@Override
protected Object doWork() {
IOUtil.assertFileIsReadable(INPUT);
IOUtil.assertFileIsReadable(REFERENCE_SEQUENCE);
IOUtil.assertFileIsReadable(CHAIN);
IOUtil.assertFileIsWritable(OUTPUT);
IOUtil.assertFileIsWritable(REJECT);
////////////////////////////////////////////////////////////////////////
// Setup the inputs
////////////////////////////////////////////////////////////////////////
final LiftOver liftOver = new LiftOver(CHAIN);
final VCFFileReader in = new VCFFileReader(INPUT, false);
logger.info("Loading up the target reference genome.");
final ReferenceSequenceFileWalker walker = new ReferenceSequenceFileWalker(REFERENCE_SEQUENCE);
final Map<String, byte[]> refSeqs = new HashMap<>();
for (final SAMSequenceRecord rec : walker.getSequenceDictionary().getSequences()) {
refSeqs.put(rec.getSequenceName(), walker.get(rec.getSequenceIndex()).getBases());
}
CloserUtil.close(walker);
////////////////////////////////////////////////////////////////////////
// Setup the outputs
////////////////////////////////////////////////////////////////////////
final VCFHeader inHeader = in.getFileHeader();
final VCFHeader outHeader = new VCFHeader(inHeader);
outHeader.setSequenceDictionary(walker.getSequenceDictionary());
final VariantContextWriter out =
new VariantContextWriterBuilder()
.setOption(Options.INDEX_ON_THE_FLY)
.setOutputFile(OUTPUT)
.setReferenceDictionary(walker.getSequenceDictionary())
.build();
out.writeHeader(outHeader);
final VariantContextWriter rejects =
new VariantContextWriterBuilder()
.setOutputFile(REJECT)
.unsetOption(Options.INDEX_ON_THE_FLY)
.build();
final VCFHeader rejectHeader = new VCFHeader(in.getFileHeader());
for (final VCFFilterHeaderLine line : FILTERS) rejectHeader.addMetaDataLine(line);
rejects.writeHeader(rejectHeader);
////////////////////////////////////////////////////////////////////////
// Read the input VCF, lift the records over and write to the sorting
// collection.
////////////////////////////////////////////////////////////////////////
long failedLiftover = 0, failedAlleleCheck = 0, total = 0;
logger.info("Lifting variants over and sorting.");
final SortingCollection<VariantContext> sorter =
SortingCollection.newInstance(
VariantContext.class,
new VCFRecordCodec(outHeader),
outHeader.getVCFRecordComparator(),
MAX_RECORDS_IN_RAM,
TMP_DIR);
ProgressLogger progress = new ProgressLogger(logger, 1000000, "read");
for (final VariantContext ctx : in) {
++total;
final Interval source =
new Interval(
ctx.getContig(),
ctx.getStart(),
ctx.getEnd(),
false,
ctx.getContig() + ":" + ctx.getStart() + "-" + ctx.getEnd());
final Interval target = liftOver.liftOver(source, 1.0);
if (target == null) {
rejects.add(new VariantContextBuilder(ctx).filter(FILTER_CANNOT_LIFTOVER).make());
failedLiftover++;
} else {
// Fix the alleles if we went from positive to negative strand
final List<Allele> alleles = new ArrayList<>();
for (final Allele oldAllele : ctx.getAlleles()) {
if (target.isPositiveStrand() || oldAllele.isSymbolic()) {
alleles.add(oldAllele);
} else {
alleles.add(
Allele.create(
SequenceUtil.reverseComplement(oldAllele.getBaseString()),
oldAllele.isReference()));
}
}
// Build the new variant context
final VariantContextBuilder builder =
new VariantContextBuilder(
ctx.getSource(), target.getContig(), target.getStart(), target.getEnd(), alleles);
builder.id(ctx.getID());
builder.attributes(ctx.getAttributes());
builder.genotypes(ctx.getGenotypes());
builder.filters(ctx.getFilters());
builder.log10PError(ctx.getLog10PError());
// Check that the reference allele still agrees with the reference sequence
boolean mismatchesReference = false;
for (final Allele allele : builder.getAlleles()) {
if (allele.isReference()) {
final byte[] ref = refSeqs.get(target.getContig());
final String refString =
StringUtil.bytesToString(ref, target.getStart() - 1, target.length());
if (!refString.equalsIgnoreCase(allele.getBaseString())) {
mismatchesReference = true;
}
break;
}
}
if (mismatchesReference) {
rejects.add(new VariantContextBuilder(ctx).filter(FILTER_MISMATCHING_REF_ALLELE).make());
failedAlleleCheck++;
} else {
sorter.add(builder.make());
}
}
progress.record(ctx.getContig(), ctx.getStart());
}
final NumberFormat pfmt = new DecimalFormat("0.0000%");
final String pct = pfmt.format((failedLiftover + failedAlleleCheck) / (double) total);
logger.info("Processed ", total, " variants.");
logger.info(Long.toString(failedLiftover), " variants failed to liftover.");
logger.info(
Long.toString(failedAlleleCheck),
" variants lifted over but had mismatching reference alleles after lift over.");
logger.info(pct, " of variants were not successfully lifted over and written to the output.");
rejects.close();
in.close();
////////////////////////////////////////////////////////////////////////
// Write the sorted outputs to the final output file
////////////////////////////////////////////////////////////////////////
sorter.doneAdding();
progress = new ProgressLogger(logger, 1000000, "written");
logger.info("Writing out sorted records to final VCF.");
for (final VariantContext ctx : sorter) {
out.add(ctx);
progress.record(ctx.getContig(), ctx.getStart());
}
out.close();
sorter.cleanup();
return null;
}
@Override
protected int doWork() {
IOUtil.assertFileIsReadable(INPUT);
IOUtil.assertFileIsWritable(OUTPUT);
IOUtil.assertFileIsReadable(REFERENCE_SEQUENCE);
// Setup all the inputs
final ProgressLogger progress = new ProgressLogger(log, 10000000, "Processed", "loci");
final ReferenceSequenceFileWalker refWalker =
new ReferenceSequenceFileWalker(REFERENCE_SEQUENCE);
final SamReader in =
SamReaderFactory.makeDefault().referenceSequence(REFERENCE_SEQUENCE).open(INPUT);
final SamLocusIterator iterator = getLocusIterator(in);
final List<SamRecordFilter> filters = new ArrayList<SamRecordFilter>();
final CountingFilter dupeFilter = new CountingDuplicateFilter();
final CountingFilter mapqFilter = new CountingMapQFilter(MINIMUM_MAPPING_QUALITY);
final CountingPairedFilter pairFilter = new CountingPairedFilter();
filters.add(mapqFilter);
filters.add(dupeFilter);
if (!COUNT_UNPAIRED) {
filters.add(pairFilter);
}
filters.add(
new SecondaryAlignmentFilter()); // Not a counting filter because we never want to count
// reads twice
iterator.setSamFilters(filters);
iterator.setEmitUncoveredLoci(true);
iterator.setMappingQualityScoreCutoff(0); // Handled separately because we want to count bases
iterator.setQualityScoreCutoff(0); // Handled separately because we want to count bases
iterator.setIncludeNonPfReads(false);
final int max = COVERAGE_CAP;
final long[] HistogramArray = new long[max + 1];
final long[] baseQHistogramArray = new long[Byte.MAX_VALUE];
final boolean usingStopAfter = STOP_AFTER > 0;
final long stopAfter = STOP_AFTER - 1;
long counter = 0;
long basesExcludedByBaseq = 0;
long basesExcludedByOverlap = 0;
long basesExcludedByCapping = 0;
// Loop through all the loci
while (iterator.hasNext()) {
final SamLocusIterator.LocusInfo info = iterator.next();
// Check that the reference is not N
final ReferenceSequence ref = refWalker.get(info.getSequenceIndex());
final byte base = ref.getBases()[info.getPosition() - 1];
if (base == 'N') continue;
// Figure out the coverage while not counting overlapping reads twice, and excluding various
// things
final HashSet<String> readNames = new HashSet<String>(info.getRecordAndPositions().size());
int pileupSize = 0;
for (final SamLocusIterator.RecordAndOffset recs : info.getRecordAndPositions()) {
if (recs.getBaseQuality() < MINIMUM_BASE_QUALITY) {
++basesExcludedByBaseq;
continue;
}
if (!readNames.add(recs.getRecord().getReadName())) {
++basesExcludedByOverlap;
continue;
}
pileupSize++;
if (pileupSize <= max) {
baseQHistogramArray[recs.getRecord().getBaseQualities()[recs.getOffset()]]++;
}
}
final int depth = Math.min(readNames.size(), max);
if (depth < readNames.size()) basesExcludedByCapping += readNames.size() - max;
HistogramArray[depth]++;
// Record progress and perhaps stop
progress.record(info.getSequenceName(), info.getPosition());
if (usingStopAfter && ++counter > stopAfter) break;
}
// Construct and write the outputs
final Histogram<Integer> histo = new Histogram<Integer>("coverage", "count");
for (int i = 0; i < HistogramArray.length; ++i) {
histo.increment(i, HistogramArray[i]);
}
// Construct and write the outputs
final Histogram<Integer> baseQHisto = new Histogram<Integer>("value", "baseq_count");
for (int i = 0; i < baseQHistogramArray.length; ++i) {
baseQHisto.increment(i, baseQHistogramArray[i]);
}
final WgsMetrics metrics = generateWgsMetrics();
metrics.GENOME_TERRITORY = (long) histo.getSumOfValues();
metrics.MEAN_COVERAGE = histo.getMean();
metrics.SD_COVERAGE = histo.getStandardDeviation();
metrics.MEDIAN_COVERAGE = histo.getMedian();
metrics.MAD_COVERAGE = histo.getMedianAbsoluteDeviation();
final long basesExcludedByDupes = getBasesExcludedBy(dupeFilter);
final long basesExcludedByMapq = getBasesExcludedBy(mapqFilter);
final long basesExcludedByPairing = getBasesExcludedBy(pairFilter);
final double total = histo.getSum();
final double totalWithExcludes =
total
+ basesExcludedByDupes
+ basesExcludedByMapq
+ basesExcludedByPairing
+ basesExcludedByBaseq
+ basesExcludedByOverlap
+ basesExcludedByCapping;
metrics.PCT_EXC_DUPE = basesExcludedByDupes / totalWithExcludes;
metrics.PCT_EXC_MAPQ = basesExcludedByMapq / totalWithExcludes;
metrics.PCT_EXC_UNPAIRED = basesExcludedByPairing / totalWithExcludes;
metrics.PCT_EXC_BASEQ = basesExcludedByBaseq / totalWithExcludes;
metrics.PCT_EXC_OVERLAP = basesExcludedByOverlap / totalWithExcludes;
metrics.PCT_EXC_CAPPED = basesExcludedByCapping / totalWithExcludes;
metrics.PCT_EXC_TOTAL = (totalWithExcludes - total) / totalWithExcludes;
metrics.PCT_1X =
MathUtil.sum(HistogramArray, 1, HistogramArray.length) / (double) metrics.GENOME_TERRITORY;
metrics.PCT_5X =
MathUtil.sum(HistogramArray, 5, HistogramArray.length) / (double) metrics.GENOME_TERRITORY;
metrics.PCT_10X =
MathUtil.sum(HistogramArray, 10, HistogramArray.length) / (double) metrics.GENOME_TERRITORY;
metrics.PCT_15X =
MathUtil.sum(HistogramArray, 15, HistogramArray.length) / (double) metrics.GENOME_TERRITORY;
metrics.PCT_20X =
MathUtil.sum(HistogramArray, 20, HistogramArray.length) / (double) metrics.GENOME_TERRITORY;
metrics.PCT_25X =
MathUtil.sum(HistogramArray, 25, HistogramArray.length) / (double) metrics.GENOME_TERRITORY;
metrics.PCT_30X =
MathUtil.sum(HistogramArray, 30, HistogramArray.length) / (double) metrics.GENOME_TERRITORY;
metrics.PCT_40X =
MathUtil.sum(HistogramArray, 40, HistogramArray.length) / (double) metrics.GENOME_TERRITORY;
metrics.PCT_50X =
MathUtil.sum(HistogramArray, 50, HistogramArray.length) / (double) metrics.GENOME_TERRITORY;
metrics.PCT_60X =
MathUtil.sum(HistogramArray, 60, HistogramArray.length) / (double) metrics.GENOME_TERRITORY;
metrics.PCT_70X =
MathUtil.sum(HistogramArray, 70, HistogramArray.length) / (double) metrics.GENOME_TERRITORY;
metrics.PCT_80X =
MathUtil.sum(HistogramArray, 80, HistogramArray.length) / (double) metrics.GENOME_TERRITORY;
metrics.PCT_90X =
MathUtil.sum(HistogramArray, 90, HistogramArray.length) / (double) metrics.GENOME_TERRITORY;
metrics.PCT_100X =
MathUtil.sum(HistogramArray, 100, HistogramArray.length)
/ (double) metrics.GENOME_TERRITORY;
final MetricsFile<WgsMetrics, Integer> out = getMetricsFile();
out.addMetric(metrics);
out.addHistogram(histo);
if (INCLUDE_BQ_HISTOGRAM) {
out.addHistogram(baseQHisto);
}
out.write(OUTPUT);
return 0;
}
