Examples with LiftOver htsjdk.samtools.liftover.LiftOver used on opensource projects

Example 1 with LiftOver

use of htsjdk.samtools.liftover.LiftOver in project gatk by broadinstitute.

the class LiftOverIntervalList method doWork.

/**
     * Do the work after command line has been parsed. RuntimeException may be
     * thrown by this method, and are reported appropriately.
     */
@Override
protected Object doWork() {
    assertFileIsReadable(INPUT);
    assertFileIsReadable(SEQUENCE_DICTIONARY);
    assertFileIsReadable(CHAIN);
    assertFileIsWritable(OUTPUT);
    final LiftOver liftOver = new LiftOver(CHAIN);
    liftOver.setLiftOverMinMatch(MIN_LIFTOVER_PCT);
    final IntervalList fromIntervals = fromFile(INPUT);
    final SAMFileHeader toHeader = makeDefault().getFileHeader(SEQUENCE_DICTIONARY);
    liftOver.validateToSequences(toHeader.getSequenceDictionary());
    final IntervalList toIntervals = new IntervalList(toHeader);
    boolean anyFailed = false;
    for (final Interval fromInterval : fromIntervals) {
        final Interval toInterval = liftOver.liftOver(fromInterval);
        if (toInterval != null) {
            toIntervals.add(toInterval);
        } else {
            anyFailed = true;
            logger.warn("Liftover failed for ", fromInterval, "(len ", fromInterval.length(), ")");
            final List<LiftOver.PartialLiftover> partials = liftOver.diagnosticLiftover(fromInterval);
            for (final LiftOver.PartialLiftover partial : partials) {
                logger.info(partial);
            }
        }
    }
    toIntervals.sorted();
    toIntervals.write(OUTPUT);
    return anyFailed ? 1 : 0;
}

Example 2 with LiftOver

use of htsjdk.samtools.liftover.LiftOver in project gatk by broadinstitute.

the class LiftOverVcf method doWork.

@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;
}