use of com.github.lindenb.jvarkit.util.vcf.ContigPosRef in project jvarkit by lindenb.
the class Biostar86363 method doWork.
@Override
public int doWork(List<String> args) {
if (genotypeFile == null) {
LOG.error("undefined genotype file");
return -1;
}
BufferedReader in = null;
try {
Pattern tab = Pattern.compile("[\t]");
in = IOUtils.openFileForBufferedReading(this.genotypeFile);
String line;
while ((line = in.readLine()) != null) {
if (line.isEmpty() || line.startsWith("#"))
continue;
String[] tokens = tab.split(line);
if (tokens.length < 4) {
LOG.error("Bad line in " + line);
in.close();
in = null;
return -1;
}
ContigPosRef cap = new ContigPosRef(tokens[0], Integer.parseInt(tokens[1]), Allele.create(tokens[2], true));
Set<String> samples = this.pos2sample.get(cap);
if (samples == null) {
samples = new HashSet<String>();
this.pos2sample.put(cap, samples);
}
samples.add(tokens[3]);
}
in.close();
return doVcfToVcf(args, outputFile);
} catch (Exception err) {
LOG.error(err);
return -1;
} finally {
CloserUtil.close(in);
}
}
use of com.github.lindenb.jvarkit.util.vcf.ContigPosRef in project jvarkit by lindenb.
the class VCFComm method doWork.
@Override
public int doWork(final List<String> args) {
CloseableIterator<LineAndFile> iter = null;
SortingCollection<LineAndFile> variants = null;
VariantContextWriter w = null;
try {
if (args.isEmpty()) {
LOG.error("Illegal number of arguments");
return -1;
}
Set<VCFHeaderLine> metaData = new HashSet<VCFHeaderLine>();
variants = SortingCollection.newInstance(LineAndFile.class, new LineAndFileCodec(), new LineAndFileComparator(), super.sortingCollectionArgs.getMaxRecordsInRam(), super.sortingCollectionArgs.getTmpPaths());
variants.setDestructiveIteration(true);
/**
* new sample names in the output vcf: one sample per file
*/
final Map<Integer, String> fileid2sampleName = new TreeMap<>();
/**
* samples names as they appear in the original VCF headers
*/
final Counter<String> countInputSamples = new Counter<String>();
/**
* dicts
*/
final List<SAMSequenceDictionary> all_dictionaries = new ArrayList<>();
for (final String vcffilename : IOUtils.unrollFiles(args)) {
LOG.info("Reading from " + vcffilename);
final Input input = super.put(variants, vcffilename);
String sampleName = vcffilename;
if (sampleName.endsWith(".vcf.gz")) {
sampleName = sampleName.substring(0, sampleName.length() - 7);
} else if (sampleName.endsWith(".vcf.gz")) {
sampleName = sampleName.substring(0, sampleName.length() - 4);
}
int slash = sampleName.lastIndexOf(File.separatorChar);
if (slash != -1)
sampleName = sampleName.substring(slash + 1);
int suffix = 1;
// loop until we find a uniq name
for (; ; ) {
final String key = sampleName + (suffix == 1 ? "" : "_" + suffix);
if (fileid2sampleName.values().contains(key)) {
suffix++;
continue;
}
fileid2sampleName.put(input.file_id, key);
metaData.add(new VCFHeaderLine(key, vcffilename));
break;
}
for (final String sname : input.codecAndHeader.header.getSampleNamesInOrder()) {
countInputSamples.incr(sname);
}
all_dictionaries.add(input.codecAndHeader.header.getSequenceDictionary());
}
variants.doneAdding();
/**
* unique sample name, if any present in all VCF
*/
Optional<String> unqueSampleName = Optional.empty();
if (countInputSamples.getCountCategories() == 1 && countInputSamples.count(countInputSamples.keySet().iterator().next()) == fileid2sampleName.size()) {
unqueSampleName = Optional.of(countInputSamples.keySet().iterator().next());
LOG.info("Unique sample name is " + unqueSampleName.get());
}
VCFStandardHeaderLines.addStandardFormatLines(metaData, true, VCFConstants.DEPTH_KEY, VCFConstants.GENOTYPE_QUALITY_KEY, VCFConstants.GENOTYPE_KEY, VCFConstants.GENOTYPE_FILTER_KEY);
VCFStandardHeaderLines.addStandardInfoLines(metaData, true, VCFConstants.DEPTH_KEY, VCFConstants.ALLELE_COUNT_KEY, VCFConstants.ALLELE_NUMBER_KEY);
metaData.add(new VCFHeaderLine(getClass().getSimpleName(), "version:" + getVersion() + " command:" + getProgramCommandLine()));
final VCFFilterHeaderLine variantNotCalledInAllVcf = new VCFFilterHeaderLine("NotCalledEveryWhere", "Variant was NOT called in all input VCF");
metaData.add(variantNotCalledInAllVcf);
final VCFFilterHeaderLine variantWasFiltered = new VCFFilterHeaderLine("VariantWasFiltered", "At least one variant was filtered");
metaData.add(variantWasFiltered);
final VCFFormatHeaderLine variantQUALFormat = new VCFFormatHeaderLine("VCQUAL", 1, VCFHeaderLineType.Float, "Variant Quality");
metaData.add(variantQUALFormat);
metaData.add(new VCFFormatHeaderLine(VCFConstants.ALLELE_NUMBER_KEY, 1, VCFHeaderLineType.Integer, "Number of allle in the src vcf"));
metaData.add(new VCFFormatHeaderLine(VCFConstants.ALLELE_COUNT_KEY, 1, VCFHeaderLineType.Integer, "Number of ALT alllele"));
final VCFInfoHeaderLine foundInCountVcfInfo = new VCFInfoHeaderLine("NVCF", 1, VCFHeaderLineType.Integer, "Number of VCF this variant was found");
metaData.add(foundInCountVcfInfo);
final VCFInfoHeaderLine variantTypesInfo = new VCFInfoHeaderLine("VTYPES", VCFHeaderLineCount.UNBOUNDED, VCFHeaderLineType.String, "Distinct Variants type");
metaData.add(variantTypesInfo);
final VCFFilterHeaderLine multipleTypeFilters = new VCFFilterHeaderLine("DiscordantTypes", "Discordant types at this position");
metaData.add(multipleTypeFilters);
final VCFFormatHeaderLine variantTypeFormat = new VCFFormatHeaderLine("VTYPE", 1, VCFHeaderLineType.String, "Variant Type");
metaData.add(variantTypeFormat);
final VCFFilterHeaderLine uniqueVariantDiscordantGTFilter;
if (unqueSampleName.isPresent()) {
metaData.add(new VCFHeaderLine("UniqSample", unqueSampleName.get()));
uniqueVariantDiscordantGTFilter = new VCFFilterHeaderLine("DiscordantGenotypeForUniqSample", "Genotype Dicordant for for sample " + unqueSampleName.get());
metaData.add(uniqueVariantDiscordantGTFilter);
} else {
uniqueVariantDiscordantGTFilter = null;
}
final VCFHeader header = new VCFHeader(metaData, new ArrayList<>(fileid2sampleName.values()));
if (// all have a dict
!normalize_chr && !all_dictionaries.contains(null)) {
SAMSequenceDictionary thedict = null;
for (int x = 0; x < all_dictionaries.size(); ++x) {
SAMSequenceDictionary d = all_dictionaries.get(x);
if (thedict == null) {
thedict = d;
} else if (!SequenceUtil.areSequenceDictionariesEqual(d, thedict)) {
thedict = null;
break;
}
}
if (thedict != null)
header.setSequenceDictionary(thedict);
}
w = super.openVariantContextWriter(super.outputFile);
w.writeHeader(header);
final List<LineAndFile> row = new ArrayList<LineAndFile>(super.inputs.size());
final Comparator<LineAndFile> posCompare = (A, B) -> A.getContigPosRef().compareTo(B.getContigPosRef());
iter = variants.iterator();
for (; ; ) {
LineAndFile rec = null;
if (iter.hasNext()) {
rec = iter.next();
}
if (rec == null || (!row.isEmpty() && posCompare.compare(row.get(0), rec) != 0)) {
if (!row.isEmpty()) {
final VariantContext first = row.get(0).getContext();
/* in which file id we find this variant */
Set<Integer> fileids_for_variant = row.stream().map(LAF -> LAF.fileIdx).collect(Collectors.toSet());
// see with HAS multiple chrom/pos/ref but different alt
if (row.size() != fileids_for_variant.size()) {
for (; ; ) {
boolean ok = true;
for (int x = 0; ok && x + 1 < row.size(); ++x) {
final VariantContext ctxx = row.get(x).getContext();
final List<Allele> altsx = ctxx.getAlternateAlleles();
for (int y = x + 1; ok && y < row.size(); ++y) {
if (row.get(x).fileIdx != row.get(y).fileIdx)
continue;
final VariantContext ctxy = row.get(y).getContext();
final List<Allele> altsy = ctxy.getAlternateAlleles();
if (altsx.equals(altsy))
continue;
if (!ctxx.isVariant() && ctxy.isVariant()) {
row.remove(x);
} else if (ctxx.isVariant() && !ctxy.isVariant()) {
row.remove(y);
} else if (!ctxx.isSNP() && ctxy.isSNP()) {
row.remove(x);
} else if (ctxx.isSNP() && !ctxy.isSNP()) {
row.remove(y);
} else if (altsx.size() > altsy.size()) {
row.remove(x);
} else if (altsx.size() < altsy.size()) {
row.remove(y);
} else {
row.remove(y);
}
ok = false;
break;
}
}
if (ok)
break;
}
fileids_for_variant = row.stream().map(LAF -> LAF.fileIdx).collect(Collectors.toSet());
}
if (row.size() != fileids_for_variant.size()) {
LOG.error("There are some duplicated variants at the position " + new ContigPosRef(first) + " in the same vcf file");
for (final LineAndFile laf : row) {
LOG.error("File [" + laf.fileIdx + "]" + fileid2sampleName.get(laf.fileIdx));
LOG.error("\t" + laf.getContigPosRef());
}
row.clear();
} else {
final Set<Allele> alleles = row.stream().flatMap(R -> R.getContext().getAlleles().stream()).collect(Collectors.toSet());
final VariantContextBuilder vcb = new VariantContextBuilder(getClass().getName(), first.getContig(), first.getStart(), first.getEnd(), alleles);
final Set<String> filters = new HashSet<>();
final Set<VariantContext.Type> variantContextTypes = new HashSet<>();
final List<Genotype> genotypes = new ArrayList<Genotype>();
for (final LineAndFile laf : row) {
if (laf.getContext().isFiltered())
filters.add(variantWasFiltered.getID());
variantContextTypes.add(laf.getContext().getType());
final GenotypeBuilder gbuilder = new GenotypeBuilder();
gbuilder.name(fileid2sampleName.get(laf.fileIdx));
if (unqueSampleName.isPresent()) {
final Genotype g0 = laf.getContext().getGenotype(unqueSampleName.get());
if (g0 == null) {
iter.close();
w.close();
throw new IllegalStateException("Cannot find genotype for " + unqueSampleName.get());
}
if (g0.hasDP())
gbuilder.DP(g0.getDP());
if (g0.hasGQ())
gbuilder.GQ(g0.getGQ());
gbuilder.alleles(g0.getAlleles());
} else {
gbuilder.alleles(Arrays.asList(first.getReference(), first.getReference()));
if (laf.getContext().hasAttribute(VCFConstants.DEPTH_KEY)) {
gbuilder.DP(laf.getContext().getAttributeAsInt(VCFConstants.DEPTH_KEY, 0));
}
}
if (laf.getContext().isFiltered()) {
gbuilder.filter("VCFFILTERED");
}
if (laf.getContext().hasLog10PError()) {
gbuilder.attribute(variantQUALFormat.getID(), laf.getContext().getPhredScaledQual());
}
gbuilder.attribute(VCFConstants.ALLELE_NUMBER_KEY, laf.getContext().getGenotypes().stream().flatMap(G -> G.getAlleles().stream()).filter(A -> !A.isNoCall()).count());
gbuilder.attribute(VCFConstants.ALLELE_COUNT_KEY, laf.getContext().getGenotypes().stream().flatMap(G -> G.getAlleles().stream()).filter(A -> !(A.isReference() || A.isNoCall())).count());
gbuilder.attribute(variantTypeFormat.getID(), laf.getContext().getType().name());
genotypes.add(gbuilder.make());
}
final String id = String.join(";", row.stream().map(LAF -> LAF.getContext()).filter(V -> V.hasID()).map(V -> V.getID()).collect(Collectors.toSet()));
if (!id.isEmpty())
vcb.id(id);
vcb.genotypes(genotypes);
if (unqueSampleName.isPresent()) {
boolean all_same = true;
for (int x = 0; all_same && x + 1 < genotypes.size(); ++x) {
if (!genotypes.get(x).isCalled())
continue;
for (int y = x + 1; all_same && y < genotypes.size(); ++y) {
if (!genotypes.get(y).isCalled())
continue;
if (!genotypes.get(x).sameGenotype(genotypes.get(y), true)) {
all_same = false;
break;
}
}
}
if (!all_same)
filters.add(uniqueVariantDiscordantGTFilter.getID());
}
// Add AN
vcb.attribute(VCFConstants.ALLELE_NUMBER_KEY, genotypes.stream().filter(G -> G.isCalled()).mapToInt(G -> G.getAlleles().size()).sum());
if (!variantContextTypes.isEmpty()) {
vcb.attribute(variantTypesInfo.getID(), new ArrayList<>(variantContextTypes.stream().map(T -> T.name()).collect(Collectors.toSet())));
if (variantContextTypes.size() > 1) {
filters.add(multipleTypeFilters.getID());
}
}
vcb.attribute(foundInCountVcfInfo.getID(), fileids_for_variant.size());
boolean print = true;
if (row.size() == super.inputs.size() && ignore_everywhere) {
print = false;
}
if (fileids_for_variant.size() != fileid2sampleName.size()) {
filters.add(variantNotCalledInAllVcf.getID());
if (only_everywhere) {
print = false;
}
}
vcb.filters(filters);
if (print) {
w.add(vcb.make());
}
}
row.clear();
}
if (rec == null)
break;
}
row.add(rec);
}
iter.close();
iter = null;
w.close();
w = null;
return 0;
} catch (Exception err) {
LOG.error(err);
return -1;
} finally {
CloserUtil.close(iter);
CloserUtil.close(w);
try {
if (variants != null)
variants.cleanup();
} catch (Exception err) {
}
}
}
use of com.github.lindenb.jvarkit.util.vcf.ContigPosRef in project jvarkit by lindenb.
the class Biostar86363 method doVcfToVcf.
@Override
protected int doVcfToVcf(String inputName, VcfIterator in, VariantContextWriter out) {
final List<Allele> empty_g = new ArrayList<Allele>(2);
empty_g.add(Allele.NO_CALL);
empty_g.add(Allele.NO_CALL);
VCFHeader h = in.getHeader();
final List<String> vcf_samples = h.getSampleNamesInOrder();
h.addMetaDataLine(new VCFFormatHeaderLine("GR", 1, VCFHeaderLineType.Integer, "(1) = Genotype was reset by " + getProgramName()));
this.recalculator.setHeader(h);
out.writeHeader(h);
while (in.hasNext()) {
VariantContext ctx = in.next();
ContigPosRef cap = new ContigPosRef(ctx);
Set<String> samplesToReset = this.pos2sample.get(cap);
if (samplesToReset != null) {
VariantContextBuilder vcb = new VariantContextBuilder(ctx);
List<Genotype> genotypes = new ArrayList<Genotype>();
for (String sample : vcf_samples) {
Genotype g = ctx.getGenotype(sample);
if (g == null)
continue;
GenotypeBuilder gb = new GenotypeBuilder(g);
if (samplesToReset.contains(sample)) {
gb.alleles(empty_g);
gb.attribute("GR", 1);
} else {
gb.attribute("GR", 0);
}
g = gb.make();
genotypes.add(g);
}
vcb.genotypes(genotypes);
ctx = this.recalculator.apply(vcb.make());
}
out.add(ctx);
}
return 0;
}
Aggregations