/** Ensure that splitting read groups still works when dealing with null read groups. */
@Test
public void testSplitByNullReadGroups() {
SAMFileHeader header = ArtificialSAMUtils.createArtificialSamHeader(1, 1, 1000);
GATKSAMRecord read1 = ArtificialSAMUtils.createArtificialRead(header, "read1", 0, 1, 10);
GATKSAMRecord read2 = ArtificialSAMUtils.createArtificialRead(header, "read2", 0, 1, 10);
GATKSAMRecord read3 = ArtificialSAMUtils.createArtificialRead(header, "read3", 0, 1, 10);
ReadBackedPileup pileup =
new ReadBackedPileupImpl(null, Arrays.asList(read1, read2, read3), Arrays.asList(1, 1, 1));
ReadBackedPileup nullRgPileup = pileup.getPileupForReadGroup(null);
List<GATKSAMRecord> nullRgReads = nullRgPileup.getReads();
Assert.assertEquals(
nullRgPileup.getNumberOfElements(), 3, "Wrong number of reads in null read group");
Assert.assertEquals(
nullRgReads.get(0),
read1,
"Read " + read1.getReadName() + " should be in null rg but isn't");
Assert.assertEquals(
nullRgReads.get(1),
read2,
"Read " + read2.getReadName() + " should be in null rg but isn't");
Assert.assertEquals(
nullRgReads.get(2),
read3,
"Read " + read3.getReadName() + " should be in null rg but isn't");
ReadBackedPileup rg1Pileup = pileup.getPileupForReadGroup("rg1");
Assert.assertNull(rg1Pileup, "Pileup for non-existent read group should return null");
}
/** Ensure that splitting read groups still works when dealing with a sample-split pileup. */
@Test
public void testSplitBySample() {
SAMReadGroupRecord readGroupOne = new SAMReadGroupRecord("rg1");
readGroupOne.setSample("sample1");
SAMReadGroupRecord readGroupTwo = new SAMReadGroupRecord("rg2");
readGroupTwo.setSample("sample2");
SAMFileHeader header = ArtificialSAMUtils.createArtificialSamHeader(1, 1, 1000);
header.addReadGroup(readGroupOne);
header.addReadGroup(readGroupTwo);
GATKSAMRecord read1 = ArtificialSAMUtils.createArtificialRead(header, "read1", 0, 1, 10);
read1.setAttribute("RG", readGroupOne.getId());
GATKSAMRecord read2 = ArtificialSAMUtils.createArtificialRead(header, "read2", 0, 1, 10);
read2.setAttribute("RG", readGroupTwo.getId());
GATKSAMRecord read3 = ArtificialSAMUtils.createArtificialRead(header, "read3", 0, 1, 10);
read3.setAttribute("RG", readGroupOne.getId());
GATKSAMRecord read4 = ArtificialSAMUtils.createArtificialRead(header, "read4", 0, 1, 10);
read4.setAttribute("RG", readGroupTwo.getId());
ReadBackedPileupImpl sample1Pileup =
new ReadBackedPileupImpl(null, Arrays.asList(read1, read3), Arrays.asList(1, 1));
ReadBackedPileupImpl sample2Pileup =
new ReadBackedPileupImpl(null, Arrays.asList(read2, read4), Arrays.asList(1, 1));
Map<String, ReadBackedPileupImpl> sampleToPileupMap =
new HashMap<String, ReadBackedPileupImpl>();
sampleToPileupMap.put(readGroupOne.getSample(), sample1Pileup);
sampleToPileupMap.put(readGroupTwo.getSample(), sample2Pileup);
ReadBackedPileup compositePileup = new ReadBackedPileupImpl(null, sampleToPileupMap);
ReadBackedPileup rg1Pileup = compositePileup.getPileupForReadGroup("rg1");
List<GATKSAMRecord> rg1Reads = rg1Pileup.getReads();
Assert.assertEquals(rg1Reads.size(), 2, "Wrong number of reads in read group rg1");
Assert.assertEquals(
rg1Reads.get(0), read1, "Read " + read1.getReadName() + " should be in rg1 but isn't");
Assert.assertEquals(
rg1Reads.get(1), read3, "Read " + read3.getReadName() + " should be in rg1 but isn't");
ReadBackedPileup rg2Pileup = compositePileup.getPileupForReadGroup("rg2");
List<GATKSAMRecord> rg2Reads = rg2Pileup.getReads();
Assert.assertEquals(rg1Reads.size(), 2, "Wrong number of reads in read group rg2");
Assert.assertEquals(
rg2Reads.get(0), read2, "Read " + read2.getReadName() + " should be in rg2 but isn't");
Assert.assertEquals(
rg2Reads.get(1), read4, "Read " + read4.getReadName() + " should be in rg2 but isn't");
}
public double[] computeReadHaplotypeLikelihoods(
ReadBackedPileup pileup, HashMap<Allele, Haplotype> haplotypesInVC) {
double[][] haplotypeLikehoodMatrix = new double[haplotypesInVC.size()][haplotypesInVC.size()];
double readLikelihoods[][] = new double[pileup.getReads().size()][haplotypesInVC.size()];
int i = 0;
for (GATKSAMRecord read : pileup.getReads()) {
if (ReadUtils.is454Read(read)) {
continue;
}
// for each read/haplotype combination, compute likelihoods, ie -10*log10(Pr(R | Hi))
// = sum_j(-10*log10(Pr(R_j | Hi) since reads are assumed to be independent
int j = 0;
for (Map.Entry<Allele, Haplotype> a : haplotypesInVC.entrySet()) {
readLikelihoods[i][j] = computeReadLikelihoodGivenHaplotype(a.getValue(), read);
if (DEBUG) {
System.out.print(read.getReadName() + " ");
System.out.format(
"%d %d S:%d US:%d E:%d UE:%d C:%s %3.4f\n",
i,
j,
read.getAlignmentStart(),
read.getUnclippedStart(),
read.getAlignmentEnd(),
read.getUnclippedEnd(),
read.getCigarString(),
readLikelihoods[i][j]);
}
j++;
}
i++;
}
for (i = 0; i < haplotypesInVC.size(); i++) {
for (int j = i; j < haplotypesInVC.size(); j++) {
// combine likelihoods of haplotypeLikelihoods[i], haplotypeLikelihoods[j]
// L(Hi, Hj) = sum_reads ( Pr(R|Hi)/2 + Pr(R|Hj)/2)
// readLikelihoods[k][j] has log10(Pr(R_k) | H[j] )
double[] readLikelihood = new double[2]; // diploid sample
for (int readIdx = 0; readIdx < pileup.getReads().size(); readIdx++) {
readLikelihood[0] = -readLikelihoods[readIdx][i] / 10;
readLikelihood[1] = -readLikelihoods[readIdx][j] / 10;
// Compute log10(10^x1/2 + 10^x2/2) = log10(10^x1+x0^x2)-log10(2)
// First term is approximated by Jacobian log with table lookup.
// Second term is a constant added to both likelihoods so will be ignored
haplotypeLikehoodMatrix[i][j] +=
MathUtils.approximateLog10SumLog10(readLikelihood[0], readLikelihood[1]);
}
}
}
return getHaplotypeLikelihoods(haplotypeLikehoodMatrix);
}
/** Ensure that basic read group splitting works. */
@Test
public void testSplitByReadGroup() {
SAMReadGroupRecord readGroupOne = new SAMReadGroupRecord("rg1");
SAMReadGroupRecord readGroupTwo = new SAMReadGroupRecord("rg2");
SAMFileHeader header = ArtificialSAMUtils.createArtificialSamHeader(1, 1, 1000);
header.addReadGroup(readGroupOne);
header.addReadGroup(readGroupTwo);
GATKSAMRecord read1 = ArtificialSAMUtils.createArtificialRead(header, "read1", 0, 1, 10);
read1.setAttribute("RG", readGroupOne.getId());
GATKSAMRecord read2 = ArtificialSAMUtils.createArtificialRead(header, "read2", 0, 1, 10);
read2.setAttribute("RG", readGroupTwo.getId());
GATKSAMRecord read3 = ArtificialSAMUtils.createArtificialRead(header, "read3", 0, 1, 10);
read3.setAttribute("RG", readGroupOne.getId());
GATKSAMRecord read4 = ArtificialSAMUtils.createArtificialRead(header, "read4", 0, 1, 10);
read4.setAttribute("RG", readGroupTwo.getId());
GATKSAMRecord read5 = ArtificialSAMUtils.createArtificialRead(header, "read5", 0, 1, 10);
read5.setAttribute("RG", readGroupTwo.getId());
GATKSAMRecord read6 = ArtificialSAMUtils.createArtificialRead(header, "read6", 0, 1, 10);
read6.setAttribute("RG", readGroupOne.getId());
GATKSAMRecord read7 = ArtificialSAMUtils.createArtificialRead(header, "read7", 0, 1, 10);
read7.setAttribute("RG", readGroupOne.getId());
ReadBackedPileup pileup =
new ReadBackedPileupImpl(
null,
Arrays.asList(read1, read2, read3, read4, read5, read6, read7),
Arrays.asList(1, 1, 1, 1, 1, 1, 1));
ReadBackedPileup rg1Pileup = pileup.getPileupForReadGroup("rg1");
List<GATKSAMRecord> rg1Reads = rg1Pileup.getReads();
Assert.assertEquals(rg1Reads.size(), 4, "Wrong number of reads in read group rg1");
Assert.assertEquals(
rg1Reads.get(0), read1, "Read " + read1.getReadName() + " should be in rg1 but isn't");
Assert.assertEquals(
rg1Reads.get(1), read3, "Read " + read3.getReadName() + " should be in rg1 but isn't");
Assert.assertEquals(
rg1Reads.get(2), read6, "Read " + read6.getReadName() + " should be in rg1 but isn't");
Assert.assertEquals(
rg1Reads.get(3), read7, "Read " + read7.getReadName() + " should be in rg1 but isn't");
ReadBackedPileup rg2Pileup = pileup.getPileupForReadGroup("rg2");
List<GATKSAMRecord> rg2Reads = rg2Pileup.getReads();
Assert.assertEquals(rg2Reads.size(), 3, "Wrong number of reads in read group rg2");
Assert.assertEquals(
rg2Reads.get(0), read2, "Read " + read2.getReadName() + " should be in rg2 but isn't");
Assert.assertEquals(
rg2Reads.get(1), read4, "Read " + read4.getReadName() + " should be in rg2 but isn't");
Assert.assertEquals(
rg2Reads.get(2), read5, "Read " + read5.getReadName() + " should be in rg2 but isn't");
}
/**
* Clips the bases in read according to this operation's start and stop. Uses the clipping
* representation used is the one provided by algorithm argument.
*
* @param algorithm clipping algorithm to use
* @param originalRead the read to be clipped
*/
public GATKSAMRecord apply(ClippingRepresentation algorithm, GATKSAMRecord originalRead) {
GATKSAMRecord read = (GATKSAMRecord) originalRead.clone();
byte[] quals = read.getBaseQualities();
byte[] bases = read.getReadBases();
byte[] newBases = new byte[bases.length];
byte[] newQuals = new byte[quals.length];
switch (algorithm) {
// important note:
// it's not safe to call read.getReadBases()[i] = 'N' or read.getBaseQualities()[i] = 0
// because you're not guaranteed to get a pointer to the actual array of bytes in the
// GATKSAMRecord
case WRITE_NS:
for (int i = 0; i < bases.length; i++) {
if (i >= start && i <= stop) {
newBases[i] = 'N';
} else {
newBases[i] = bases[i];
}
}
read.setReadBases(newBases);
break;
case WRITE_Q0S:
for (int i = 0; i < quals.length; i++) {
if (i >= start && i <= stop) {
newQuals[i] = 0;
} else {
newQuals[i] = quals[i];
}
}
read.setBaseQualities(newQuals);
break;
case WRITE_NS_Q0S:
for (int i = 0; i < bases.length; i++) {
if (i >= start && i <= stop) {
newQuals[i] = 0;
newBases[i] = 'N';
} else {
newQuals[i] = quals[i];
newBases[i] = bases[i];
}
}
read.setBaseQualities(newBases);
read.setReadBases(newBases);
break;
case HARDCLIP_BASES:
read = hardClip(read, start, stop);
break;
case SOFTCLIP_BASES:
if (read.getReadUnmappedFlag()) {
// we can't process unmapped reads
throw new UserException("Read Clipper cannot soft clip unmapped reads");
}
// System.out.printf("%d %d %d%n", stop, start, read.getReadLength());
int myStop = stop;
if ((stop + 1 - start) == read.getReadLength()) {
// BAM representation issue -- we can't SOFTCLIP away all bases in a read, just leave it
// alone
// Walker.logger.info(String.format("Warning, read %s has all bases clip but this can't be
// represented with SOFTCLIP_BASES, just leaving it alone", read.getReadName()));
// break;
myStop--; // just decrement stop
}
if (start > 0 && myStop != read.getReadLength() - 1)
throw new RuntimeException(
String.format(
"Cannot apply soft clipping operator to the middle of a read: %s to be clipped at %d-%d",
read.getReadName(), start, myStop));
Cigar oldCigar = read.getCigar();
int scLeft = 0, scRight = read.getReadLength();
if (start == 0) scLeft = myStop + 1;
else scRight = start;
Cigar newCigar = softClip(oldCigar, scLeft, scRight);
read.setCigar(newCigar);
int newClippedStart = getNewAlignmentStartOffset(newCigar, oldCigar);
int newStart = read.getAlignmentStart() + newClippedStart;
read.setAlignmentStart(newStart);
break;
case REVERT_SOFTCLIPPED_BASES:
read = revertSoftClippedBases(read);
break;
default:
throw new IllegalStateException("Unexpected Clipping operator type " + algorithm);
}
return read;
}
