Examples with CloseableIterator htsjdk.samtools.util.CloseableIterator used on opensource projects

Example 41 with CloseableIterator

use of htsjdk.samtools.util.CloseableIterator in project jvarkit by lindenb.

the class VcfPeekVcf method getOverlappingBuffer.

private List<VariantContext> getOverlappingBuffer(final String contig, final int start, final int end) {
    if (!(this.last_buffer_interval != null && this.last_buffer_interval.getContig().equals(contig) && this.last_buffer_interval.getStart() < start && end < this.last_buffer_interval.getEnd())) {
        this.buffer.clear();
        this.last_buffer_interval = new Interval(contig, Math.max(0, start - 1), (end + 1 + this.buffer_size));
        try (CloseableIterator<VariantContext> t = this.indexedVcfFileReader.query(contig, Math.max(0, start - 1), (end + 1 + this.buffer_size))) {
            while (t.hasNext()) {
                VariantContext ctx = t.next();
                if (// reduce memory
                ctx.hasGenotypes()) {
                    ctx = new VariantContextBuilder(ctx).noGenotypes().make();
                }
                this.buffer.add(ctx);
            }
        }
    }
    return this.buffer.stream().filter(V -> V.getContig().equals(contig) && CoordMath.overlaps(V.getStart(), V.getEnd(), start, end)).collect(Collectors.toList());
}

Example 42 with CloseableIterator

use of htsjdk.samtools.util.CloseableIterator 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) {
        }
    }
}

Example 43 with CloseableIterator

use of htsjdk.samtools.util.CloseableIterator in project gridss by PapenfussLab.

the class SAMEvidenceSource method iterator.

public CloseableIterator<DirectedEvidence> iterator(final QueryInterval[] intervals, EvidenceSortOrder eso) {
    SamReader reader = getReader();
    // expand query bounds as the alignment for a discordant read pair could fall before or after the breakend interval we are extracting
    QueryInterval[] expandedIntervals = QueryIntervalUtil.padIntervals(getContext().getDictionary(), intervals, getMaxConcordantFragmentSize() + 1);
    // ignore blacklisted regions
    IntervalBed queryInterval = new IntervalBed(getContext().getLinear(), expandedIntervals);
    queryInterval.remove(getBlacklistedRegions());
    CloseableIterator it = tryOpenReader(reader, queryInterval.asQueryInterval());
    if (Defaults.SANITY_CHECK_DUMP_ITERATORS) {
        it = new AutoClosingIterator<>(new DebugSpammingIterator<>(it, "SAMEvidenceSource.iterator().rawiterator"));
    }
    Iterator<DirectedEvidence> eit = asEvidence(it, eso);
    eit = Iterators.filter(eit, e -> QueryIntervalUtil.overlaps(intervals, e.getBreakendSummary()));
    if (Defaults.SANITY_CHECK_DUMP_ITERATORS) {
        eit = new AutoClosingIterator<>(new DebugSpammingIterator<>(eit, "SAMEvidenceSource.iterator(QueryInterval[]).filter"));
    }
    return new AutoClosingIterator<>(eit, reader, it);
}

Example 44 with CloseableIterator

use of htsjdk.samtools.util.CloseableIterator in project gatk-protected by broadinstitute.

the class CreateSomaticPanelOfNormals method doWork.

public Object doWork() {
    final List<File> inputVcfs = new ArrayList<>(vcfs);
    final Collection<CloseableIterator<VariantContext>> iterators = new ArrayList<>(inputVcfs.size());
    final Collection<VCFHeader> headers = new HashSet<>(inputVcfs.size());
    final VCFHeader headerOfFirstVcf = new VCFFileReader(inputVcfs.get(0), false).getFileHeader();
    final SAMSequenceDictionary sequenceDictionary = headerOfFirstVcf.getSequenceDictionary();
    final VariantContextComparator comparator = headerOfFirstVcf.getVCFRecordComparator();
    for (final File vcf : inputVcfs) {
        final VCFFileReader reader = new VCFFileReader(vcf, false);
        iterators.add(reader.iterator());
        final VCFHeader header = reader.getFileHeader();
        Utils.validateArg(comparator.isCompatible(header.getContigLines()), () -> vcf.getAbsolutePath() + " has incompatible contigs.");
        headers.add(header);
    }
    final VariantContextWriter writer = GATKVariantContextUtils.createVCFWriter(outputVcf, sequenceDictionary, false, Options.INDEX_ON_THE_FLY);
    writer.writeHeader(new VCFHeader(VCFUtils.smartMergeHeaders(headers, false)));
    final MergingIterator<VariantContext> mergingIterator = new MergingIterator<>(comparator, iterators);
    SimpleInterval currentPosition = new SimpleInterval("FAKE", 1, 1);
    final List<VariantContext> variantsAtThisPosition = new ArrayList<>(20);
    while (mergingIterator.hasNext()) {
        final VariantContext vc = mergingIterator.next();
        if (!currentPosition.overlaps(vc)) {
            processVariantsAtSamePosition(variantsAtThisPosition, writer);
            variantsAtThisPosition.clear();
            currentPosition = new SimpleInterval(vc.getContig(), vc.getStart(), vc.getStart());
        }
        variantsAtThisPosition.add(vc);
    }
    mergingIterator.close();
    writer.close();
    return "SUCCESS";
}

Example 45 with CloseableIterator

use of htsjdk.samtools.util.CloseableIterator in project gatk by broadinstitute.

the class MultiVariantDataSource method getMergedIteratorFromDataSources.

/**
     * Close any existing iterator, create a new iterator and update the local cached iterator reference.
     * @param iteratorFromSource function to retrieve individual iterator, to be applied to each data source
     * @return
     */
private Iterator<VariantContext> getMergedIteratorFromDataSources(final Function<FeatureDataSource<VariantContext>, Iterator<VariantContext>> iteratorFromSource) {
    // Tribble documentation states that having multiple iterators open simultaneously over the same FeatureReader
    // results in undefined behavior
    closeOpenIterationIfNecessary();
    if (featureDataSources.size() > 1) {
        final List<CloseableIterator<VariantContext>> iterators = new ArrayList<>(featureDataSources.size());
        featureDataSources.forEach(ds -> iterators.add(getCloseableIteratorWrapper(iteratorFromSource.apply((ds)))));
        final VariantContextComparator varComparator = new VariantContextComparator(getSequenceDictionary());
        currentIterator = new MergingIterator<>(varComparator, iterators);
    } else {
        currentIterator = getCloseableIteratorWrapper(iteratorFromSource.apply(featureDataSources.get(0)));
    }
    return currentIterator;
}