@Test
public void testVCFHeaderSampleRenamingSingleSampleVCF() throws Exception {
final VCFCodec codec = new VCFCodec();
codec.setRemappedSampleName("FOOSAMPLE");
final AsciiLineReaderIterator vcfIterator =
new AsciiLineReaderIterator(
new AsciiLineReader(new FileInputStream(variantTestDataRoot + "HiSeq.10000.vcf")));
final VCFHeader header = (VCFHeader) codec.readHeader(vcfIterator).getHeaderValue();
Assert.assertEquals(
header.getNGenotypeSamples(), 1, "Wrong number of samples in remapped header");
Assert.assertEquals(
header.getGenotypeSamples().get(0),
"FOOSAMPLE",
"Sample name in remapped header has incorrect value");
int recordCount = 0;
while (vcfIterator.hasNext() && recordCount < 10) {
recordCount++;
final VariantContext vcfRecord = codec.decode(vcfIterator.next());
Assert.assertEquals(
vcfRecord.getSampleNames().size(),
1,
"Wrong number of samples in vcf record after remapping");
Assert.assertEquals(
vcfRecord.getSampleNames().iterator().next(),
"FOOSAMPLE",
"Wrong sample in vcf record after remapping");
}
}
// this method is intended to reconcile uniquified sample names
// it comes into play when calling this annotation from GenotypeGVCFs with --uniquifySamples
// because founderIds
// is derived from the sampleDB, which comes from the input sample names, but vc will have
// uniquified (i.e. different)
// sample names. Without this check, the founderIds won't be found in the vc and the annotation
// won't be calculated.
protected static Set<String> validateFounderIDs(
final Set<String> founderIds, final VariantContext vc) {
Set<String> vcSamples = new HashSet<>();
Set<String> returnIDs = founderIds;
vcSamples.addAll(vc.getSampleNames());
if (!vcSamples.isEmpty()) {
if (founderIds != null) {
vcSamples.removeAll(founderIds);
if (vcSamples.equals(vc.getSampleNames())) returnIDs = vc.getSampleNames();
}
}
return returnIDs;
}
private void addVariant(final VariantContext ctx) {
if (!ctx.getChr().equals(genes.get(0).getChromosome())) return;
if (ctx.getStart() >= chromEnd) return;
if (ctx.getStart() < chromStart) return;
positions.add(ctx.getStart());
for (String sample : ctx.getSampleNames()) {
Genotype g = ctx.getGenotype(sample);
if (!g.isAvailable()) continue;
if (!g.isCalled()) continue;
if (g.isNoCall()) continue;
if (g.isNonInformative()) continue;
Set<Integer> set = sample2positions.get(sample);
if (set == null) {
set = new HashSet<Integer>();
sample2positions.put(sample, set);
}
set.add(ctx.getStart());
}
}
private void writeRecord(VariantContext vc, RefMetaDataTracker tracker, GenomeLoc loc) {
if (!wroteHeader) {
wroteHeader = true;
// setup the header fields
Set<VCFHeaderLine> hInfo = new HashSet<VCFHeaderLine>();
hInfo.addAll(GATKVCFUtils.getHeaderFields(getToolkit(), Arrays.asList(variants.getName())));
hInfo.add(VCFStandardHeaderLines.getFormatLine(VCFConstants.GENOTYPE_KEY));
allowedGenotypeFormatStrings.add(VCFConstants.GENOTYPE_KEY);
for (VCFHeaderLine field : hInfo) {
if (field instanceof VCFFormatHeaderLine) {
allowedGenotypeFormatStrings.add(((VCFFormatHeaderLine) field).getID());
}
}
samples = new LinkedHashSet<String>();
if (sampleName != null) {
samples.add(sampleName);
} else {
// try VCF first
samples =
SampleUtils.getSampleListWithVCFHeader(getToolkit(), Arrays.asList(variants.getName()));
if (samples.isEmpty()) {
List<Feature> features = tracker.getValues(variants, loc);
if (features.size() == 0)
throw new IllegalStateException(
"No rod data is present, but we just created a VariantContext");
Feature f = features.get(0);
if (f instanceof RawHapMapFeature)
samples.addAll(Arrays.asList(((RawHapMapFeature) f).getSampleIDs()));
else samples.addAll(vc.getSampleNames());
}
}
vcfwriter.writeHeader(new VCFHeader(hInfo, samples));
}
vc = GATKVariantContextUtils.purgeUnallowedGenotypeAttributes(vc, allowedGenotypeFormatStrings);
vcfwriter.add(vc);
}
