public void initialize() {
// Initialize VCF header
if (vcfWriter != null) {
Map<String, VCFHeader> header =
GATKVCFUtils.getVCFHeadersFromRodPrefix(getToolkit(), alleles.getName());
samples =
SampleUtils.getSampleList(
header, GATKVariantContextUtils.GenotypeMergeType.REQUIRE_UNIQUE);
Set<VCFHeaderLine> headerLines = VCFUtils.smartMergeHeaders(header.values(), true);
headerLines.add(new VCFHeaderLine("source", "GenotypeAndValidate"));
headerLines.add(
GATKVCFHeaderLines.getInfoLine(GATKVCFConstants.GENOTYPE_AND_VALIDATE_STATUS_KEY));
vcfWriter.writeHeader(new VCFHeader(headerLines, samples));
}
// Filling in SNP calling arguments for UG
UnifiedArgumentCollection uac = new UnifiedArgumentCollection();
uac.outputMode = OutputMode.EMIT_ALL_SITES;
uac.alleles = alleles;
// TODO -- if we change this tool to actually validate against the called allele, then this if
// statement is needed;
// TODO -- for now, though, we need to be able to validate the right allele (because we only
// test isVariant below) [EB]
// if (!bamIsTruth)
uac.genotypingOutputMode = GenotypingOutputMode.GENOTYPE_GIVEN_ALLELES;
if (mbq >= 0) uac.MIN_BASE_QUALTY_SCORE = mbq;
if (deletions >= 0) uac.MAX_DELETION_FRACTION = deletions;
else uac.MAX_DELETION_FRACTION = 1.0;
if (emitConf >= 0) uac.genotypeArgs.STANDARD_CONFIDENCE_FOR_EMITTING = emitConf;
if (callConf >= 0) uac.genotypeArgs.STANDARD_CONFIDENCE_FOR_CALLING = callConf;
final GenomeAnalysisEngine toolkit = getToolkit();
uac.GLmodel = GenotypeLikelihoodsCalculationModel.Model.SNP;
snpEngine =
new UnifiedGenotypingEngine(
uac, FixedAFCalculatorProvider.createThreadSafeProvider(toolkit, uac, logger), toolkit);
// Adding the INDEL calling arguments for UG
UnifiedArgumentCollection uac_indel = uac.clone();
uac_indel.GLmodel = GenotypeLikelihoodsCalculationModel.Model.INDEL;
indelEngine =
new UnifiedGenotypingEngine(
uac_indel,
FixedAFCalculatorProvider.createThreadSafeProvider(toolkit, uac, logger),
toolkit);
// make sure we have callConf set to the threshold set by the UAC so we can use it later.
callConf = uac.genotypeArgs.STANDARD_CONFIDENCE_FOR_CALLING;
}
/**
* Gets the header lines for the VCF writer
*
* @return A set of VCF header lines
*/
private static Set<VCFHeaderLine> getHeaderInfo() {
Set<VCFHeaderLine> headerLines = new HashSet<>();
// INFO fields for overall data
headerLines.add(VCFStandardHeaderLines.getInfoLine(VCFConstants.END_KEY));
headerLines.add(GATKVCFHeaderLines.getInfoLine(GATKVCFConstants.AVG_INTERVAL_DP_KEY));
headerLines.add(GATKVCFHeaderLines.getInfoLine(GATKVCFConstants.INTERVAL_GC_CONTENT_KEY));
headerLines.add(
new VCFInfoHeaderLine(
"Diagnose Targets", 0, VCFHeaderLineType.Flag, "DiagnoseTargets mode"));
// FORMAT fields for each genotype
headerLines.add(VCFStandardHeaderLines.getFormatLine(VCFConstants.GENOTYPE_FILTER_KEY));
headerLines.add(
GATKVCFHeaderLines.getFormatLine(GATKVCFConstants.AVG_INTERVAL_DP_BY_SAMPLE_KEY));
headerLines.add(GATKVCFHeaderLines.getFormatLine(GATKVCFConstants.LOW_COVERAGE_LOCI));
headerLines.add(GATKVCFHeaderLines.getFormatLine(GATKVCFConstants.ZERO_COVERAGE_LOCI));
// FILTER fields
for (CallableStatus stat : CallableStatus.values())
headerLines.add(new VCFFilterHeaderLine(stat.name(), stat.description));
return headerLines;
}
@Override
public List<VCFInfoHeaderLine> getDescriptions() {
return Arrays.asList(GATKVCFHeaderLines.getInfoLine(getKeyNames().get(0)));
}