@Test
public void testWholeIndelReadInIsolation() {
final int firstLocus = 44367789;
// create a test version of the Reads object
ReadProperties readAttributes = createTestReadProperties();
SAMRecord indelOnlyRead =
ArtificialSAMUtils.createArtificialRead(header, "indelOnly", 0, firstLocus, 76);
indelOnlyRead.setReadBases(Utils.dupBytes((byte) 'A', 76));
indelOnlyRead.setBaseQualities(Utils.dupBytes((byte) '@', 76));
indelOnlyRead.setCigarString("76I");
List<SAMRecord> reads = Arrays.asList(indelOnlyRead);
// create the iterator by state with the fake reads and fake records
li = makeLTBS(reads, readAttributes);
// Traditionally, reads that end with indels bleed into the pileup at the following locus.
// Verify that the next pileup contains this read
// and considers it to be an indel-containing read.
Assert.assertTrue(
li.hasNext(),
"Should have found a whole-indel read in the normal base pileup without extended events enabled");
AlignmentContext alignmentContext = li.next();
Assert.assertEquals(
alignmentContext.getLocation().getStart(),
firstLocus,
"Base pileup is at incorrect location.");
ReadBackedPileup basePileup = alignmentContext.getBasePileup();
Assert.assertEquals(basePileup.getReads().size(), 1, "Pileup is of incorrect size");
Assert.assertSame(basePileup.getReads().get(0), indelOnlyRead, "Read in pileup is incorrect");
}
@Test(dataProvider = "LIBSTest")
public void testLIBS(LIBSTest params) {
final int locus = 44367788;
SAMRecord read =
ArtificialSAMUtils.createArtificialRead(header, "read", 0, locus, params.readLength);
read.setReadBases(Utils.dupBytes((byte) 'A', params.readLength));
read.setBaseQualities(Utils.dupBytes((byte) '@', params.readLength));
read.setCigarString(params.cigar);
// create the iterator by state with the fake reads and fake records
li = makeLTBS(Arrays.asList(read), createTestReadProperties());
final LIBS_position tester = new LIBS_position(read);
while (li.hasNext()) {
AlignmentContext alignmentContext = li.next();
ReadBackedPileup p = alignmentContext.getBasePileup();
Assert.assertTrue(p.getNumberOfElements() == 1);
PileupElement pe = p.iterator().next();
tester.stepForwardOnGenome();
Assert.assertEquals(pe.isBeforeDeletedBase(), tester.isBeforeDeletedBase);
Assert.assertEquals(pe.isBeforeDeletionStart(), tester.isBeforeDeletionStart);
Assert.assertEquals(pe.isAfterDeletedBase(), tester.isAfterDeletedBase);
Assert.assertEquals(pe.isAfterDeletionEnd(), tester.isAfterDeletionEnd);
Assert.assertEquals(pe.isBeforeInsertion(), tester.isBeforeInsertion);
Assert.assertEquals(pe.isAfterInsertion(), tester.isAfterInsertion);
Assert.assertEquals(pe.isNextToSoftClip(), tester.isNextToSoftClip);
Assert.assertEquals(pe.getOffset(), tester.getCurrentReadOffset());
}
}
@Override
public String map(RefMetaDataTracker tracker, ReferenceContext ref, AlignmentContext context) {
final String rods = getReferenceOrderedData(tracker);
ReadBackedPileup basePileup = context.getBasePileup();
final StringBuilder s = new StringBuilder();
s.append(String.format("%s %s", basePileup.getPileupString((char) ref.getBase()), rods));
if (SHOW_VERBOSE) s.append(" ").append(createVerboseOutput(basePileup));
s.append("\n");
return s.toString();
}
private Map<String, AlignmentContext> getFilteredAndStratifiedContexts(
UnifiedArgumentCollection UAC,
ReferenceContext refContext,
AlignmentContext rawContext,
final GenotypeLikelihoodsCalculationModel.Model model) {
if (!BaseUtils.isRegularBase(refContext.getBase())) return null;
Map<String, AlignmentContext> stratifiedContexts = null;
if (model.name().contains("INDEL")) {
final ReadBackedPileup pileup =
rawContext.getBasePileup().getMappingFilteredPileup(UAC.MIN_BASE_QUALTY_SCORE);
// don't call when there is no coverage
if (pileup.getNumberOfElements() == 0 && UAC.OutputMode != OUTPUT_MODE.EMIT_ALL_SITES)
return null;
// stratify the AlignmentContext and cut by sample
stratifiedContexts = AlignmentContextUtils.splitContextBySampleName(pileup);
} else if (model.name().contains("SNP")) {
// stratify the AlignmentContext and cut by sample
stratifiedContexts =
AlignmentContextUtils.splitContextBySampleName(rawContext.getBasePileup());
if (!(UAC.OutputMode == OUTPUT_MODE.EMIT_ALL_SITES
&& UAC.GenotypingMode
!= GenotypeLikelihoodsCalculationModel.GENOTYPING_MODE.GENOTYPE_GIVEN_ALLELES)) {
int numDeletions = 0;
for (final PileupElement p : rawContext.getBasePileup()) {
if (p.isDeletion()) numDeletions++;
}
if (((double) numDeletions) / ((double) rawContext.getBasePileup().getNumberOfElements())
> UAC.MAX_DELETION_FRACTION) {
return null;
}
}
}
return stratifiedContexts;
}
private Map<String, Object> annotateSNP(AlignmentContext stratifiedContext, VariantContext vc) {
if (!stratifiedContext.hasBasePileup()) return null;
HashMap<Byte, Integer> alleleCounts = new HashMap<Byte, Integer>();
for (Allele allele : vc.getAlternateAlleles()) alleleCounts.put(allele.getBases()[0], 0);
ReadBackedPileup pileup = stratifiedContext.getBasePileup();
int totalDepth = pileup.size();
Map<String, Object> map = new HashMap<String, Object>();
map.put(getKeyNames().get(0), totalDepth); // put total depth in right away
if (totalDepth == 0) return map; // done, can not compute FA at 0 coverage!!
int mq0 = 0; // number of "ref" reads that are acually mq0
for (PileupElement p : pileup) {
if (p.getMappingQual() == 0) {
mq0++;
continue;
}
if (alleleCounts.containsKey(p.getBase())) // non-mq0 read and it's an alt
alleleCounts.put(p.getBase(), alleleCounts.get(p.getBase()) + 1);
}
if (mq0 == totalDepth) return map; // if all reads are mq0, there is nothing left to do
// we need to add counts in the correct order
String[] fracs = new String[alleleCounts.size()];
for (int i = 0; i < vc.getAlternateAlleles().size(); i++) {
fracs[i] =
String.format(
"%.3f",
((float) alleleCounts.get(vc.getAlternateAllele(i).getBases()[0]))
/ (totalDepth - mq0));
}
map.put(getKeyNames().get(1), fracs);
return map;
}
@Test
public void testXandEQOperators() {
final byte[] bases1 = new byte[] {'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A'};
final byte[] bases2 = new byte[] {'A', 'A', 'A', 'C', 'A', 'A', 'A', 'A', 'A', 'C'};
// create a test version of the Reads object
ReadProperties readAttributes = createTestReadProperties();
SAMRecord r1 = ArtificialSAMUtils.createArtificialRead(header, "r1", 0, 1, 10);
r1.setReadBases(bases1);
r1.setBaseQualities(new byte[] {20, 20, 20, 20, 20, 20, 20, 20, 20, 20});
r1.setCigarString("10M");
SAMRecord r2 = ArtificialSAMUtils.createArtificialRead(header, "r2", 0, 1, 10);
r2.setReadBases(bases2);
r2.setBaseQualities(new byte[] {20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20});
r2.setCigarString("3=1X5=1X");
SAMRecord r3 = ArtificialSAMUtils.createArtificialRead(header, "r3", 0, 1, 10);
r3.setReadBases(bases2);
r3.setBaseQualities(new byte[] {20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20});
r3.setCigarString("3=1X5M1X");
SAMRecord r4 = ArtificialSAMUtils.createArtificialRead(header, "r4", 0, 1, 10);
r4.setReadBases(bases2);
r4.setBaseQualities(new byte[] {20, 20, 20, 20, 20, 20, 20, 20, 20, 20});
r4.setCigarString("10M");
List<SAMRecord> reads = Arrays.asList(r1, r2, r3, r4);
// create the iterator by state with the fake reads and fake records
li = makeLTBS(reads, readAttributes);
while (li.hasNext()) {
AlignmentContext context = li.next();
ReadBackedPileup pileup = context.getBasePileup();
Assert.assertEquals(pileup.depthOfCoverage(), 4);
}
}
private static String createVerboseOutput(final ReadBackedPileup pileup) {
final StringBuilder sb = new StringBuilder();
boolean isFirst = true;
sb.append(pileup.getNumberOfDeletions());
sb.append(" ");
for (PileupElement p : pileup) {
if (isFirst) isFirst = false;
else sb.append(",");
sb.append(p.getRead().getReadName());
sb.append(verboseDelimiter);
sb.append(p.getOffset());
sb.append(verboseDelimiter);
sb.append(p.getRead().getReadLength());
sb.append(verboseDelimiter);
sb.append(p.getRead().getMappingQuality());
}
return sb.toString();
}
/**
* Test to make sure that reads supporting only an indel (example cigar string: 76I) are
* represented properly
*/
@Test
public void testWholeIndelReadRepresentedTest() {
final int firstLocus = 44367788, secondLocus = firstLocus + 1;
SAMRecord read1 = ArtificialSAMUtils.createArtificialRead(header, "read1", 0, secondLocus, 1);
read1.setReadBases(Utils.dupBytes((byte) 'A', 1));
read1.setBaseQualities(Utils.dupBytes((byte) '@', 1));
read1.setCigarString("1I");
List<SAMRecord> reads = Arrays.asList(read1);
// create the iterator by state with the fake reads and fake records
li = makeLTBS(reads, createTestReadProperties());
while (li.hasNext()) {
AlignmentContext alignmentContext = li.next();
ReadBackedPileup p = alignmentContext.getBasePileup();
Assert.assertTrue(p.getNumberOfElements() == 1);
PileupElement pe = p.iterator().next();
Assert.assertTrue(pe.isBeforeInsertion());
Assert.assertFalse(pe.isAfterInsertion());
Assert.assertEquals(pe.getEventBases(), "A");
}
SAMRecord read2 = ArtificialSAMUtils.createArtificialRead(header, "read2", 0, secondLocus, 10);
read2.setReadBases(Utils.dupBytes((byte) 'A', 10));
read2.setBaseQualities(Utils.dupBytes((byte) '@', 10));
read2.setCigarString("10I");
reads = Arrays.asList(read2);
// create the iterator by state with the fake reads and fake records
li = makeLTBS(reads, createTestReadProperties());
while (li.hasNext()) {
AlignmentContext alignmentContext = li.next();
ReadBackedPileup p = alignmentContext.getBasePileup();
Assert.assertTrue(p.getNumberOfElements() == 1);
PileupElement pe = p.iterator().next();
Assert.assertTrue(pe.isBeforeInsertion());
Assert.assertFalse(pe.isAfterInsertion());
Assert.assertEquals(pe.getEventBases(), "AAAAAAAAAA");
}
}