protected void printVerboseData(
String pos,
VariantContext vc,
double PofF,
double phredScaledConfidence,
final GenotypeLikelihoodsCalculationModel.Model model) {
Allele refAllele = null, altAllele = null;
for (Allele allele : vc.getAlleles()) {
if (allele.isReference()) refAllele = allele;
else altAllele = allele;
}
for (int i = 0; i <= N; i++) {
StringBuilder AFline = new StringBuilder("AFINFO\t");
AFline.append(pos);
AFline.append("\t");
AFline.append(refAllele);
AFline.append("\t");
if (altAllele != null) AFline.append(altAllele);
else AFline.append("N/A");
AFline.append("\t");
AFline.append(i + "/" + N + "\t");
AFline.append(String.format("%.2f\t", ((float) i) / N));
AFline.append(String.format("%.8f\t", getAlleleFrequencyPriors(model)[i]));
verboseWriter.println(AFline.toString());
}
verboseWriter.println("P(f>0) = " + PofF);
verboseWriter.println("Qscore = " + phredScaledConfidence);
verboseWriter.println();
}
protected final void printCallInfo(
final VariantContext vc,
final double[] log10AlleleFrequencyPriors,
final long runtimeNano,
final AFCalcResult result) {
printCallElement(vc, "type", "ignore", vc.getType());
int allelei = 0;
for (final Allele a : vc.getAlleles())
printCallElement(vc, "allele", allelei++, a.getDisplayString());
for (final Genotype g : vc.getGenotypes())
printCallElement(vc, "PL", g.getSampleName(), g.getLikelihoodsString());
for (int priorI = 0; priorI < log10AlleleFrequencyPriors.length; priorI++)
printCallElement(vc, "priorI", priorI, log10AlleleFrequencyPriors[priorI]);
printCallElement(vc, "runtime.nano", "ignore", runtimeNano);
printCallElement(vc, "log10PosteriorOfAFEq0", "ignore", result.getLog10PosteriorOfAFEq0());
printCallElement(vc, "log10PosteriorOfAFGt0", "ignore", result.getLog10PosteriorOfAFGT0());
for (final Allele allele : result.getAllelesUsedInGenotyping()) {
if (allele.isNonReference()) {
printCallElement(vc, "MLE", allele, result.getAlleleCountAtMLE(allele));
printCallElement(
vc, "pNonRefByAllele", allele, result.getLog10PosteriorOfAFGt0ForAllele(allele));
}
}
callReport.flush();
}
private String getAlleleRepresentation(Allele allele) {
if (allele.isNull()) { // deletion wrt the ref
return DEL;
} else { // insertion, pass actual bases
return allele.getBaseString();
}
}
@Requires({"eval != null", "comp != null"})
private EvalCompMatchType doEvalAndCompMatch(
final VariantContext eval, final VariantContext comp, boolean requireStrictAlleleMatch) {
// find all of the matching comps
if (comp.getType() != eval.getType()) return EvalCompMatchType.NO_MATCH;
// find the comp which matches both the reference allele and alternate allele from eval
final Allele altEval =
eval.getAlternateAlleles().size() == 0 ? null : eval.getAlternateAllele(0);
final Allele altComp =
comp.getAlternateAlleles().size() == 0 ? null : comp.getAlternateAllele(0);
if ((altEval == null && altComp == null)
|| (altEval != null
&& altEval.equals(altComp)
&& eval.getReference().equals(comp.getReference()))) return EvalCompMatchType.STRICT;
else return requireStrictAlleleMatch ? EvalCompMatchType.NO_MATCH : EvalCompMatchType.LENIENT;
}
private VariantCallContext generateEmptyContext(
RefMetaDataTracker tracker,
ReferenceContext ref,
Map<String, AlignmentContext> stratifiedContexts,
AlignmentContext rawContext) {
VariantContext vc;
if (UAC.GenotypingMode
== GenotypeLikelihoodsCalculationModel.GENOTYPING_MODE.GENOTYPE_GIVEN_ALLELES) {
VariantContext vcInput =
UnifiedGenotyperEngine.getVCFromAllelesRod(
tracker, ref, rawContext.getLocation(), false, logger, UAC.alleles);
if (vcInput == null) return null;
vc =
new VariantContextBuilder(
"UG_call",
ref.getLocus().getContig(),
vcInput.getStart(),
vcInput.getEnd(),
vcInput.getAlleles())
.make();
} else {
// deal with bad/non-standard reference bases
if (!Allele.acceptableAlleleBases(new byte[] {ref.getBase()})) return null;
Set<Allele> alleles = new HashSet<Allele>();
alleles.add(Allele.create(ref.getBase(), true));
vc =
new VariantContextBuilder(
"UG_call",
ref.getLocus().getContig(),
ref.getLocus().getStart(),
ref.getLocus().getStart(),
alleles)
.make();
}
if (annotationEngine != null) {
// Note: we want to use the *unfiltered* and *unBAQed* context for the annotations
final ReadBackedPileup pileup = rawContext.getBasePileup();
stratifiedContexts = AlignmentContextUtils.splitContextBySampleName(pileup);
vc = annotationEngine.annotateContext(tracker, ref, stratifiedContexts, vc);
}
return new VariantCallContext(vc, false);
}
static boolean someSampleHasDoubleNonReferenceAllele(VariantContext vc1, VariantContext vc2) {
for (final Genotype gt1 : vc1.getGenotypes()) {
Genotype gt2 = vc2.getGenotype(gt1.getSampleName());
List<Allele> site1Alleles = gt1.getAlleles();
List<Allele> site2Alleles = gt2.getAlleles();
Iterator<Allele> all2It = site2Alleles.iterator();
for (Allele all1 : site1Alleles) {
Allele all2 = all2It.next(); // this is OK, since allSamplesAreMergeable()
if (all1.isNonReference() && all2.isNonReference()) // corresponding alleles are alternate
return true;
}
}
return false;
}
private Allele ensureMergedAllele(
Allele all1, Allele all2, boolean creatingReferenceForFirstTime) {
AlleleOneAndTwo all12 = new AlleleOneAndTwo(all1, all2);
Allele mergedAllele = mergedAlleles.get(all12);
if (mergedAllele == null) {
byte[] bases1 = all1.getBases();
byte[] bases2 = all2.getBases();
byte[] mergedBases = new byte[bases1.length + intermediateLength + bases2.length];
System.arraycopy(bases1, 0, mergedBases, 0, bases1.length);
if (intermediateBases != null)
System.arraycopy(intermediateBases, 0, mergedBases, bases1.length, intermediateLength);
System.arraycopy(bases2, 0, mergedBases, bases1.length + intermediateLength, bases2.length);
mergedAllele = Allele.create(mergedBases, creatingReferenceForFirstTime);
mergedAlleles.put(all12, mergedAllele);
}
return mergedAllele;
}
private Map<String, Object> annotateSNP(AlignmentContext stratifiedContext, VariantContext vc) {
if (!stratifiedContext.hasBasePileup()) return null;
HashMap<Byte, Integer> alleleCounts = new HashMap<Byte, Integer>();
for (Allele allele : vc.getAlternateAlleles()) alleleCounts.put(allele.getBases()[0], 0);
ReadBackedPileup pileup = stratifiedContext.getBasePileup();
int totalDepth = pileup.size();
Map<String, Object> map = new HashMap<String, Object>();
map.put(getKeyNames().get(0), totalDepth); // put total depth in right away
if (totalDepth == 0) return map; // done, can not compute FA at 0 coverage!!
int mq0 = 0; // number of "ref" reads that are acually mq0
for (PileupElement p : pileup) {
if (p.getMappingQual() == 0) {
mq0++;
continue;
}
if (alleleCounts.containsKey(p.getBase())) // non-mq0 read and it's an alt
alleleCounts.put(p.getBase(), alleleCounts.get(p.getBase()) + 1);
}
if (mq0 == totalDepth) return map; // if all reads are mq0, there is nothing left to do
// we need to add counts in the correct order
String[] fracs = new String[alleleCounts.size()];
for (int i = 0; i < vc.getAlternateAlleles().size(); i++) {
fracs[i] =
String.format(
"%.3f",
((float) alleleCounts.get(vc.getAlternateAllele(i).getBases()[0]))
/ (totalDepth - mq0));
}
map.put(getKeyNames().get(1), fracs);
return map;
}
static boolean doubleAllelesSegregatePerfectlyAmongSamples(
VariantContext vc1, VariantContext vc2) {
// Check that Alleles at vc1 and at vc2 always segregate together in all samples (including
// reference):
Map<Allele, Allele> allele1ToAllele2 = new HashMap<Allele, Allele>();
Map<Allele, Allele> allele2ToAllele1 = new HashMap<Allele, Allele>();
// Note the segregation of the alleles for the reference genome:
allele1ToAllele2.put(vc1.getReference(), vc2.getReference());
allele2ToAllele1.put(vc2.getReference(), vc1.getReference());
// Note the segregation of the alleles for each sample (and check that it is consistent with the
// reference and all previous samples).
for (final Genotype gt1 : vc1.getGenotypes()) {
Genotype gt2 = vc2.getGenotype(gt1.getSampleName());
List<Allele> site1Alleles = gt1.getAlleles();
List<Allele> site2Alleles = gt2.getAlleles();
Iterator<Allele> all2It = site2Alleles.iterator();
for (Allele all1 : site1Alleles) {
Allele all2 = all2It.next();
Allele all1To2 = allele1ToAllele2.get(all1);
if (all1To2 == null) allele1ToAllele2.put(all1, all2);
else if (!all1To2.equals(all2)) // all1 segregates with two different alleles at site 2
return false;
Allele all2To1 = allele2ToAllele1.get(all2);
if (all2To1 == null) allele2ToAllele1.put(all2, all1);
else if (!all2To1.equals(all1)) // all2 segregates with two different alleles at site 1
return false;
}
}
return true;
}
/**
* Read in a list of ExactCall objects from reader, keeping only those with starts in startsToKeep
* or all sites (if this is empty)
*
* @param reader a just-opened reader sitting at the start of the file
* @param startsToKeep a list of start position of the calls to keep, or empty if all calls should
* be kept
* @param parser a genome loc parser to create genome locs
* @return a list of ExactCall objects in reader
* @throws IOException
*/
public static List<ExactCall> readExactLog(
final BufferedReader reader, final List<Integer> startsToKeep, GenomeLocParser parser)
throws IOException {
if (reader == null) throw new IllegalArgumentException("reader cannot be null");
if (startsToKeep == null) throw new IllegalArgumentException("startsToKeep cannot be null");
if (parser == null) throw new IllegalArgumentException("GenomeLocParser cannot be null");
List<ExactCall> calls = new LinkedList<ExactCall>();
// skip the header line
reader.readLine();
// skip the first "type" line
reader.readLine();
while (true) {
final VariantContextBuilder builder = new VariantContextBuilder();
final List<Allele> alleles = new ArrayList<Allele>();
final List<Genotype> genotypes = new ArrayList<Genotype>();
final double[] posteriors = new double[2];
final double[] priors = MathUtils.normalizeFromLog10(new double[] {0.5, 0.5}, true);
final List<Integer> mle = new ArrayList<Integer>();
final Map<Allele, Double> log10pNonRefByAllele = new HashMap<Allele, Double>();
long runtimeNano = -1;
GenomeLoc currentLoc = null;
while (true) {
final String line = reader.readLine();
if (line == null) return calls;
final String[] parts = line.split("\t");
final GenomeLoc lineLoc = parser.parseGenomeLoc(parts[0]);
final String variable = parts[1];
final String key = parts[2];
final String value = parts[3];
if (currentLoc == null) currentLoc = lineLoc;
if (variable.equals("type")) {
if (startsToKeep.isEmpty() || startsToKeep.contains(currentLoc.getStart())) {
builder.alleles(alleles);
final int stop = currentLoc.getStart() + alleles.get(0).length() - 1;
builder.chr(currentLoc.getContig()).start(currentLoc.getStart()).stop(stop);
builder.genotypes(genotypes);
final int[] mleInts = ArrayUtils.toPrimitive(mle.toArray(new Integer[] {}));
final AFCalcResult result =
new AFCalcResult(mleInts, 1, alleles, posteriors, priors, log10pNonRefByAllele);
calls.add(new ExactCall(builder.make(), runtimeNano, result));
}
break;
} else if (variable.equals("allele")) {
final boolean isRef = key.equals("0");
alleles.add(Allele.create(value, isRef));
} else if (variable.equals("PL")) {
final GenotypeBuilder gb = new GenotypeBuilder(key);
gb.PL(GenotypeLikelihoods.fromPLField(value).getAsPLs());
genotypes.add(gb.make());
} else if (variable.equals("log10PosteriorOfAFEq0")) {
posteriors[0] = Double.valueOf(value);
} else if (variable.equals("log10PosteriorOfAFGt0")) {
posteriors[1] = Double.valueOf(value);
} else if (variable.equals("MLE")) {
mle.add(Integer.valueOf(value));
} else if (variable.equals("pNonRefByAllele")) {
final Allele a = Allele.create(key);
log10pNonRefByAllele.put(a, Double.valueOf(value));
} else if (variable.equals("runtime.nano")) {
runtimeNano = Long.valueOf(value);
} else {
// nothing to do
}
}
}
}
public void writeBeagleOutput(
VariantContext preferredVC, VariantContext otherVC, boolean isValidationSite, double prior) {
GenomeLoc currentLoc =
VariantContextUtils.getLocation(getToolkit().getGenomeLocParser(), preferredVC);
StringBuffer beagleOut = new StringBuffer();
String marker = String.format("%s:%d ", currentLoc.getContig(), currentLoc.getStart());
beagleOut.append(marker);
if (markers != null)
markers.append(marker).append("\t").append(Integer.toString(markerCounter++)).append("\t");
for (Allele allele : preferredVC.getAlleles()) {
String bglPrintString;
if (allele.isNoCall() || allele.isNull()) bglPrintString = "-";
else bglPrintString = allele.getBaseString(); // get rid of * in case of reference allele
beagleOut.append(String.format("%s ", bglPrintString));
if (markers != null) markers.append(bglPrintString).append("\t");
}
if (markers != null) markers.append("\n");
GenotypesContext preferredGenotypes = preferredVC.getGenotypes();
GenotypesContext otherGenotypes = goodSite(otherVC) ? otherVC.getGenotypes() : null;
for (String sample : samples) {
boolean isMaleOnChrX = CHECK_IS_MALE_ON_CHR_X && getSample(sample).getGender() == Gender.MALE;
Genotype genotype;
boolean isValidation;
// use sample as key into genotypes structure
if (preferredGenotypes.containsSample(sample)) {
genotype = preferredGenotypes.get(sample);
isValidation = isValidationSite;
} else if (otherGenotypes != null && otherGenotypes.containsSample(sample)) {
genotype = otherGenotypes.get(sample);
isValidation = !isValidationSite;
} else {
// there is magically no genotype for this sample.
throw new StingException(
"Sample "
+ sample
+ " arose with no genotype in variant or validation VCF. This should never happen.");
}
/*
* Use likelihoods if: is validation, prior is negative; or: is not validation, has genotype key
*/
double[] log10Likelihoods = null;
if ((isValidation && prior < 0.0) || genotype.hasLikelihoods()) {
log10Likelihoods = genotype.getLikelihoods().getAsVector();
// see if we need to randomly mask out genotype in this position.
if (GenomeAnalysisEngine.getRandomGenerator().nextDouble() <= insertedNoCallRate) {
// we are masking out this genotype
log10Likelihoods =
isMaleOnChrX ? HAPLOID_FLAT_LOG10_LIKELIHOODS : DIPLOID_FLAT_LOG10_LIKELIHOODS;
}
if (isMaleOnChrX) {
log10Likelihoods[1] = -255; // todo -- warning this is dangerous for multi-allele case
}
}
/** otherwise, use the prior uniformly */
else if (!isValidation && genotype.isCalled() && !genotype.hasLikelihoods()) {
// hack to deal with input VCFs with no genotype likelihoods. Just assume the called
// genotype
// is confident. This is useful for Hapmap and 1KG release VCFs.
double AA = (1.0 - prior) / 2.0;
double AB = (1.0 - prior) / 2.0;
double BB = (1.0 - prior) / 2.0;
if (genotype.isHomRef()) {
AA = prior;
} else if (genotype.isHet()) {
AB = prior;
} else if (genotype.isHomVar()) {
BB = prior;
}
log10Likelihoods = MathUtils.toLog10(new double[] {AA, isMaleOnChrX ? 0.0 : AB, BB});
} else {
log10Likelihoods =
isMaleOnChrX ? HAPLOID_FLAT_LOG10_LIKELIHOODS : DIPLOID_FLAT_LOG10_LIKELIHOODS;
}
writeSampleLikelihoods(beagleOut, preferredVC, log10Likelihoods);
}
beagleWriter.println(beagleOut.toString());
}
/**
* add a record to the file
*
* @param vc the Variant Context object
* @param refBase the ref base used for indels
* @param refBaseShouldBeAppliedToEndOfAlleles *** THIS SHOULD BE FALSE EXCEPT FOR AN INDEL AT THE
* EXTREME BEGINNING OF A CONTIG (WHERE THERE IS NO PREVIOUS BASE, SO WE USE THE BASE AFTER
* THE EVENT INSTEAD)
*/
public void add(VariantContext vc, byte refBase, boolean refBaseShouldBeAppliedToEndOfAlleles) {
if (mHeader == null)
throw new IllegalStateException(
"The VCF Header must be written before records can be added: " + locationString());
if (doNotWriteGenotypes) vc = VariantContext.modifyGenotypes(vc, null);
try {
vc =
VariantContext.createVariantContextWithPaddedAlleles(
vc, refBase, refBaseShouldBeAppliedToEndOfAlleles);
// if we are doing on the fly indexing, add the record ***before*** we write any bytes
if (indexer != null) indexer.addFeature(vc, positionalStream.getPosition());
Map<Allele, String> alleleMap = new HashMap<Allele, String>(vc.getAlleles().size());
alleleMap.put(Allele.NO_CALL, VCFConstants.EMPTY_ALLELE); // convenience for lookup
// CHROM
mWriter.write(vc.getChr());
mWriter.write(VCFConstants.FIELD_SEPARATOR);
// POS
mWriter.write(String.valueOf(vc.getStart()));
mWriter.write(VCFConstants.FIELD_SEPARATOR);
// ID
String ID = vc.hasID() ? vc.getID() : VCFConstants.EMPTY_ID_FIELD;
mWriter.write(ID);
mWriter.write(VCFConstants.FIELD_SEPARATOR);
// REF
alleleMap.put(vc.getReference(), "0");
String refString = vc.getReference().getDisplayString();
mWriter.write(refString);
mWriter.write(VCFConstants.FIELD_SEPARATOR);
// ALT
if (vc.isVariant()) {
Allele altAllele = vc.getAlternateAllele(0);
alleleMap.put(altAllele, "1");
String alt = altAllele.getDisplayString();
mWriter.write(alt);
for (int i = 1; i < vc.getAlternateAlleles().size(); i++) {
altAllele = vc.getAlternateAllele(i);
alleleMap.put(altAllele, String.valueOf(i + 1));
alt = altAllele.getDisplayString();
mWriter.write(",");
mWriter.write(alt);
}
} else {
mWriter.write(VCFConstants.EMPTY_ALTERNATE_ALLELE_FIELD);
}
mWriter.write(VCFConstants.FIELD_SEPARATOR);
// QUAL
if (!vc.hasNegLog10PError()) mWriter.write(VCFConstants.MISSING_VALUE_v4);
else mWriter.write(getQualValue(vc.getPhredScaledQual()));
mWriter.write(VCFConstants.FIELD_SEPARATOR);
// FILTER
String filters =
vc.isFiltered()
? ParsingUtils.join(";", ParsingUtils.sortList(vc.getFilters()))
: (filtersWereAppliedToContext || vc.filtersWereApplied()
? VCFConstants.PASSES_FILTERS_v4
: VCFConstants.UNFILTERED);
mWriter.write(filters);
mWriter.write(VCFConstants.FIELD_SEPARATOR);
// INFO
Map<String, String> infoFields = new TreeMap<String, String>();
for (Map.Entry<String, Object> field : vc.getAttributes().entrySet()) {
String key = field.getKey();
if (key.equals(VariantContext.ID_KEY)
|| key.equals(VariantContext.REFERENCE_BASE_FOR_INDEL_KEY)
|| key.equals(VariantContext.UNPARSED_GENOTYPE_MAP_KEY)
|| key.equals(VariantContext.UNPARSED_GENOTYPE_PARSER_KEY)) continue;
String outputValue = formatVCFField(field.getValue());
if (outputValue != null) infoFields.put(key, outputValue);
}
writeInfoString(infoFields);
// FORMAT
if (vc.hasAttribute(VariantContext.UNPARSED_GENOTYPE_MAP_KEY)) {
mWriter.write(VCFConstants.FIELD_SEPARATOR);
mWriter.write(vc.getAttributeAsString(VariantContext.UNPARSED_GENOTYPE_MAP_KEY, ""));
} else {
List<String> genotypeAttributeKeys = new ArrayList<String>();
if (vc.hasGenotypes()) {
genotypeAttributeKeys.addAll(calcVCFGenotypeKeys(vc));
} else if (mHeader.hasGenotypingData()) {
// this needs to be done in case all samples are no-calls
genotypeAttributeKeys.add(VCFConstants.GENOTYPE_KEY);
}
if (genotypeAttributeKeys.size() > 0) {
String genotypeFormatString =
ParsingUtils.join(VCFConstants.GENOTYPE_FIELD_SEPARATOR, genotypeAttributeKeys);
mWriter.write(VCFConstants.FIELD_SEPARATOR);
mWriter.write(genotypeFormatString);
addGenotypeData(vc, alleleMap, genotypeAttributeKeys);
}
}
mWriter.write("\n");
mWriter.flush(); // necessary so that writing to an output stream will work
} catch (IOException e) {
throw new RuntimeException("Unable to write the VCF object to " + locationString());
}
}
/**
* Main entry function to calculate genotypes of a given VC with corresponding GL's
*
* @param tracker Tracker
* @param refContext Reference context
* @param rawContext Raw context
* @param stratifiedContexts Stratified alignment contexts
* @param vc Input VC
* @param model GL calculation model
* @param inheritAttributesFromInputVC Output VC will contain attributes inherited from input vc
* @return VC with assigned genotypes
*/
public VariantCallContext calculateGenotypes(
final RefMetaDataTracker tracker,
final ReferenceContext refContext,
final AlignmentContext rawContext,
Map<String, AlignmentContext> stratifiedContexts,
final VariantContext vc,
final GenotypeLikelihoodsCalculationModel.Model model,
final boolean inheritAttributesFromInputVC,
final Map<String, org.broadinstitute.sting.utils.genotyper.PerReadAlleleLikelihoodMap>
perReadAlleleLikelihoodMap) {
boolean limitedContext =
tracker == null || refContext == null || rawContext == null || stratifiedContexts == null;
// initialize the data for this thread if that hasn't been done yet
if (afcm.get() == null) {
afcm.set(AFCalcFactory.createAFCalc(UAC, N, logger));
}
// estimate our confidence in a reference call and return
if (vc.getNSamples() == 0) {
if (limitedContext) return null;
return (UAC.OutputMode != OUTPUT_MODE.EMIT_ALL_SITES
? estimateReferenceConfidence(vc, stratifiedContexts, getTheta(model), false, 1.0)
: generateEmptyContext(tracker, refContext, stratifiedContexts, rawContext));
}
AFCalcResult AFresult = afcm.get().getLog10PNonRef(vc, getAlleleFrequencyPriors(model));
// is the most likely frequency conformation AC=0 for all alternate alleles?
boolean bestGuessIsRef = true;
// determine which alternate alleles have AF>0
final List<Allele> myAlleles = new ArrayList<Allele>(vc.getAlleles().size());
final List<Integer> alleleCountsofMLE = new ArrayList<Integer>(vc.getAlleles().size());
myAlleles.add(vc.getReference());
for (int i = 0; i < AFresult.getAllelesUsedInGenotyping().size(); i++) {
final Allele alternateAllele = AFresult.getAllelesUsedInGenotyping().get(i);
if (alternateAllele.isReference()) continue;
// we are non-ref if the probability of being non-ref > the emit confidence.
// the emit confidence is phred-scaled, say 30 => 10^-3.
// the posterior AF > 0 is log10: -5 => 10^-5
// we are non-ref if 10^-5 < 10^-3 => -5 < -3
final boolean isNonRef =
AFresult.isPolymorphic(alternateAllele, UAC.STANDARD_CONFIDENCE_FOR_EMITTING / -10.0);
// if the most likely AC is not 0, then this is a good alternate allele to use
if (isNonRef) {
myAlleles.add(alternateAllele);
alleleCountsofMLE.add(AFresult.getAlleleCountAtMLE(alternateAllele));
bestGuessIsRef = false;
}
// if in GENOTYPE_GIVEN_ALLELES mode, we still want to allow the use of a poor allele
else if (UAC.GenotypingMode
== GenotypeLikelihoodsCalculationModel.GENOTYPING_MODE.GENOTYPE_GIVEN_ALLELES) {
myAlleles.add(alternateAllele);
alleleCountsofMLE.add(AFresult.getAlleleCountAtMLE(alternateAllele));
}
}
final double PoFGT0 = Math.pow(10, AFresult.getLog10PosteriorOfAFGT0());
// note the math.abs is necessary because -10 * 0.0 => -0.0 which isn't nice
final double phredScaledConfidence =
Math.abs(
!bestGuessIsRef
|| UAC.GenotypingMode
== GenotypeLikelihoodsCalculationModel.GENOTYPING_MODE
.GENOTYPE_GIVEN_ALLELES
? -10 * AFresult.getLog10PosteriorOfAFEq0()
: -10 * AFresult.getLog10PosteriorOfAFGT0());
// return a null call if we don't pass the confidence cutoff or the most likely allele frequency
// is zero
if (UAC.OutputMode != OUTPUT_MODE.EMIT_ALL_SITES
&& !passesEmitThreshold(phredScaledConfidence, bestGuessIsRef)) {
// technically, at this point our confidence in a reference call isn't accurately estimated
// because it didn't take into account samples with no data, so let's get a better estimate
return limitedContext
? null
: estimateReferenceConfidence(vc, stratifiedContexts, getTheta(model), true, PoFGT0);
}
// start constructing the resulting VC
final GenomeLoc loc = genomeLocParser.createGenomeLoc(vc);
final VariantContextBuilder builder =
new VariantContextBuilder(
"UG_call", loc.getContig(), loc.getStart(), loc.getStop(), myAlleles);
builder.log10PError(phredScaledConfidence / -10.0);
if (!passesCallThreshold(phredScaledConfidence)) builder.filters(filter);
// create the genotypes
final GenotypesContext genotypes = afcm.get().subsetAlleles(vc, myAlleles, true, ploidy);
builder.genotypes(genotypes);
// print out stats if we have a writer
if (verboseWriter != null && !limitedContext)
printVerboseData(refContext.getLocus().toString(), vc, PoFGT0, phredScaledConfidence, model);
// *** note that calculating strand bias involves overwriting data structures, so we do that
// last
final HashMap<String, Object> attributes = new HashMap<String, Object>();
// inherit attributed from input vc if requested
if (inheritAttributesFromInputVC) attributes.putAll(vc.getAttributes());
// if the site was downsampled, record that fact
if (!limitedContext && rawContext.hasPileupBeenDownsampled())
attributes.put(VCFConstants.DOWNSAMPLED_KEY, true);
if (UAC.ANNOTATE_NUMBER_OF_ALLELES_DISCOVERED)
attributes.put(NUMBER_OF_DISCOVERED_ALLELES_KEY, vc.getAlternateAlleles().size());
// add the MLE AC and AF annotations
if (alleleCountsofMLE.size() > 0) {
attributes.put(VCFConstants.MLE_ALLELE_COUNT_KEY, alleleCountsofMLE);
final int AN = builder.make().getCalledChrCount();
final ArrayList<Double> MLEfrequencies = new ArrayList<Double>(alleleCountsofMLE.size());
// the MLEAC is allowed to be larger than the AN (e.g. in the case of all PLs being 0, the GT
// is ./. but the exact model may arbitrarily choose an AC>1)
for (int AC : alleleCountsofMLE) MLEfrequencies.add(Math.min(1.0, (double) AC / (double) AN));
attributes.put(VCFConstants.MLE_ALLELE_FREQUENCY_KEY, MLEfrequencies);
}
if (UAC.COMPUTE_SLOD && !limitedContext && !bestGuessIsRef) {
// final boolean DEBUG_SLOD = false;
// the overall lod
// double overallLog10PofNull = AFresult.log10AlleleFrequencyPosteriors[0];
double overallLog10PofF = AFresult.getLog10LikelihoodOfAFGT0();
// if ( DEBUG_SLOD ) System.out.println("overallLog10PofF=" + overallLog10PofF);
List<Allele> allAllelesToUse = builder.make().getAlleles();
// the forward lod
VariantContext vcForward =
calculateLikelihoods(
tracker,
refContext,
stratifiedContexts,
AlignmentContextUtils.ReadOrientation.FORWARD,
allAllelesToUse,
false,
model,
perReadAlleleLikelihoodMap);
AFresult = afcm.get().getLog10PNonRef(vcForward, getAlleleFrequencyPriors(model));
// double[] normalizedLog10Posteriors =
// MathUtils.normalizeFromLog10(AFresult.log10AlleleFrequencyPosteriors, true);
double forwardLog10PofNull = AFresult.getLog10LikelihoodOfAFEq0();
double forwardLog10PofF = AFresult.getLog10LikelihoodOfAFGT0();
// if ( DEBUG_SLOD ) System.out.println("forwardLog10PofNull=" + forwardLog10PofNull + ",
// forwardLog10PofF=" + forwardLog10PofF);
// the reverse lod
VariantContext vcReverse =
calculateLikelihoods(
tracker,
refContext,
stratifiedContexts,
AlignmentContextUtils.ReadOrientation.REVERSE,
allAllelesToUse,
false,
model,
perReadAlleleLikelihoodMap);
AFresult = afcm.get().getLog10PNonRef(vcReverse, getAlleleFrequencyPriors(model));
// normalizedLog10Posteriors =
// MathUtils.normalizeFromLog10(AFresult.log10AlleleFrequencyPosteriors, true);
double reverseLog10PofNull = AFresult.getLog10LikelihoodOfAFEq0();
double reverseLog10PofF = AFresult.getLog10LikelihoodOfAFGT0();
// if ( DEBUG_SLOD ) System.out.println("reverseLog10PofNull=" + reverseLog10PofNull + ",
// reverseLog10PofF=" + reverseLog10PofF);
double forwardLod = forwardLog10PofF + reverseLog10PofNull - overallLog10PofF;
double reverseLod = reverseLog10PofF + forwardLog10PofNull - overallLog10PofF;
// if ( DEBUG_SLOD ) System.out.println("forward lod=" + forwardLod + ", reverse lod=" +
// reverseLod);
// strand score is max bias between forward and reverse strands
double strandScore = Math.max(forwardLod, reverseLod);
// rescale by a factor of 10
strandScore *= 10.0;
// logger.debug(String.format("SLOD=%f", strandScore));
if (!Double.isNaN(strandScore)) attributes.put("SB", strandScore);
}
// finish constructing the resulting VC
builder.attributes(attributes);
VariantContext vcCall = builder.make();
// if we are subsetting alleles (either because there were too many or because some were not
// polymorphic)
// then we may need to trim the alleles (because the original VariantContext may have had to pad
// at the end).
if (myAlleles.size() != vc.getAlleles().size()
&& !limitedContext) // limitedContext callers need to handle allele trimming on their own to
// keep their perReadAlleleLikelihoodMap alleles in sync
vcCall = VariantContextUtils.reverseTrimAlleles(vcCall);
if (annotationEngine != null
&& !limitedContext) { // limitedContext callers need to handle annotations on their own by
// calling their own annotationEngine
// Note: we want to use the *unfiltered* and *unBAQed* context for the annotations
final ReadBackedPileup pileup = rawContext.getBasePileup();
stratifiedContexts = AlignmentContextUtils.splitContextBySampleName(pileup);
vcCall =
annotationEngine.annotateContext(
tracker, refContext, stratifiedContexts, vcCall, perReadAlleleLikelihoodMap);
}
return new VariantCallContext(vcCall, confidentlyCalled(phredScaledConfidence, PoFGT0));
}
public Allele getLikelihoods(
RefMetaDataTracker tracker,
ReferenceContext ref,
Map<String, AlignmentContext> contexts,
AlignmentContextUtils.ReadOrientation contextType,
GenotypePriors priors,
Map<String, MultiallelicGenotypeLikelihoods> GLs,
Allele alternateAlleleToUse,
boolean useBAQedPileup) {
if (tracker == null) return null;
GenomeLoc loc = ref.getLocus();
Allele refAllele, altAllele;
VariantContext vc = null;
if (!ref.getLocus().equals(lastSiteVisited)) {
// starting a new site: clear allele list
alleleList.clear();
lastSiteVisited = ref.getLocus();
indelLikelihoodMap.set(new HashMap<PileupElement, LinkedHashMap<Allele, Double>>());
haplotypeMap.clear();
if (getAlleleListFromVCF) {
for (final VariantContext vc_input : tracker.getValues(UAC.alleles, loc)) {
if (vc_input != null
&& allowableTypes.contains(vc_input.getType())
&& ref.getLocus().getStart() == vc_input.getStart()) {
vc = vc_input;
break;
}
}
// ignore places where we don't have a variant
if (vc == null) return null;
alleleList.clear();
if (ignoreSNPAllelesWhenGenotypingIndels) {
// if there's an allele that has same length as the reference (i.e. a SNP or MNP), ignore
// it and don't genotype it
for (Allele a : vc.getAlleles())
if (a.isNonReference() && a.getBases().length == vc.getReference().getBases().length)
continue;
else alleleList.add(a);
} else {
for (Allele a : vc.getAlleles()) alleleList.add(a);
}
} else {
alleleList = computeConsensusAlleles(ref, contexts, contextType);
if (alleleList.isEmpty()) return null;
}
}
// protect against having an indel too close to the edge of a contig
if (loc.getStart() <= HAPLOTYPE_SIZE) return null;
// check if there is enough reference window to create haplotypes (can be an issue at end of
// contigs)
if (ref.getWindow().getStop() < loc.getStop() + HAPLOTYPE_SIZE) return null;
if (!(priors instanceof DiploidIndelGenotypePriors))
throw new StingException(
"Only diploid-based Indel priors are supported in the DINDEL GL model");
if (alleleList.isEmpty()) return null;
refAllele = alleleList.get(0);
altAllele = alleleList.get(1);
// look for alt allele that has biggest length distance to ref allele
int maxLenDiff = 0;
for (Allele a : alleleList) {
if (a.isNonReference()) {
int lenDiff = Math.abs(a.getBaseString().length() - refAllele.getBaseString().length());
if (lenDiff > maxLenDiff) {
maxLenDiff = lenDiff;
altAllele = a;
}
}
}
final int eventLength = altAllele.getBaseString().length() - refAllele.getBaseString().length();
final int hsize = (int) ref.getWindow().size() - Math.abs(eventLength) - 1;
final int numPrefBases = ref.getLocus().getStart() - ref.getWindow().getStart() + 1;
haplotypeMap =
Haplotype.makeHaplotypeListFromAlleles(
alleleList, loc.getStart(), ref, hsize, numPrefBases);
// For each sample, get genotype likelihoods based on pileup
// compute prior likelihoods on haplotypes, and initialize haplotype likelihood matrix with
// them.
// initialize the GenotypeLikelihoods
GLs.clear();
for (Map.Entry<String, AlignmentContext> sample : contexts.entrySet()) {
AlignmentContext context = AlignmentContextUtils.stratify(sample.getValue(), contextType);
ReadBackedPileup pileup = null;
if (context.hasExtendedEventPileup()) pileup = context.getExtendedEventPileup();
else if (context.hasBasePileup()) pileup = context.getBasePileup();
if (pileup != null) {
final double[] genotypeLikelihoods =
pairModel.computeReadHaplotypeLikelihoods(
pileup, haplotypeMap, ref, eventLength, getIndelLikelihoodMap());
GLs.put(
sample.getKey(),
new MultiallelicGenotypeLikelihoods(
sample.getKey(), alleleList, genotypeLikelihoods, getFilteredDepth(pileup)));
if (DEBUG) {
System.out.format("Sample:%s Alleles:%s GL:", sample.getKey(), alleleList.toString());
for (int k = 0; k < genotypeLikelihoods.length; k++)
System.out.format("%1.4f ", genotypeLikelihoods[k]);
System.out.println();
}
}
}
return refAllele;
}
public int hashCode() {
return all1.hashCode() + all2.hashCode();
}
private ArrayList<Allele> computeConsensusAlleles(
ReferenceContext ref,
Map<String, AlignmentContext> contexts,
AlignmentContextUtils.ReadOrientation contextType) {
Allele refAllele = null, altAllele = null;
GenomeLoc loc = ref.getLocus();
ArrayList<Allele> aList = new ArrayList<Allele>();
HashMap<String, Integer> consensusIndelStrings = new HashMap<String, Integer>();
int insCount = 0, delCount = 0;
// quick check of total number of indels in pileup
for (Map.Entry<String, AlignmentContext> sample : contexts.entrySet()) {
AlignmentContext context = AlignmentContextUtils.stratify(sample.getValue(), contextType);
final ReadBackedExtendedEventPileup indelPileup = context.getExtendedEventPileup();
insCount += indelPileup.getNumberOfInsertions();
delCount += indelPileup.getNumberOfDeletions();
}
if (insCount < minIndelCountForGenotyping && delCount < minIndelCountForGenotyping)
return aList;
for (Map.Entry<String, AlignmentContext> sample : contexts.entrySet()) {
// todo -- warning, can be duplicating expensive partition here
AlignmentContext context = AlignmentContextUtils.stratify(sample.getValue(), contextType);
final ReadBackedExtendedEventPileup indelPileup = context.getExtendedEventPileup();
for (ExtendedEventPileupElement p : indelPileup.toExtendedIterable()) {
// SAMRecord read = p.getRead();
GATKSAMRecord read = ReadUtils.hardClipAdaptorSequence(p.getRead());
if (read == null) continue;
if (ReadUtils.is454Read(read)) {
continue;
}
/* if (DEBUG && p.isIndel()) {
System.out.format("Read: %s, cigar: %s, aln start: %d, aln end: %d, p.len:%d, Type:%s, EventBases:%s\n",
read.getReadName(),read.getCigar().toString(),read.getAlignmentStart(),read.getAlignmentEnd(),
p.getEventLength(),p.getType().toString(), p.getEventBases());
}
*/
String indelString = p.getEventBases();
if (p.isInsertion()) {
boolean foundKey = false;
if (read.getAlignmentEnd() == loc.getStart()) {
// first corner condition: a read has an insertion at the end, and we're right at the
// insertion.
// In this case, the read could have any of the inserted bases and we need to build a
// consensus
for (String s : consensusIndelStrings.keySet()) {
int cnt = consensusIndelStrings.get(s);
if (s.startsWith(indelString)) {
// case 1: current insertion is prefix of indel in hash map
consensusIndelStrings.put(s, cnt + 1);
foundKey = true;
break;
} else if (indelString.startsWith(s)) {
// case 2: indel stored in hash table is prefix of current insertion
// In this case, new bases are new key.
consensusIndelStrings.remove(s);
consensusIndelStrings.put(indelString, cnt + 1);
foundKey = true;
break;
}
}
if (!foundKey)
// none of the above: event bases not supported by previous table, so add new key
consensusIndelStrings.put(indelString, 1);
} else if (read.getAlignmentStart() == loc.getStart() + 1) {
// opposite corner condition: read will start at current locus with an insertion
for (String s : consensusIndelStrings.keySet()) {
int cnt = consensusIndelStrings.get(s);
if (s.endsWith(indelString)) {
// case 1: current insertion is suffix of indel in hash map
consensusIndelStrings.put(s, cnt + 1);
foundKey = true;
break;
} else if (indelString.endsWith(s)) {
// case 2: indel stored in hash table is suffix of current insertion
// In this case, new bases are new key.
consensusIndelStrings.remove(s);
consensusIndelStrings.put(indelString, cnt + 1);
foundKey = true;
break;
}
}
if (!foundKey)
// none of the above: event bases not supported by previous table, so add new key
consensusIndelStrings.put(indelString, 1);
} else {
// normal case: insertion somewhere in the middle of a read: add count to hash map
int cnt =
consensusIndelStrings.containsKey(indelString)
? consensusIndelStrings.get(indelString)
: 0;
consensusIndelStrings.put(indelString, cnt + 1);
}
} else if (p.isDeletion()) {
indelString = String.format("D%d", p.getEventLength());
int cnt =
consensusIndelStrings.containsKey(indelString)
? consensusIndelStrings.get(indelString)
: 0;
consensusIndelStrings.put(indelString, cnt + 1);
}
}
/* if (DEBUG) {
int icount = indelPileup.getNumberOfInsertions();
int dcount = indelPileup.getNumberOfDeletions();
if (icount + dcount > 0)
{
List<Pair<String,Integer>> eventStrings = indelPileup.getEventStringsWithCounts(ref.getBases());
System.out.format("#ins: %d, #del:%d\n", insCount, delCount);
for (int i=0 ; i < eventStrings.size() ; i++ ) {
System.out.format("%s:%d,",eventStrings.get(i).first,eventStrings.get(i).second);
// int k=0;
}
System.out.println();
}
} */
}
int maxAlleleCnt = 0;
String bestAltAllele = "";
for (String s : consensusIndelStrings.keySet()) {
int curCnt = consensusIndelStrings.get(s);
if (curCnt > maxAlleleCnt) {
maxAlleleCnt = curCnt;
bestAltAllele = s;
}
// if (DEBUG)
// System.out.format("Key:%s, number: %d\n",s,consensusIndelStrings.get(s) );
} // gdebug-
if (maxAlleleCnt < minIndelCountForGenotyping) return aList;
if (bestAltAllele.startsWith("D")) {
// get deletion length
int dLen = Integer.valueOf(bestAltAllele.substring(1));
// get ref bases of accurate deletion
int startIdxInReference = (int) (1 + loc.getStart() - ref.getWindow().getStart());
// System.out.println(new String(ref.getBases()));
byte[] refBases =
Arrays.copyOfRange(ref.getBases(), startIdxInReference, startIdxInReference + dLen);
if (Allele.acceptableAlleleBases(refBases)) {
refAllele = Allele.create(refBases, true);
altAllele = Allele.create(Allele.NULL_ALLELE_STRING, false);
}
} else {
// insertion case
if (Allele.acceptableAlleleBases(bestAltAllele)) {
refAllele = Allele.create(Allele.NULL_ALLELE_STRING, true);
altAllele = Allele.create(bestAltAllele, false);
}
}
if (refAllele != null && altAllele != null) {
aList.add(0, refAllele);
aList.add(1, altAllele);
}
return aList;
}
private Map<String, Object> annotateIndel(AlignmentContext stratifiedContext, VariantContext vc) {
if (!stratifiedContext.hasExtendedEventPileup()) {
return null;
}
ReadBackedExtendedEventPileup pileup = stratifiedContext.getExtendedEventPileup();
if (pileup == null) return null;
int totalDepth = pileup.size();
Map<String, Object> map = new HashMap<String, Object>();
map.put(getKeyNames().get(0), totalDepth); // put total depth in right away
if (totalDepth == 0) return map;
int mq0 = 0; // number of "ref" reads that are acually mq0
HashMap<String, Integer> alleleCounts = new HashMap<String, Integer>();
Allele refAllele = vc.getReference();
for (Allele allele : vc.getAlternateAlleles()) {
if (allele.isNoCall()) {
continue; // this does not look so good, should we die???
}
alleleCounts.put(getAlleleRepresentation(allele), 0);
}
for (ExtendedEventPileupElement e : pileup.toExtendedIterable()) {
if (e.getMappingQual() == 0) {
mq0++;
continue;
}
if (e.isInsertion()) {
final String b = e.getEventBases();
if (alleleCounts.containsKey(b)) {
alleleCounts.put(b, alleleCounts.get(b) + 1);
}
} else {
if (e.isDeletion()) {
if (e.getEventLength() == refAllele.length()) {
// this is indeed the deletion allele recorded in VC
final String b = DEL;
if (alleleCounts.containsKey(b)) {
alleleCounts.put(b, alleleCounts.get(b) + 1);
}
}
// else {
// System.out.print(" deletion of WRONG length found");
// }
}
}
}
if (mq0 == totalDepth) return map;
String[] fracs = new String[alleleCounts.size()];
for (int i = 0; i < vc.getAlternateAlleles().size(); i++)
fracs[i] =
String.format(
"%.3f",
((float) alleleCounts.get(getAlleleRepresentation(vc.getAlternateAllele(i))))
/ (totalDepth - mq0));
map.put(getKeyNames().get(1), fracs);
// map.put(getKeyNames().get(0), counts);
return map;
}