSRArgumentCollection}.
@@ -75,6 +87,8 @@ public ApplyBQSRArgumentCollection toApplyBQSRArgumentCollection(RecalibrationAr
         ret.roundDown = this.roundDown;
         ret.emitOriginalQuals = this.emitOriginalQuals;
         ret.globalQScorePrior = this.globalQScorePrior;
+        ret.allowMissingReadGroups = this.allowMissingReadGroups;
+
         // include all the fields from RecalibrationArgumentCollection
         ret.PRESERVE_QSCORES_LESS_THAN = bqsrArgs.PRESERVE_QSCORES_LESS_THAN;
         ret.useOriginalBaseQualities = bqsrArgs.useOriginalBaseQualities;

src/main/java/org/broadinstitute/hellbender/tools/spark/transforms/BaseRecalibratorSparkFn.java

@@ -40,7 +40,8 @@ public static RecalibrationReport apply(final JavaPairRDD<GATKRead, Iterable<GAT
                 RecalibrationTables::inPlaceCombine,
                 Math.max(1, (int)(Math.log(unmergedTables.partitions().size()) / Math.log(2))));
 
-        BaseRecalibrationEngine.finalizeRecalibrationTables(combinedTables);
+        BaseRecalibrationEngine.collapseQualityScoreTableToReadGroupTable(combinedTables.getQualityScoreTable(), combinedTables.getReadGroupTable());
+        BaseRecalibrationEngine.roundTableValues(combinedTables);
 
         final QuantizationInfo quantizationInfo = new QuantizationInfo(combinedTables, recalArgs.QUANTIZING_LEVELS);
 

src/main/java/org/broadinstitute/hellbender/tools/walkers/bqsr/ApplyBQSR.java

@@ -81,7 +81,7 @@ public final class ApplyBQSR extends ReadWalker{
 
     @Argument(fullName = StandardArgumentDefinitions.OUTPUT_LONG_NAME, shortName = StandardArgumentDefinitions.OUTPUT_SHORT_NAME, doc="Write output to this file")
     @WorkflowOutput(optionalCompanions = {StandardArgumentDefinitions.OUTPUT_INDEX_COMPANION})
-    public GATKPath OUTPUT;
+    public GATKPath output;
 
     /**
      * This argument is required for recalibration of base qualities. The recalibration table is a file produced by
@@ -90,7 +90,7 @@ public final class ApplyBQSR extends ReadWalker{
      */
     @Argument(fullName=StandardArgumentDefinitions.BQSR_TABLE_LONG_NAME, shortName=StandardArgumentDefinitions.BQSR_TABLE_SHORT_NAME, doc="Input recalibration table for BQSR")
     @WorkflowInput
-    public File BQSR_RECAL_FILE;
+    public File bqsrRecalFile;
 
     /**
      * Command-line arguments to fine tune the recalibration.
@@ -105,12 +105,12 @@ public final class ApplyBQSR extends ReadWalker{
      */
     @Override
     public ReadTransformer makePostReadFilterTransformer(){
-        return new BQSRReadTransformer(getHeaderForReads(), BQSR_RECAL_FILE, bqsrArgs);
+        return new BQSRReadTransformer(getHeaderForReads(), bqsrRecalFile, bqsrArgs);
     }
 
     @Override
     public void onTraversalStart() {
-        outputWriter = createSAMWriter(OUTPUT, true);
+        outputWriter = createSAMWriter(output, true);
         Utils.warnOnNonIlluminaReadGroups(getHeaderForReads(), logger);
     }