private static void mergeGenotypes(
GenotypesContext mergedGenotypes,
VariantContext oneVC,
AlleleMapper alleleMapping,
boolean uniqifySamples) {
for (Genotype g : oneVC.getGenotypes()) {
String name = mergedSampleName(oneVC.getSource(), g.getSampleName(), uniqifySamples);
if (!mergedGenotypes.containsSample(name)) {
// only add if the name is new
Genotype newG = g;
if (uniqifySamples || alleleMapping.needsRemapping()) {
final List<Allele> alleles =
alleleMapping.needsRemapping() ? alleleMapping.remap(g.getAlleles()) : g.getAlleles();
newG =
new Genotype(
name,
alleles,
g.getLog10PError(),
g.getFilters(),
g.getAttributes(),
g.isPhased());
}
mergedGenotypes.add(newG);
}
}
}
public static Genotype removePLs(Genotype g) {
Map<String, Object> attrs = new HashMap<String, Object>(g.getAttributes());
attrs.remove(VCFConstants.PHRED_GENOTYPE_LIKELIHOODS_KEY);
attrs.remove(VCFConstants.GENOTYPE_LIKELIHOODS_KEY);
return new Genotype(
g.getSampleName(),
g.getAlleles(),
g.getLog10PError(),
g.filtersWereApplied() ? g.getFilters() : null,
attrs,
g.isPhased());
}
public static void assertEquals(final Genotype actual, final Genotype expected) {
Assert.assertEquals(actual.getSampleName(), expected.getSampleName(), "Genotype names");
Assert.assertEquals(actual.getAlleles(), expected.getAlleles(), "Genotype alleles");
Assert.assertEquals(
actual.getGenotypeString(), expected.getGenotypeString(), "Genotype string");
Assert.assertEquals(actual.getType(), expected.getType(), "Genotype type");
// filters are the same
Assert.assertEquals(actual.getFilters(), expected.getFilters(), "Genotype fields");
Assert.assertEquals(actual.isFiltered(), expected.isFiltered(), "Genotype isFiltered");
// inline attributes
Assert.assertEquals(actual.getDP(), expected.getDP(), "Genotype dp");
Assert.assertTrue(Arrays.equals(actual.getAD(), expected.getAD()));
Assert.assertEquals(actual.getGQ(), expected.getGQ(), "Genotype gq");
Assert.assertEquals(actual.hasPL(), expected.hasPL(), "Genotype hasPL");
Assert.assertEquals(actual.hasAD(), expected.hasAD(), "Genotype hasAD");
Assert.assertEquals(actual.hasGQ(), expected.hasGQ(), "Genotype hasGQ");
Assert.assertEquals(actual.hasDP(), expected.hasDP(), "Genotype hasDP");
Assert.assertEquals(
actual.hasLikelihoods(), expected.hasLikelihoods(), "Genotype haslikelihoods");
Assert.assertEquals(
actual.getLikelihoodsString(),
expected.getLikelihoodsString(),
"Genotype getlikelihoodsString");
Assert.assertEquals(
actual.getLikelihoods(), expected.getLikelihoods(), "Genotype getLikelihoods");
Assert.assertTrue(Arrays.equals(actual.getPL(), expected.getPL()));
Assert.assertEquals(
actual.getPhredScaledQual(), expected.getPhredScaledQual(), "Genotype phredScaledQual");
assertAttributesEquals(actual.getExtendedAttributes(), expected.getExtendedAttributes());
Assert.assertEquals(actual.isPhased(), expected.isPhased(), "Genotype isPhased");
Assert.assertEquals(actual.getPloidy(), expected.getPloidy(), "Genotype getPloidy");
}
public static VariantContextBuilder pruneVariantContext(
final VariantContextBuilder builder, Collection<String> keysToPreserve) {
final VariantContext vc = builder.make();
if (keysToPreserve == null) keysToPreserve = Collections.emptyList();
// VC info
final Map<String, Object> attributes = subsetAttributes(vc.commonInfo, keysToPreserve);
// Genotypes
final GenotypesContext genotypes = GenotypesContext.create(vc.getNSamples());
for (final Genotype g : vc.getGenotypes()) {
Map<String, Object> genotypeAttributes = subsetAttributes(g.commonInfo, keysToPreserve);
genotypes.add(
new Genotype(
g.getSampleName(),
g.getAlleles(),
g.getLog10PError(),
g.getFilters(),
genotypeAttributes,
g.isPhased()));
}
return builder.genotypes(genotypes).attributes(attributes);
}
public static VariantContext createVariantContextWithPaddedAlleles(
VariantContext inputVC, boolean refBaseShouldBeAppliedToEndOfAlleles) {
// see if we need to pad common reference base from all alleles
boolean padVC;
// We need to pad a VC with a common base if the length of the reference allele is less than the
// length of the VariantContext.
// This happens because the position of e.g. an indel is always one before the actual event (as
// per VCF convention).
long locLength = (inputVC.getEnd() - inputVC.getStart()) + 1;
if (inputVC.hasSymbolicAlleles()) padVC = true;
else if (inputVC.getReference().length() == locLength) padVC = false;
else if (inputVC.getReference().length() == locLength - 1) padVC = true;
else
throw new IllegalArgumentException(
"Badly formed variant context at location "
+ String.valueOf(inputVC.getStart())
+ " in contig "
+ inputVC.getChr()
+ ". Reference length must be at most one base shorter than location size");
// nothing to do if we don't need to pad bases
if (padVC) {
if (!inputVC.hasReferenceBaseForIndel())
throw new ReviewedStingException(
"Badly formed variant context at location "
+ inputVC.getChr()
+ ":"
+ inputVC.getStart()
+ "; no padded reference base is available.");
Byte refByte = inputVC.getReferenceBaseForIndel();
List<Allele> alleles = new ArrayList<Allele>();
for (Allele a : inputVC.getAlleles()) {
// get bases for current allele and create a new one with trimmed bases
if (a.isSymbolic()) {
alleles.add(a);
} else {
String newBases;
if (refBaseShouldBeAppliedToEndOfAlleles)
newBases = a.getBaseString() + new String(new byte[] {refByte});
else newBases = new String(new byte[] {refByte}) + a.getBaseString();
alleles.add(Allele.create(newBases, a.isReference()));
}
}
// now we can recreate new genotypes with trimmed alleles
GenotypesContext genotypes = GenotypesContext.create(inputVC.getNSamples());
for (final Genotype g : inputVC.getGenotypes()) {
List<Allele> inAlleles = g.getAlleles();
List<Allele> newGenotypeAlleles = new ArrayList<Allele>(g.getAlleles().size());
for (Allele a : inAlleles) {
if (a.isCalled()) {
if (a.isSymbolic()) {
newGenotypeAlleles.add(a);
} else {
String newBases;
if (refBaseShouldBeAppliedToEndOfAlleles)
newBases = a.getBaseString() + new String(new byte[] {refByte});
else newBases = new String(new byte[] {refByte}) + a.getBaseString();
newGenotypeAlleles.add(Allele.create(newBases, a.isReference()));
}
} else {
// add no-call allele
newGenotypeAlleles.add(Allele.NO_CALL);
}
}
genotypes.add(
new Genotype(
g.getSampleName(),
newGenotypeAlleles,
g.getLog10PError(),
g.getFilters(),
g.getAttributes(),
g.isPhased()));
}
return new VariantContextBuilder(inputVC).alleles(alleles).genotypes(genotypes).make();
} else return inputVC;
}
