Examples with SAMReadGroupRecord htsjdk.samtools.SAMReadGroupRecord used on opensource projects

Example 56 with SAMReadGroupRecord

use of htsjdk.samtools.SAMReadGroupRecord in project gatk by broadinstitute.

the class ReadGroupBlackListReadFilterUnitTest method testFilterOutByFile.

@Test
public void testFilterOutByFile() throws IOException {
    int recordsPerGroup = 3;
    List<GATKRead> records = new ArrayList<>();
    int alignmentStart = 0;
    for (int x = 0; x < GROUP_COUNT; x++) {
        SAMReadGroupRecord groupRecord = header.getReadGroup(getReadGroupId(x));
        for (int y = 1; y <= recordsPerGroup; y++) {
            GATKRead record = ArtificialReadUtils.createArtificialRead(header, "readUno", 0, ++alignmentStart, 20);
            record.setReadGroup(groupRecord.getReadGroupId());
            records.add(record);
        }
    }
    List<String> filterList = new ArrayList<>();
    filterList.add(publicTestDir + "readgroupblacklisttest.txt");
    ReadGroupBlackListReadFilter filter = new ReadGroupBlackListReadFilter(filterList, header);
    int filtered = 0;
    int unfiltered = 0;
    for (GATKRead record : records) {
        String readGroup = record.getReadGroup();
        if (!filter.test(record)) {
            if (!("ReadGroup3".equals(readGroup) || "ReadGroup4".equals(readGroup)))
                Assert.fail("Read group " + readGroup + " was filtered");
            filtered++;
        } else {
            if ("ReadGroup3".equals(readGroup) || "ReadGroup4".equals(readGroup))
                Assert.fail("Read group " + readGroup + " was not filtered");
            unfiltered++;
        }
    }
    int filteredExpected = recordsPerGroup * 2;
    int unfilteredExpected = recordsPerGroup * (GROUP_COUNT - 2);
    Assert.assertEquals(filtered, filteredExpected, "Filtered");
    Assert.assertEquals(unfiltered, unfilteredExpected, "Uniltered");
}

Example 57 with SAMReadGroupRecord

use of htsjdk.samtools.SAMReadGroupRecord in project gatk by broadinstitute.

the class RecalUtils method parsePlatformForRead.

/**
     * Section of code shared between the two recalibration walkers which uses the command line arguments to adjust attributes of the read such as quals or platform string
     *
     * @param read The read to adjust
     * @param RAC  The list of shared command line arguments
     */
public static void parsePlatformForRead(final GATKRead read, final SAMFileHeader header, final RecalibrationArgumentCollection RAC) {
    final SAMReadGroupRecord readGroup = ReadUtils.getSAMReadGroupRecord(read, header);
    if (RAC.FORCE_PLATFORM != null && (readGroup.getPlatform() == null || !readGroup.getPlatform().equals(RAC.FORCE_PLATFORM))) {
        readGroup.setPlatform(RAC.FORCE_PLATFORM);
    }
    if (readGroup.getPlatform() == null) {
        if (RAC.DEFAULT_PLATFORM != null) {
            if (!warnUserNullPlatform) {
                Utils.warnUser("The input .bam file contains reads with no platform information. " + "Defaulting to platform = " + RAC.DEFAULT_PLATFORM + ". " + "First observed at read with name = " + read.getName());
                warnUserNullPlatform = true;
            }
            readGroup.setPlatform(RAC.DEFAULT_PLATFORM);
        } else {
            throw new UserException.MalformedRead(read, "The input .bam file contains reads with no platform information. First observed at read with name = " + read.getName());
        }
    }
}

Example 58 with SAMReadGroupRecord

use of htsjdk.samtools.SAMReadGroupRecord in project gatk by broadinstitute.

the class ReadGroupCovariate method recordValues.

@Override
public void recordValues(final GATKRead read, final SAMFileHeader header, final ReadCovariates values, final boolean recordIndelValues) {
    final SAMReadGroupRecord rg = ReadUtils.getSAMReadGroupRecord(read, header);
    final String readGroupId = getID(rg);
    final int key = keyForReadGroup(readGroupId);
    final int readLength = read.getLength();
    for (int i = 0; i < readLength; i++) {
        values.addCovariate(key, key, key, i);
    }
}

Example 59 with SAMReadGroupRecord

use of htsjdk.samtools.SAMReadGroupRecord in project gatk by broadinstitute.

the class AlignmentContextUnitTest method testNoSample.

@Test(expectedExceptions = UserException.ReadMissingReadGroup.class)
public void testNoSample() throws Exception {
    final SAMReadGroupRecord readGroupOne = new SAMReadGroupRecord("rg1");
    final SAMFileHeader header = ArtificialReadUtils.createArtificialSamHeader(1, 1, 1000);
    header.addReadGroup(readGroupOne);
    final Locatable loc = new SimpleInterval("chr1", 1, 1);
    final GATKRead read1 = ArtificialReadUtils.createArtificialRead(header, "read1", 0, 1, 10);
    read1.setReadGroup(readGroupOne.getId());
    final ReadPileup pileup = new ReadPileup(loc, Arrays.asList(read1), 1);
    final AlignmentContext ac = new AlignmentContext(loc, pileup);
    ac.splitContextBySampleName(header);
}

Example 60 with SAMReadGroupRecord

use of htsjdk.samtools.SAMReadGroupRecord in project gatk by broadinstitute.

the class LocusWalkerSpark method getAlignmentsFunction.

/**
     * Return a function that maps a {@link Shard} of reads into a tuple of alignments and their corresponding reference and features.
     * @param bReferenceSource the reference source broadcast
     * @param bFeatureManager the feature manager broadcast
     * @param sequenceDictionary the sequence dictionary for the reads
     * @param header the reads header
     * @param downsamplingInfo the downsampling method for the reads
     * @return a function that maps a {@link Shard} of reads into a tuple of alignments and their corresponding reference and features.
     */
private static FlatMapFunction<Shard<GATKRead>, LocusWalkerContext> getAlignmentsFunction(Broadcast<ReferenceMultiSource> bReferenceSource, Broadcast<FeatureManager> bFeatureManager, SAMSequenceDictionary sequenceDictionary, SAMFileHeader header, LIBSDownsamplingInfo downsamplingInfo) {
    return (FlatMapFunction<Shard<GATKRead>, LocusWalkerContext>) shardedRead -> {
        SimpleInterval interval = shardedRead.getInterval();
        SimpleInterval paddedInterval = shardedRead.getPaddedInterval();
        Iterator<GATKRead> readIterator = shardedRead.iterator();
        ReferenceDataSource reference = bReferenceSource == null ? null : new ReferenceMemorySource(bReferenceSource.getValue().getReferenceBases(null, paddedInterval), sequenceDictionary);
        FeatureManager fm = bFeatureManager == null ? null : bFeatureManager.getValue();
        final Set<String> samples = header.getReadGroups().stream().map(SAMReadGroupRecord::getSample).collect(Collectors.toSet());
        LocusIteratorByState libs = new LocusIteratorByState(readIterator, downsamplingInfo, false, samples, header, true, false);
        IntervalOverlappingIterator<AlignmentContext> alignmentContexts = new IntervalOverlappingIterator<>(libs, ImmutableList.of(interval), sequenceDictionary);
        final Spliterator<AlignmentContext> alignmentContextSpliterator = Spliterators.spliteratorUnknownSize(alignmentContexts, 0);
        return StreamSupport.stream(alignmentContextSpliterator, false).map(alignmentContext -> {
            final SimpleInterval alignmentInterval = new SimpleInterval(alignmentContext);
            return new LocusWalkerContext(alignmentContext, new ReferenceContext(reference, alignmentInterval), new FeatureContext(fm, alignmentInterval));
        }).iterator();
    };
}