@Test
public void testLastToCompact1() throws IOException {
// test LastToCompact convert+filtering
// read fake data and convert+filter
final LastToCompactMode processor = new LastToCompactMode();
final int LAST_TO_COMPACT_M_PARAM = 2;
processor.setAmbiguityThreshold(LAST_TO_COMPACT_M_PARAM);
processor.setInputFile("test-results/alignments/last-to-compact/last-101.maf");
processor.setOutputFile("test-results/alignments/last-to-compact/last-101.compact");
processor.setTargetReferenceIdsFilename(
"test-results/alignments/last-to-compact/last-reference.compact-reads");
processor.setOnlyMafFile(true);
processor.setNumberOfReads(2857819);
processor.setLargestQueryIndex(2857819);
processor.setSmallestQueryIndex(0);
processor.setPropagateQueryIds(false);
processor.setPropagateTargetIds(true);
processor.execute();
// read compact alignment results
final AlignmentReaderImpl reader = new AlignmentReaderImpl(processor.getOutputFile());
reader.readHeader();
assertEquals(2857819, reader.getNumberOfQueries());
assertEquals(1, reader.getNumberOfTargets());
assertTrue(reader.hasQueryIndexOccurrences());
// lookup tables
final Int2IntOpenHashMap queryIndex2NumberOfHits = new Int2IntOpenHashMap();
final Int2FloatOpenHashMap queryIndex2Score = new Int2FloatOpenHashMap();
final Int2IntOpenHashMap queryIndex2Multiplicity = new Int2IntOpenHashMap();
final Int2IntOpenHashMap queryIndex2NumberOfIndels = new Int2IntOpenHashMap();
final Int2IntOpenHashMap queryIndex2NumberOfMismatches = new Int2IntOpenHashMap();
final Int2IntOpenHashMap queryIndex2Position = new Int2IntOpenHashMap();
final Int2IntOpenHashMap queryIndex2QueryAlignedLength = new Int2IntOpenHashMap();
final Int2IntOpenHashMap queryIndex2QueryPosition = new Int2IntOpenHashMap();
final Int2IntOpenHashMap queryIndex2TargetIndex = new Int2IntOpenHashMap();
final Int2BooleanOpenHashMap queryIndex2MatchingReverseStrand = new Int2BooleanOpenHashMap();
final Int2IntOpenHashMap queryIndex2TargetAlignedLength = new Int2IntOpenHashMap();
final Int2IntOpenHashMap queryIndex2QueryIndexOcc = new Int2IntOpenHashMap();
final Int2IntOpenHashMap queryIndex2Ambiguity = new Int2IntOpenHashMap();
// enter alignment data
int qii;
while (reader.hasNext()) {
final Alignments.AlignmentEntry aln = reader.next();
qii = aln.getQueryIndex();
final int numHits = queryIndex2NumberOfHits.get(qii);
queryIndex2NumberOfHits.put(qii, numHits + 1);
queryIndex2Score.put(qii, aln.getScore());
queryIndex2Multiplicity.put(qii, aln.getMultiplicity());
queryIndex2NumberOfIndels.put(qii, aln.getNumberOfIndels());
queryIndex2NumberOfMismatches.put(qii, aln.getNumberOfMismatches());
queryIndex2Position.put(qii, aln.getPosition());
queryIndex2QueryAlignedLength.put(qii, aln.getQueryAlignedLength());
queryIndex2QueryPosition.put(qii, aln.getQueryPosition());
queryIndex2TargetIndex.put(qii, aln.getTargetIndex());
queryIndex2MatchingReverseStrand.put(qii, aln.getMatchingReverseStrand());
queryIndex2TargetAlignedLength.put(qii, aln.getTargetAlignedLength());
queryIndex2QueryIndexOcc.put(qii, aln.getQueryIndexOccurrences());
queryIndex2Ambiguity.put(qii, aln.getAmbiguity());
}
//
// validate alignment data using values from getMafInput() below
//
// there are a total of 5 entries with ID 2857818
// 2 entries have score = 35 (the maximum score for this ID)
// 3 entries are filtered b/c their scores are below 35
qii = 2857818;
assertEquals(queryIndex2NumberOfHits.get(qii), 2);
assertEquals((int) queryIndex2Score.get(qii), 35);
assertEquals(queryIndex2Multiplicity.get(qii), 1);
assertEquals(queryIndex2NumberOfIndels.get(qii), 0);
assertEquals(queryIndex2NumberOfMismatches.get(qii), 0);
assertEquals(queryIndex2Position.get(qii), 1614); // last entry added
assertEquals(queryIndex2QueryAlignedLength.get(qii), 35);
assertEquals(queryIndex2QueryPosition.get(qii), 0);
assertEquals(queryIndex2TargetIndex.get(qii), 0);
assertEquals(queryIndex2MatchingReverseStrand.get(qii), false);
assertEquals(queryIndex2TargetAlignedLength.get(qii), 35);
assertEquals(2, queryIndex2QueryIndexOcc.get(qii));
assertEquals(2, queryIndex2Ambiguity.get(qii));
// there are 5 entries with the score = 35 (the maximum score for this ID)
// filtered due to ambiguity
qii = 577287;
assertEquals(queryIndex2NumberOfHits.get(qii), 0);
assertEquals((int) queryIndex2Score.get(qii), 0);
assertEquals(queryIndex2Multiplicity.get(qii), 0);
assertEquals(queryIndex2NumberOfIndels.get(qii), 0);
assertEquals(queryIndex2NumberOfMismatches.get(qii), 0);
assertEquals(queryIndex2Position.get(qii), 0);
assertEquals(queryIndex2QueryAlignedLength.get(qii), 0);
assertEquals(queryIndex2QueryPosition.get(qii), 0);
assertEquals(queryIndex2TargetIndex.get(qii), 0);
assertEquals(queryIndex2MatchingReverseStrand.get(qii), false);
assertEquals(queryIndex2TargetAlignedLength.get(qii), 0);
assertEquals(0, queryIndex2QueryIndexOcc.get(qii));
assertEquals(0, queryIndex2Ambiguity.get(qii));
//
reader.close();
}
/**
* Construct alignment blocks from the Goby alignment entry. This method uses the convention that
* '=' denotes a match to the reference.
*
* <p>Conventions for storing sequence variations in Goby alignments are described <a
* href="http://tinyurl.com/goby-sequence-variations">here</a>
*
* @param alignmentEntry The Goby alignment entry to use
*/
public void buildBlocks(Alignments.AlignmentEntry alignmentEntry) {
ObjectArrayList<AlignmentBlock> blocks = new ObjectArrayList<AlignmentBlock>();
ObjectArrayList<AlignmentBlock> insertionBlocks = new ObjectArrayList<AlignmentBlock>();
int start = alignmentEntry.getPosition();
ByteArrayList bases = new ByteArrayList();
ByteArrayList scores = new ByteArrayList();
int readLength = alignmentEntry.getQueryLength();
byte[] readBases = new byte[readLength];
byte[] readQual = new byte[readLength];
Arrays.fill(readBases, (byte) '=');
if (alignmentEntry.hasReadQualityScores()) {
readQual = alignmentEntry.getReadQualityScores().toByteArray();
} else {
Arrays.fill(readQual, (byte) 40);
}
int j = 0;
int insertedBases = 0;
int deletedBases = 0;
final int leftPadding = alignmentEntry.getQueryPosition();
boolean showSoftClipped =
PreferenceManager.getInstance().getAsBoolean(PreferenceManager.SAM_SHOW_SOFT_CLIPPED);
if (showSoftClipped && entry.hasSoftClippedBasesLeft()) {
int clipLength = entry.getSoftClippedBasesLeft().length();
addSoftClipBlock(
blocks,
Math.max(0, entry.getPosition() - clipLength),
entry.getSoftClippedBasesLeft(),
readQual,
entry.hasSoftClippedQualityLeft(),
entry.getSoftClippedQualityLeft().toByteArray(),
0);
}
for (Alignments.SequenceVariation var : alignmentEntry.getSequenceVariationsList()) {
final String from = var.getFrom();
final int fromLength = from.length();
final String to = var.getTo();
final int toLength = from.length();
final int sequenceVariationLength = Math.max(fromLength, toLength);
final ByteString toQuality = var.getToQuality();
if (hasReadInsertion(from)) {
bases.clear();
scores.clear();
for (int i = 0; i < sequenceVariationLength; i++) {
final char toChar = i >= toLength ? '-' : to.charAt(i);
int size = toQuality.size();
final byte qual = size > 0 && i < size ? toQuality.byteAt(i) : 40;
bases.add((byte) toChar);
scores.add(qual);
deletedBases++;
}
addBlock(insertionBlocks, alignmentEntry.getPosition() + var.getPosition(), bases, scores);
bases.clear();
scores.clear();
} else if (!to.contains("-")) {
for (int i = 0; i < toLength; i++) {
final int offset = j + var.getPosition() + i - 1 + leftPadding - insertedBases;
if (offset > 0 && offset < readBases.length) {
readBases[offset] = (byte) to.charAt(i);
if (i < toQuality.size()) {
readQual[offset] = toQuality.byteAt(i);
}
}
}
} else {
// has read deletion:
insertedBases++;
}
}
int pos = start;
int matchLength = alignmentEntry.getQueryAlignedLength() - deletedBases;
int endAlignmentRefPosition = matchLength + start;
bases.clear();
scores.clear();
int maxIndex = Math.min(readBases.length, readQual.length);
while (pos < endAlignmentRefPosition) {
final int index = pos - start + leftPadding;
if (index < maxIndex) {
bases.add(readBases[index]);
scores.add(readQual[index]);
} else {
break;
}
++pos;
}
addBlock(blocks, start, bases, scores);
blocks = introduceDeletions(blocks, entry);
if (showSoftClipped && entry.hasSoftClippedBasesRight()) {
int targetAlignedLength = entry.getTargetAlignedLength();
addSoftClipBlock(
blocks,
entry.getPosition() + targetAlignedLength,
entry.getSoftClippedBasesRight(),
readQual,
entry.hasSoftClippedQualityRight(),
entry.getSoftClippedQualityRight().toByteArray(),
entry.getQueryAlignedLength() + entry.getSoftClippedBasesLeft().length());
}
block = blocks.toArray(new AlignmentBlock[blocks.size()]);
Arrays.sort(block, blockComparator);
insertionBlock = insertionBlocks.toArray(new AlignmentBlock[insertionBlocks.size()]);
Arrays.sort(insertionBlock, blockComparator);
ObjectArrayList<GobyAlignment> list = null;
if (alignmentEntry.hasSplicedForwardAlignmentLink()
|| alignmentEntry.hasSplicedBackwardAlignmentLink()) {
// if has a forward link, store a reference to this alignment in the reader (which represents
// the window scope)
list = iterator.cacheSpliceComponent(this);
if (list.size() > 1 && spliceListIsValid(list)) {
final GobyAlignment spliceHeadAlignment = list.get(0);
ObjectArrayList<AlignmentBlock> splicedBlocks = new ObjectArrayList<AlignmentBlock>();
splicedBlocks.addAll(ObjectArrayList.wrap(spliceHeadAlignment.block));
splicedBlocks.addAll(blocks);
spliceHeadAlignment.block = splicedBlocks.toArray(new AlignmentBlock[splicedBlocks.size()]);
ObjectArrayList<AlignmentBlock> splicedInsertionBlocks =
new ObjectArrayList<AlignmentBlock>();
splicedInsertionBlocks.addAll(ObjectArrayList.wrap(spliceHeadAlignment.insertionBlock));
splicedInsertionBlocks.addAll(insertionBlocks);
spliceHeadAlignment.insertionBlock =
splicedInsertionBlocks.toArray(new AlignmentBlock[splicedInsertionBlocks.size()]);
if (spliceHeadAlignment.gapTypes == null) {
spliceHeadAlignment.gapTypes = new CharArrayList(10);
}
spliceHeadAlignment.gapTypes.add(SamAlignment.SKIPPED_REGION);
// Since the previous alignment carries this information, we clear up block and
// insertionBlock
// in this alignment, but keep any softClips:
this.block = keepSoftClips(block);
this.insertionBlock = new AlignmentBlock[0];
}
}
}
