private void loadVCFListFile(ResourceLocator locator, List<Track> newTracks, Genome genome)
throws IOException, TribbleIndexNotFoundException {
TribbleListFeatureSource src = new TribbleListFeatureSource(locator.getPath(), genome);
VCFHeader header = (VCFHeader) src.getHeader();
// Test if the input VCF file contains methylation rate data:
// This is determined by testing for the presence of two sample format fields: MR and GB, used
// in the
// rendering of methylation rate.
// MR is the methylation rate on a scale of 0 to 100% and GB is the number of bases that pass
// filter for the position. GB is needed to avoid displaying positions for which limited
// coverage
// prevents reliable estimation of methylation rate.
boolean enableMethylationRateSupport =
(header.getFormatHeaderLine("MR") != null && header.getFormatHeaderLine("GB") != null);
List<String> allSamples = new ArrayList(header.getGenotypeSamples());
VariantTrack t = new VariantTrack(locator, src, allSamples, enableMethylationRateSupport);
// VCF tracks handle their own margin
t.setMargin(0);
newTracks.add(t);
}
@Override
protected void doWork(String inputSource, VcfIterator r, VariantContextWriter w)
throws IOException {
VCFHeader header = r.getHeader();
VCFHeader h2 = new VCFHeader(header.getMetaDataInInputOrder(), header.getSampleNamesInOrder());
h2.addMetaDataLine(
new VCFInfoHeaderLine(
TAG,
VCFHeaderLineCount.UNBOUNDED,
VCFHeaderLineType.String,
"metadata added from " + TABIX + " . Format was " + FORMAT));
h2.addMetaDataLine(
new VCFHeaderLine(
getClass().getSimpleName() + "CmdLine", String.valueOf(getProgramCommandLine())));
h2.addMetaDataLine(
new VCFHeaderLine(getClass().getSimpleName() + "Version", String.valueOf(getVersion())));
h2.addMetaDataLine(
new VCFHeaderLine(
getClass().getSimpleName() + "HtsJdkVersion", HtsjdkVersion.getVersion()));
h2.addMetaDataLine(
new VCFHeaderLine(getClass().getSimpleName() + "HtsJdkHome", HtsjdkVersion.getHome()));
SAMSequenceDictionaryProgress progress = new SAMSequenceDictionaryProgress(header);
w.writeHeader(h2);
while (r.hasNext()) {
VariantContext ctx = progress.watch(r.next());
Set<String> annotations = new HashSet<String>();
CloseableIterator<BedLine> iter =
this.bedReader.iterator(ctx.getContig(), ctx.getStart() - 1, ctx.getEnd() + 1);
while (iter.hasNext()) {
BedLine bedLine = iter.next();
if (!ctx.getContig().equals(bedLine.getContig())) continue;
if (ctx.getStart() - 1 >= bedLine.getEnd()) continue;
if (ctx.getEnd() - 1 < bedLine.getStart()) continue;
String newannot = this.parsedFormat.toString(bedLine);
if (!newannot.isEmpty()) annotations.add(VCFUtils.escapeInfoField(newannot));
}
CloserUtil.close(iter);
if (annotations.isEmpty()) {
w.add(ctx);
continue;
}
VariantContextBuilder vcb = new VariantContextBuilder(ctx);
vcb.attribute(TAG, annotations.toArray());
w.add(vcb.make());
incrVariantCount();
if (checkOutputError()) break;
}
progress.finish();
}
@Override
public void initialize(
AnnotatorCompatible walker, GenomeAnalysisEngine toolkit, Set<VCFHeaderLine> headerLines) {
// Make sure that we actually have a valid SnpEff rod binding (just in case the user specified
// -A SnpEff
// without providing a SnpEff rod via --snpEffFile):
if (!isValidRodBinding(walker.getSnpEffRodBinding())) {
canAnnotate = false;
return;
}
RodBinding<VariantContext> snpEffRodBinding = walker.getSnpEffRodBinding();
// Make sure that the SnpEff version number and command-line header lines are present in the VCF
// header of
// the SnpEff rod, and that the file was generated by a supported version of SnpEff:
VCFHeader snpEffVCFHeader =
GATKVCFUtils.getVCFHeadersFromRods(toolkit, Arrays.asList(snpEffRodBinding.getName()))
.get(snpEffRodBinding.getName());
VCFHeaderLine snpEffVersionLine =
snpEffVCFHeader.getOtherHeaderLine(SNPEFF_VCF_HEADER_VERSION_LINE_KEY);
VCFHeaderLine snpEffCommandLine =
snpEffVCFHeader.getOtherHeaderLine(SNPEFF_VCF_HEADER_COMMAND_LINE_KEY);
if (!isValidSnpEffVersionAndCommandLine(snpEffVersionLine, snpEffCommandLine)) {
canAnnotate = false;
return;
}
// If everything looks ok, add the SnpEff version number and command-line header lines to the
// header of the VCF output file, changing the key names so that our output file won't be
// mistaken in the future for a SnpEff output file:
headerLines.add(
new VCFHeaderLine(OUTPUT_VCF_HEADER_VERSION_LINE_KEY, snpEffVersionLine.getValue()));
headerLines.add(
new VCFHeaderLine(OUTPUT_VCF_HEADER_COMMAND_LINE_KEY, snpEffCommandLine.getValue()));
// Can only be called from VariantAnnotator
if (!(walker instanceof VariantAnnotator)) {
if (walker != null)
logger.warn(
"Annotation will not be calculated, must be called from VariantAnnotator, not "
+ walker.getClass().getName());
else logger.warn("Annotation will not be calculated, must be called from VariantAnnotator");
canAnnotate = false;
return;
}
}
@Override
public void execute() {
VariantContextWriter vcw =
new VariantContextWriterBuilder()
.setOutputFile(out)
.clearIndexCreator()
.setReferenceDictionary(VCF.getFileHeader().getSequenceDictionary())
.build();
VCFHeader vh = VCF.getFileHeader();
vh.addMetaDataLine(
new VCFInfoHeaderLine(
"region", 1, VCFHeaderLineType.String, "region type in which the variant occurs"));
vcw.writeHeader(VCF.getFileHeader());
IntervalTreeMap<String> itm = loadRegions();
int num = 0;
for (VariantContext vc : VCF) {
if (!vc.isFiltered()
&& vc.getGenotype(CHILD).isCalled()
&& vc.getGenotype(MOTHER).isCalled()
&& vc.getGenotype(FATHER).isCalled()
&& vc.getGenotype(CHILD).getType().equals(GenotypeType.HET)
&& vc.getGenotype(MOTHER).getType().equals(GenotypeType.HOM_REF)
&& vc.getGenotype(FATHER).getType().equals(GenotypeType.HOM_REF)
&& (!vc.getGenotype(CHILD).hasAnyAttribute("GQ")
|| (vc.getGenotype(CHILD).getGQ() > GQ_THRESHOLD
&& vc.getGenotype(MOTHER).getGQ() > GQ_THRESHOLD
&& vc.getGenotype(FATHER).getGQ() > GQ_THRESHOLD))
&& (!vc.getGenotype(CHILD).hasAnyAttribute("DP")
|| (vc.getGenotype(CHILD).getDP() > DP_THRESHOLD
&& vc.getGenotype(MOTHER).getDP() > DP_THRESHOLD
&& vc.getGenotype(FATHER).getDP() > DP_THRESHOLD))) {
vcw.add(vc);
num++;
}
}
vcw.close();
log.info("num: {}", num);
}
public VariantContextTestData(final VCFHeader header, final List<VariantContext> vcs) {
final Set<String> samples = new HashSet<String>();
for (final VariantContext vc : vcs)
if (vc.hasGenotypes()) samples.addAll(vc.getSampleNames());
this.header =
samples.isEmpty() ? header : new VCFHeader(header.getMetaDataInSortedOrder(), samples);
this.vcs = vcs;
}
public static void assertEquals(final VCFHeader actual, final VCFHeader expected) {
Assert.assertEquals(
actual.getMetaDataInSortedOrder().size(),
expected.getMetaDataInSortedOrder().size(),
"No VCF header lines");
// for some reason set.equals() is returning false but all paired elements are .equals().
// Perhaps compare to is busted?
// Assert.assertEquals(actual.getMetaDataInInputOrder(), expected.getMetaDataInInputOrder());
final List<VCFHeaderLine> actualLines =
new ArrayList<VCFHeaderLine>(actual.getMetaDataInSortedOrder());
final List<VCFHeaderLine> expectedLines =
new ArrayList<VCFHeaderLine>(expected.getMetaDataInSortedOrder());
for (int i = 0; i < actualLines.size(); i++) {
Assert.assertEquals(actualLines.get(i), expectedLines.get(i), "VCF header lines");
}
}
@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 void doWork(String source, VcfIterator in, VariantContextWriter out)
throws IOException {
try {
VCFHeader header = in.getHeader();
VCFHeader h2 = new VCFHeader(header);
h2.addMetaDataLine(
new VCFHeaderLine(
getClass().getSimpleName() + "CmdLine", String.valueOf(getProgramCommandLine())));
h2.addMetaDataLine(
new VCFHeaderLine(getClass().getSimpleName() + "Version", String.valueOf(getVersion())));
h2.addMetaDataLine(
new VCFHeaderLine(
getClass().getSimpleName() + "HtsJdkVersion", HtsjdkVersion.getVersion()));
h2.addMetaDataLine(
new VCFHeaderLine(getClass().getSimpleName() + "HtsJdkHome", HtsjdkVersion.getHome()));
out.writeHeader(h2);
final VepPredictionParser vepParser = new VepPredictionParser(header);
final SnpEffPredictionParser snpEffparser = new SnpEffPredictionParser(header);
final MyPredictionParser myPredParser = new MyPredictionParser(header);
SAMSequenceDictionaryProgress progress =
new SAMSequenceDictionaryProgress(header.getSequenceDictionary());
while (in.hasNext()) {
this.checkKnimeCancelled();
VariantContext ctx = progress.watch(in.next());
boolean keep = false;
for (SnpEffPredictionParser.SnpEffPrediction pred : snpEffparser.getPredictions(ctx)) {
if (hasUserTem(pred.getSOTerms())) {
keep = true;
break;
}
}
if (!keep) {
for (VepPredictionParser.VepPrediction pred : vepParser.getPredictions(ctx)) {
if (hasUserTem(pred.getSOTerms())) {
keep = true;
break;
}
}
}
if (!keep) {
for (MyPredictionParser.MyPrediction pred : myPredParser.getPredictions(ctx)) {
if (hasUserTem(pred.getSOTerms())) {
keep = true;
break;
}
}
}
if (isInverseResult()) keep = !keep;
if (keep) {
incrVariantCount();
out.add(ctx);
}
if (checkOutputError()) break;
}
progress.finish();
} finally {
}
}
@Override
public void runCommand() {
logger.info("MergeVCFColumnsCommand");
/*
* Assumptions
* (1) Only two vcfs that are sorted with the same contig order
* (2) if contigs on it same order, then we will just skip that contig
* (3) No overlapping samples allowed
*
* Output:
* A vcf where intersecting sites are merged together and will only return biallelic markers
* the info field will be cleared
* the only GT FORMAT field will be there
*/
Collection<File> vcfs = applicationOptions.getVcfs();
String outfile = applicationOptions.getOutFile();
if (vcfs.size() != 2) {
throw new IllegalArgumentException("This function requires exactly two vcfs");
}
Iterator<File> vcfFileIter = vcfs.iterator();
File vcf1 = vcfFileIter.next();
File vcf2 = vcfFileIter.next();
VCFFileReader reader1 = new VCFFileReader(vcf1, false);
VCFFileReader reader2 = new VCFFileReader(vcf2, false);
Iterator<VariantContext> iter1 = reader1.iterator();
Iterator<VariantContext> iter2 = reader2.iterator();
VariantContextComparator comparator = new VariantContextComparator();
/*
* Merge headers
*/
VCFHeader header1 = reader1.getFileHeader();
VCFHeader header2 = reader2.getFileHeader();
List<String> samples1 = header1.getGenotypeSamples();
List<String> samples2 = header2.getGenotypeSamples();
List<String> mergedSamples = new ArrayList<>(samples1.size() + samples2.size());
mergedSamples.addAll(samples1);
mergedSamples.addAll(samples2);
// Validate that there are no duplicates
HashSet<String> sampleSet = new HashSet<String>();
for (String id : mergedSamples) {
if (sampleSet.contains(id)) {
throw new IllegalArgumentException("Duplicate id found: " + id);
} else {
sampleSet.add(id);
}
}
HashSet<VCFHeaderLine> meta = new HashSet<>();
meta.add(new VCFFormatHeaderLine("GT", 1, VCFHeaderLineType.String, "GT"));
meta.addAll(header1.getContigLines());
VCFHeader mergedHeader = new VCFHeader(meta, mergedSamples);
/*
* Create encoder
*/
VCFEncoder encoder = new VCFEncoder(mergedHeader, false, false);
BufferedWriter writer = null;
try {
if (outfile.endsWith(".gz")) {
BlockCompressedOutputStream outstream = new BlockCompressedOutputStream(new File(outfile));
writer = new BufferedWriter(new OutputStreamWriter(outstream));
} else {
writer =
Files.newBufferedWriter(
Paths.get(outfile),
Charset.defaultCharset(),
StandardOpenOption.CREATE,
StandardOpenOption.TRUNCATE_EXISTING);
}
/*
* Write header
*/
VCFHeaderWriter.writeHeader(writer, mergedHeader);
logger.info("Wrote header");
VariantContext previous1 = null;
VariantContext previous2 = null;
int count = 0;
int countFile1 = 0;
int countFile2 = 0;
boolean usePrevious1 = false;
boolean usePrevious2 = false;
while (iter1.hasNext() || iter2.hasNext()) {
if ((iter1.hasNext() || usePrevious1) && (iter2.hasNext() || usePrevious2)) {
VariantContext variant1 = null;
VariantContext variant2 = null;
// if(usePrevious1 == true && usePrevious2 == true &&
// comparator.compare(previous1,previous2) != 0) {
// //then skip both
// usePrevious1 = false;
// usePrevious2 = false;
// }
if (usePrevious1) {
variant1 = previous1;
} else {
variant1 = iter1.next();
countFile1++;
}
if (usePrevious2) {
variant2 = previous2;
} else {
variant2 = iter2.next();
countFile2++;
}
// check that variants are ordered correctly
if (previous1 != null
&& previous1 != variant1
&& comparator.compare(previous1, variant1) > 0) {
throw new IllegalStateException(
previous1.getContig()
+ ":"
+ previous1.getStart()
+ " > "
+ variant1.getContig()
+ ":"
+ variant1.getStart());
}
if (previous2 != null
&& previous2 != variant2
&& comparator.compare(previous2, variant2) > 0) {
throw new IllegalStateException(
previous2.getContig()
+ ":"
+ previous2.getStart()
+ " > "
+ variant2.getContig()
+ ":"
+ variant2.getStart());
}
int cmp = comparator.compare(variant1, variant2);
if (cmp < 0) {
// logger.info("Skipping VCF1: " + variant1.getContig() + ":" + variant1.getStart() +
// "\t" + variant1.getReference().toString() + "\t" + variant1.getAlternateAlleles());
if (usePrevious1 == true && usePrevious2 == true) {
// variant1 < variant2
// we need to go to next variant in vcf1
usePrevious1 = false;
}
usePrevious2 = true;
} else if (cmp > 0) {
if (usePrevious1 == true && usePrevious2 == true) {
// variant1 > variant2
// we need to go to next variant in vcf2
usePrevious2 = false;
}
usePrevious1 = true;
// logger.info("Skipping VCF2: " + variant2.getContig() + ":" + variant2.getStart() +
// "\t" + variant2.getReference().toString() + "\t" + variant2.getAlternateAlleles());
} else {
// they equal position
usePrevious1 = false;
usePrevious2 = false;
if (variant1.isBiallelic()
&& variant2.isBiallelic()
&& variant1.getReference().equals(variant2.getReference())
&& variant1.getAlternateAllele(0).equals(variant2.getAlternateAllele(0))) {
// TODO: Finish merging
// both variants are bialleleic and the reference and alternative alleles match
count++;
if (count % 10000 == 0) {
logger.info(count + " mergeable variants found");
}
VariantContext merged = VariantContextMerger.merge(variant1, variant2);
writer.write(encoder.encode(merged));
writer.write("\n");
} else {
// skip if they do not equal
// logger.info("Skipping: " + variant1.getContig() + ":" + variant1.getStart() +
// "\t" + variant1.getReference().toString() + "\t" +
// variant1.getAlternateAlleles());
// logger.info("Skipping: " + variant2.getContig() + ":" + variant2.getStart() +
// "\t" + variant2.getReference().toString() + "\t" +
// variant2.getAlternateAlleles());
}
}
previous1 = variant1;
previous2 = variant2;
} else if (iter1.hasNext()) {
// just skip remaining variants
VariantContext current = iter1.next();
countFile1++;
if (previous1 != null && current != null && comparator.compare(previous1, current) > 0) {
throw new IllegalStateException(
previous1.getContig()
+ ":"
+ previous1.getStart()
+ " > "
+ current.getContig()
+ ":"
+ current.getStart());
}
previous1 = current;
// logger.info("Skipping: " + previous1.getContig() + ":" + previous1.getStart() + "\t" +
// previous1.getReference().toString() + "\t" + previous1.getAlternateAlleles());
} else if (iter2.hasNext()) {
// just skip remaining variants
// fixed bug/ was iter1 changed to iter2
VariantContext current = iter2.next();
countFile2++;
if (previous2 != null && current != null && comparator.compare(previous2, current) > 0) {
throw new IllegalStateException(
previous2.getContig()
+ ":"
+ previous2.getStart()
+ " > "
+ current.getContig()
+ ":"
+ current.getStart());
}
previous2 = current;
// logger.info("Skipping: " + previous2.getContig() + ":" + previous2.getStart() + "\t" +
// previous2.getReference().toString() + "\t" + previous2.getAlternateAlleles());
} else {
throw new IllegalStateException("Error should not of reached this point");
}
}
reader1.close();
reader2.close();
logger.info(count + " merged variants");
logger.info(countFile1 + " variants in " + vcf1.getAbsolutePath());
logger.info(countFile2 + " variants in " + vcf2.getAbsolutePath());
} catch (Exception e) {
e.printStackTrace();
} finally {
if (writer != null) {
try {
logger.info("Flushing writer");
writer.close();
} catch (IOException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
}
}
logger.info("finished merging vcfs");
}
@Override
protected void doWork(VcfIterator r, VariantContextWriter w) throws IOException {
AbstractVCFCodec codeIn3 = VCFUtils.createDefaultVCFCodec();
String line;
StringWriter sw = new StringWriter();
LOG.info("opening tabix file: " + this.TABIX);
TabixReader tabix = new TabixReader(this.TABIX);
while ((line = tabix.readLine()) != null) {
if (!line.startsWith(VCFHeader.HEADER_INDICATOR)) {
break;
}
sw.append(line).append("\n");
}
VCFHeader header3 =
(VCFHeader)
codeIn3.readActualHeader(
new LineIteratorImpl(
LineReaderUtil.fromBufferedStream(
new ByteArrayInputStream(sw.toString().getBytes()))));
VCFHeader header1 = r.getHeader();
VCFHeader h2 =
new VCFHeader(header1.getMetaDataInInputOrder(), header1.getSampleNamesInOrder());
for (String infoId : this.INFO_IDS) {
VCFInfoHeaderLine vihl = header3.getInfoHeaderLine(infoId);
if (vihl == null) {
LOG.warn("Not INFO=" + infoId + " in " + TABIX);
continue;
}
if (h2.getInfoHeaderLine(infoId) != null) {
LOG.warn("Input already contains INFO=" + vihl);
}
h2.addMetaDataLine(vihl);
}
if (ALT_CONFLICT_FLAG != null) {
h2.addMetaDataLine(
new VCFInfoHeaderLine(
ALT_CONFLICT_FLAG,
1,
VCFHeaderLineType.Flag,
"conflict ALT allele with " + this.TABIX));
}
w.writeHeader(h2);
while (r.hasNext()) {
VariantContext ctx1 = r.next();
VariantContextBuilder vcb = new VariantContextBuilder(ctx1);
String line2;
String BEST_ID = null;
boolean best_id_match_alt = false;
List<VariantContext> variantsList = new ArrayList<VariantContext>();
int[] array = tabix.parseReg(ctx1.getChr() + ":" + (ctx1.getStart()) + "-" + (ctx1.getEnd()));
TabixReader.Iterator iter = null;
if (array != null && array.length == 3 && array[0] != -1 && array[1] >= 0 && array[2] >= 0) {
iter = tabix.query(array[0], array[1], array[2]);
} else {
LOG.info("Cannot get " + ctx1.getChr() + ":" + (ctx1.getStart()) + "-" + (ctx1.getEnd()));
}
while (iter != null && (line2 = iter.next()) != null) {
VariantContext ctx3 = codeIn3.decode(line2);
if (ctx3.getStart() != ctx1.getStart()) continue;
if (ctx3.getEnd() != ctx1.getEnd()) continue;
if (ctx1.getReference().equals(ctx3.getReference())
&& ctx1.getAlternateAlleles().equals(ctx3.getAlternateAlleles())) {
variantsList.clear();
variantsList.add(ctx3);
break;
} else {
variantsList.add(ctx3);
}
}
for (VariantContext ctx3 : variantsList) {
if (this.REF_ALLELE_MATTERS && !ctx1.getReference().equals(ctx3.getReference())) {
continue;
}
if (this.ALT_ALLELES_MATTERS
&& !ctx1.getAlternateAlleles().equals(ctx3.getAlternateAlleles())) {
continue;
}
if (ctx3.getID() != null && this.REPLACE_ID) {
if (BEST_ID != null && best_id_match_alt) {
// nothing
} else {
BEST_ID = ctx3.getID();
best_id_match_alt = ctx1.getAlternateAlleles().equals(ctx3.getAlternateAlleles());
}
}
for (String id : this.INFO_IDS) {
Object info3 = ctx3.getAttribute(id);
if (info3 == null) {
continue;
}
Object info1 = ctx1.getAttribute(id);
if (info1 != null && !this.REPLACE_INFO_FIELD) {
continue;
}
vcb.attribute(id, info3);
}
if (ALT_CONFLICT_FLAG != null
&& !ctx1.getAlternateAlleles().equals(ctx3.getAlternateAlleles())) {
vcb.attribute(ALT_CONFLICT_FLAG, true);
}
}
if (BEST_ID != null) {
vcb.id(BEST_ID);
}
w.add(vcb.make());
}
tabix.close();
}
@Override
public ListVector convert(Iterable<Genotype> genotypes) {
NamedBuilder fullBuilder = ListVector.newNamedBuilder();
// add a named element in the build for each sample
for (Genotype gt : genotypes) {
NamedBuilder itemBuilder = ListVector.newNamedBuilder();
// add id
itemBuilder.add("ID", gt.getSampleName());
if (header == null) {
// only add genotype
itemBuilder.add("GT", getGT(gt));
} else {
itemBuilder.add("GT", getGT(gt));
Collection<VCFFormatHeaderLine> formatLines = header.getFormatHeaderLines();
for (VCFFormatHeaderLine hl : formatLines) {
String key = hl.getID();
VCFHeaderLineType type = hl.getType();
if (key.equals("GT")) {
continue;
}
Object val = gt.getAnyAttribute(key);
switch (type) {
case String:
{
if (val instanceof List<?>) {
List<String> items = (List<String>) val;
StringArrayVector vec = new StringArrayVector(items);
itemBuilder.add(key, vec);
} else if (val instanceof String[]) {
String[] items = (String[]) val;
StringArrayVector vec = new StringArrayVector(items);
itemBuilder.add(key, vec);
} else if (val instanceof String) {
String items = (String) val;
StringArrayVector vec = new StringArrayVector(items);
itemBuilder.add(key, vec);
}
break;
}
case Float:
{
if (val instanceof List<?>) {
List<Double> items = (List<Double>) val;
DoubleArrayVector vec = new DoubleArrayVector(items);
itemBuilder.add(key, vec);
} else if (val instanceof Double[]) {
double[] items = ArrayUtils.toPrimitive((Double[]) val);
DoubleArrayVector vec = new DoubleArrayVector(items);
itemBuilder.add(key, vec);
} else if (val instanceof Double) {
Double items = (Double) val;
DoubleArrayVector vec = new DoubleArrayVector(items);
itemBuilder.add(key, vec);
} else if (val instanceof String) {
Double items = Double.parseDouble((String) val);
DoubleArrayVector vec = new DoubleArrayVector(items);
itemBuilder.add(key, vec);
}
break;
}
case Integer:
{
if (val instanceof List<?>) {
List<Integer> items = (List<Integer>) val;
int[] arr = new int[items.size()];
for (int i = 0; i < items.size(); i++) {
arr[i] = items.get(i);
}
IntArrayVector vec = new IntArrayVector(arr);
itemBuilder.add(key, vec);
} else if (val instanceof Integer[]) {
int[] items = ArrayUtils.toPrimitive((Integer[]) val);
IntArrayVector vec = new IntArrayVector(items);
itemBuilder.add(key, vec);
} else if (val instanceof Integer) {
Integer items = (Integer) val;
IntArrayVector vec = new IntArrayVector(items);
itemBuilder.add(key, vec);
} else if (val instanceof String) {
Integer items = Integer.parseInt((String) val);
IntArrayVector vec = new IntArrayVector(items);
itemBuilder.add(key, vec);
}
break;
}
case Character:
{
char value = (char) gt.getAnyAttribute(key);
itemBuilder.add(key, new String(new char[] {value}));
break;
}
default:
throw new IllegalStateException(type + " is not supported");
}
}
}
fullBuilder.add(gt.getSampleName(), itemBuilder.build());
}
return fullBuilder.build();
}
@Override
protected void doWork(VcfIterator r, VariantContextWriter w) throws IOException {
long nChanged = 0L;
final String TAG = "INDELFIXED";
VCFHeader header = r.getHeader();
VCFHeader h2 = new VCFHeader(header.getMetaDataInInputOrder(), header.getSampleNamesInOrder());
h2.addMetaDataLine(
new VCFInfoHeaderLine(TAG, 1, VCFHeaderLineType.String, "Fix Indels for @SolenaLS."));
w.writeHeader(h2);
final Pattern dna = Pattern.compile("[ATGCatgc]+");
while (r.hasNext()) {
VariantContext ctx = r.next();
VariantContextBuilder b = new VariantContextBuilder(ctx);
List<Allele> alleles = ctx.getAlternateAlleles();
if (alleles.size() != 1
|| !dna.matcher(ctx.getReference().getBaseString()).matches()
|| !dna.matcher(alleles.get(0).getBaseString()).matches()) {
w.add(ctx);
continue;
}
StringBuffer ref = new StringBuffer(ctx.getReference().getBaseString().toUpperCase());
StringBuffer alt = new StringBuffer(alleles.get(0).getBaseString().toUpperCase());
int start = ctx.getStart();
int end = ctx.getEnd();
boolean changed = false;
/** ** we trim on the right side *** */
// REF=TGCTGCGGGGGCCGCTGCGGGGG ALT=TGCTGCGGGGG
while (alt.length() > 1
&& alt.length() < ref.length()
&& ref.charAt(ref.length() - 1) == alt.charAt(alt.length() - 1)) {
changed = true;
ref.setLength(ref.length() - 1);
alt.deleteCharAt(alt.length() - 1);
end--;
}
// REF=TGCTGCGGGGG ALT= TGCTGCGGGGGCCGCTGCGGGGG
while (ref.length() > 1
&& alt.length() > ref.length()
&& ref.charAt(ref.length() - 1) == alt.charAt(alt.length() - 1)) {
changed = true;
ref.setLength(ref.length() - 1);
alt.deleteCharAt(alt.length() - 1);
end--;
}
/** ** we trim on the left side *** */
// REF=TGCTGCGGGGGCCGCTGCGGGGG ALT=TGCTGCGGGGG
while (alt.length() > 1 && alt.length() < ref.length() && ref.charAt(0) == alt.charAt(0)) {
changed = true;
ref.deleteCharAt(0);
alt.deleteCharAt(0);
start++;
}
// REF=TGCTGCGGGGG ALT= TGCTGCGGGGGCCGCTGCGGGGG
while (ref.length() > 1 && alt.length() > ref.length() && ref.charAt(0) == alt.charAt(0)) {
changed = true;
ref.deleteCharAt(0);
alt.deleteCharAt(0);
start++;
}
if (!changed) {
w.add(ctx);
continue;
}
/*
LOG.info(line);
LOG.info("ctx.getStart() "+ctx.getStart());
LOG.info("ctx.getEnd() "+ ctx.getEnd());
LOG.info("start " + start);
LOG.info("end "+end);
LOG.info("ref " + ref.toString());
LOG.info("alt "+alt.toString());
*/
Allele newRef = Allele.create(ref.toString(), true);
Allele newAlt = Allele.create(alt.toString(), false);
Allele newalleles[] = new Allele[] {newRef, newAlt};
b.attribute(
TAG,
ctx.getReference().getBaseString()
+ "|"
+ alleles.get(0).getBaseString()
+ "|"
+ ctx.getStart());
b.start(start);
b.stop(end);
b.alleles(Arrays.asList(newalleles));
nChanged++;
VariantContext ctx2 = b.make();
try {
w.add(ctx2);
} catch (TribbleException err) {
error(err, "Cannot convert new context:" + ctx2 + " old context:" + ctx);
w.add(ctx);
}
}
info("indels changed:" + nChanged);
}