zzh
/
gatk-3.8
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Merge pull request #240 from broadinstitute/rp_gga_implies_both_models 

Bugfix for GGA mode in UG silently ignoring indels
This commit is contained in:
Ryan Poplin 2013-05-24 11:13:57 -07:00
parent b7fab31e25 85905dba92
commit 3516daff52
2 changed files with 18 additions and 22 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

38
protected/java/src/org/broadinstitute/sting/gatk/walkers/genotyper/UnifiedGenotyperEngine.java
View File

@ -83,6 +83,9 @@ public class UnifiedGenotyperEngine {
public static final double HUMAN_SNP_HETEROZYGOSITY = 1e-3;
public static final double HUMAN_INDEL_HETEROZYGOSITY = 1e-4;
private static final int SNP_MODEL = 0;
private static final int INDEL_MODEL = 1;
public enum OUTPUT_MODE {
/** produces calls only at variant sites */
EMIT_VARIANTS_ONLY,
@ -693,13 +696,13 @@ public class UnifiedGenotyperEngine {
}
private void determineGLModelsToUse() {
String modelPrefix = "";
if ( !UAC.GLmodel.name().contains(GPSTRING) && UAC.samplePloidy != GATKVariantContextUtils.DEFAULT_PLOIDY )
modelPrefix = GPSTRING;
if ( UAC.GLmodel.name().toUpperCase().contains("BOTH") ) {
modelPrefix += UAC.GLmodel.name().toUpperCase().replaceAll("BOTH","");
// GGA mode => must initialize both the SNP and indel models
if ( UAC.GenotypingMode == GenotypeLikelihoodsCalculationModel.GENOTYPING_MODE.GENOTYPE_GIVEN_ALLELES ||
UAC.GLmodel.name().toUpperCase().contains("BOTH") ) {
modelsToUse.add(GenotypeLikelihoodsCalculationModel.Model.valueOf(modelPrefix+"SNP"));
modelsToUse.add(GenotypeLikelihoodsCalculationModel.Model.valueOf(modelPrefix+"INDEL"));
}
@ -712,31 +715,24 @@ public class UnifiedGenotyperEngine {
private List<GenotypeLikelihoodsCalculationModel.Model> getGLModelsToUse(final RefMetaDataTracker tracker,
final ReferenceContext refContext,
final AlignmentContext rawContext) {
if ( UAC.GenotypingMode != GenotypeLikelihoodsCalculationModel.GENOTYPING_MODE.GENOTYPE_GIVEN_ALLELES )
return modelsToUse;
if ( modelsToUse.size() != 2 )
throw new IllegalStateException("GGA mode assumes that we have initialized both the SNP and indel models but found " + modelsToUse);
// if we're genotyping given alleles then we need to choose the model corresponding to the variant type requested
final List<GenotypeLikelihoodsCalculationModel.Model> GGAmodel = new ArrayList<GenotypeLikelihoodsCalculationModel.Model>(1);
final VariantContext vcInput = getVCFromAllelesRod(tracker, refContext, rawContext.getLocation(), false, logger, UAC.alleles);
if ( vcInput == null )
return GGAmodel; // no work to be done
if ( vcInput.isSNP() ) {
// use the SNP model unless the user chose INDEL mode only
if ( modelsToUse.size() == 2 || modelsToUse.get(0).name().endsWith("SNP") )
GGAmodel.add(modelsToUse.get(0));
if ( vcInput == null ) {
return Collections.emptyList(); // no work to be done
} else if ( vcInput.isSNP() ) {
return Collections.singletonList(modelsToUse.get(SNP_MODEL));
} else if ( vcInput.isIndel() || vcInput.isMixed() ) {
return Collections.singletonList(modelsToUse.get(INDEL_MODEL));
} else {
return Collections.emptyList(); // No support for other types yet
}
else if ( vcInput.isIndel() || vcInput.isMixed() ) {
// use the INDEL model unless the user chose SNP mode only
if ( modelsToUse.size() == 2 )
GGAmodel.add(modelsToUse.get(1));
else if ( modelsToUse.get(0).name().endsWith("INDEL") )
GGAmodel.add(modelsToUse.get(0));
}
// No support for other types yet
return GGAmodel;
}
/**

2
protected/java/test/org/broadinstitute/sting/gatk/walkers/genotyper/UnifiedGenotyperGeneralPloidySuite1IntegrationTest.java
View File

@ -74,7 +74,7 @@ public class UnifiedGenotyperGeneralPloidySuite1IntegrationTest extends WalkerTe
@Test(enabled = true)
public void testINDEL_GGA_Pools() {
executor.PC_LSV_Test(String.format(" -maxAltAlleles 1 -ploidy 24 -gt_mode GENOTYPE_GIVEN_ALLELES -out_mode EMIT_ALL_SITES -alleles %s", LSV_ALLELES), "LSV_INDEL_GGA", "INDEL", "3f7d763c654f1d708323f369ea4a099b");
executor.PC_LSV_Test(String.format(" -maxAltAlleles 1 -ploidy 24 -gt_mode GENOTYPE_GIVEN_ALLELES -out_mode EMIT_ALL_SITES -alleles %s", LSV_ALLELES), "LSV_INDEL_GGA", "INDEL", "ceb105e3db0f2b993e3d725b0d60b6a3");
}
@Test(enabled = true)