// note: takes smart pointer as it maybe added to the results
static hal_size_t mapUp(MappedSegmentPtr mappedSeg, list<MappedSegmentPtr> &results, bool doDupes, hal_size_t minLength) {
const Genome *parent = mappedSeg->getGenome()->getParent();
assert(parent != NULL);
hal_size_t added = 0;
if (mappedSeg->isTop() == true) {
BottomSegmentIteratorPtr botSegIt = parent->getBottomSegmentIterator();
TopSegmentIteratorPtr topSegIt = mappedSeg->targetAsTop();
if (topSegIt->tseg()->hasParent() == true && topSegIt->getLength() >= minLength &&
(doDupes == true || topSegIt->tseg()->isCanonicalParalog() == true)) {
botSegIt->toParent(topSegIt);
mappedSeg->setTarget(std::dynamic_pointer_cast<SegmentIterator>(botSegIt));
results.push_back(mappedSeg);
++added;
}
} else {
hal_index_t rightCutoff = mappedSeg->getEndPosition();
BottomSegmentIteratorPtr botSegIt = mappedSeg->targetAsBottom();
hal_index_t startOffset = (hal_index_t)botSegIt->getStartOffset();
hal_index_t endOffset = (hal_index_t)botSegIt->getEndOffset();
TopSegmentIteratorPtr topSegIt = mappedSeg->getGenome()->getTopSegmentIterator();
topSegIt->toParseUp(botSegIt);
do {
TopSegmentIteratorPtr newTopSegIt = topSegIt->clone();
// we map the new target back to see how the offsets have
// changed. these changes are then applied to the source segment
// as deltas
BottomSegmentIteratorPtr backBotSegIt = botSegIt->clone();
backBotSegIt->toParseDown(newTopSegIt);
hal_index_t startBack = (hal_index_t)backBotSegIt->getStartOffset();
hal_index_t endBack = (hal_index_t)backBotSegIt->getEndOffset();
assert(startBack >= startOffset);
assert(endBack >= endOffset);
SegmentIteratorPtr newSourceSegIt = mappedSeg->sourceClone();
hal_index_t startDelta = startBack - startOffset;
hal_index_t endDelta = endBack - endOffset;
assert((hal_index_t)newSourceSegIt->getLength() > startDelta + endDelta);
newSourceSegIt->slice(newSourceSegIt->getStartOffset() + startDelta, newSourceSegIt->getEndOffset() + endDelta);
MappedSegmentPtr newMappedSeg(new MappedSegment(newSourceSegIt, newTopSegIt));
assert(newMappedSeg->isTop() == true);
assert(newMappedSeg->getSource()->getGenome() == mappedSeg->getSource()->getGenome());
added += mapUp(newMappedSeg, results, doDupes, minLength);
// stupid that we have to make this check but odn't want to
// make fundamental api change now
if (topSegIt->getEndPosition() != rightCutoff) {
topSegIt->toRight(rightCutoff);
} else {
break;
}
} while (true);
}
return added;
}
// Set top segments to be equal width and so that segment 1, 2, 3,
// etc. corresponds to parent segment 1, 2, 3, etc.
void setTopSegments(Genome *genome, hal_size_t width) {
TopSegmentIteratorPtr topIt = genome->getTopSegmentIterator();
hal_size_t n = genome->getNumTopSegments();
hal_index_t startPos = 0;
for (; topIt->getArrayIndex() < n; topIt->toRight(), startPos += width) {
topIt->setCoordinates(startPos, width);
topIt->tseg()->setParentIndex(topIt->getArrayIndex());
topIt->tseg()->setParentReversed(false);
topIt->tseg()->setBottomParseIndex(NULL_INDEX);
topIt->tseg()->setNextParalogyIndex(NULL_INDEX);
}
}
int MappedSegment::boundComp(const SegmentIteratorPtr &s1, const SegmentIteratorPtr &s2) {
int res = 0;
bool flip = s2->getReversed();
if (flip) {
s2->toReverse();
}
if (s1->isTop() && !s2->isTop()) {
BottomSegmentIteratorPtr bot = std::dynamic_pointer_cast<BottomSegmentIterator>(s2);
hal_index_t lb = bot->bseg()->getTopParseIndex();
hal_index_t ub = lb;
if ((hal_size_t)bot->getArrayIndex() < bot->getGenome()->getNumBottomSegments() - 1) {
bot = bot->clone();
bot->slice(0, 0);
bot->toRight();
ub = bot->bseg()->getTopParseIndex();
}
if (s1->getArrayIndex() < lb) {
res = -1;
} else if (s1->getArrayIndex() > ub) {
res = 1;
}
} else if (!s1->isTop() && s2->isTop()) {
TopSegmentIteratorPtr top = std::dynamic_pointer_cast<TopSegmentIterator>(s2);
hal_index_t lb = top->tseg()->getBottomParseIndex();
hal_index_t ub = lb;
if ((hal_size_t)top->getArrayIndex() < top->getGenome()->getNumTopSegments() - 1) {
top = top->clone();
top->slice(0, 0);
top->toRight();
ub = top->tseg()->getBottomParseIndex();
}
if (s1->getArrayIndex() < lb) {
res = -1;
} else if (s1->getArrayIndex() > ub) {
res = 1;
}
}
if (flip) {
s2->toReverse();
}
return res;
}
void GappedBottomSegmentIterator::toRightNextUngapped(TopSegmentIteratorPtr topSeqIt) const {
while (topSeqIt->tseg()->hasParent() == false && topSeqIt->getLength() <= _gapThreshold) {
if ((!topSeqIt->getReversed() && topSeqIt->getTopSegment()->isLast()) ||
(topSeqIt->getReversed() && topSeqIt->getTopSegment()->isFirst())) {
break;
}
topSeqIt->toRight();
}
}
void GenomeCopyTest::checkCallBack(const Alignment *alignment) {
// FIXME: halAlignment->open() fails miserably but
// openHalAlignmentReadOnly works? Probably some state isn't cleared
// on close.
AlignmentPtr tmp(getTestAlignmentInstances(alignment->getStorageFormat(), _path, WRITE_ACCESS));
_secondAlignment = tmp;
const Genome *ancGenome = alignment->openGenome("AncGenome");
CuAssertTrue(_testCase, ancGenome->getName() == "AncGenome");
CuAssertTrue(_testCase, ancGenome->getSequenceLength() == 1000000);
CuAssertTrue(_testCase, ancGenome->getNumTopSegments() == 0);
CuAssertTrue(_testCase, ancGenome->getNumBottomSegments() == 700000);
const MetaData *ancMeta = ancGenome->getMetaData();
CuAssertTrue(_testCase, ancMeta->get("Young") == "Jeezy");
const Genome *leafGenome = alignment->openGenome("LeafGenome1");
string ancSeq = "CAT";
hal_index_t n = ancGenome->getSequenceLength();
DnaIteratorPtr dnaIt = ancGenome->getDnaIterator();
for (; dnaIt->getArrayIndex() < n; dnaIt->toRight()) {
size_t i = dnaIt->getArrayIndex() % ancSeq.size();
CuAssertTrue(_testCase, dnaIt->getBase() == ancSeq[i]);
}
TopSegmentIteratorPtr topIt = leafGenome->getTopSegmentIterator();
n = leafGenome->getNumTopSegments();
for (; topIt->getArrayIndex() < n; topIt->toRight()) {
CuAssertTrue(_testCase, topIt->getStartPosition() == topIt->getArrayIndex());
CuAssertTrue(_testCase, topIt->getLength() == 1);
CuAssertTrue(_testCase, topIt->tseg()->getParentIndex() == 3);
CuAssertTrue(_testCase, topIt->tseg()->getParentReversed() == true);
CuAssertTrue(_testCase, topIt->tseg()->getBottomParseIndex() == 5);
if (topIt->getArrayIndex() != 6) {
CuAssertTrue(_testCase, topIt->tseg()->getNextParalogyIndex() == 6);
} else {
CuAssertTrue(_testCase, topIt->tseg()->getNextParalogyIndex() == 7);
}
}
BottomSegmentIteratorPtr botIt = ancGenome->getBottomSegmentIterator();
n = ancGenome->getNumBottomSegments();
for (; botIt->getArrayIndex() < n; botIt->toRight()) {
CuAssertTrue(_testCase, botIt->getStartPosition() == botIt->getArrayIndex());
CuAssertTrue(_testCase, botIt->getLength() == 1);
CuAssertTrue(_testCase, botIt->bseg()->getChildIndex(0) == 3);
CuAssertTrue(_testCase, botIt->bseg()->getChildReversed(0) == true);
CuAssertTrue(_testCase, botIt->bseg()->getTopParseIndex() == 5);
}
const Genome *copyRootGenome = _secondAlignment->openGenome("copyRootGenome");
const Genome *copyLeafGenome = _secondAlignment->openGenome("LeafGenome1");
CuAssertTrue(_testCase, copyRootGenome->getName() == "copyRootGenome");
CuAssertTrue(_testCase, copyRootGenome->getSequenceLength() == 1000000);
CuAssertTrue(_testCase, copyRootGenome->getNumTopSegments() == 0);
CuAssertTrue(_testCase, copyRootGenome->getNumBottomSegments() == 700000);
CuAssertTrue(_testCase, copyLeafGenome->getName() == "LeafGenome1");
CuAssertTrue(_testCase, copyLeafGenome->getSequenceLength() == 1000000);
CuAssertTrue(_testCase, copyLeafGenome->getNumTopSegments() == 5000);
CuAssertTrue(_testCase, copyLeafGenome->getNumBottomSegments() == 0);
const MetaData *copyMeta = copyRootGenome->getMetaData();
CuAssertTrue(_testCase, copyMeta->get("Young") == "Jeezy");
n = copyRootGenome->getSequenceLength();
dnaIt = copyRootGenome->getDnaIterator();
for (; dnaIt->getArrayIndex() < n; dnaIt->toRight()) {
size_t i = dnaIt->getArrayIndex() % ancSeq.size();
CuAssertTrue(_testCase, dnaIt->getBase() == ancSeq[i]);
}
topIt = copyLeafGenome->getTopSegmentIterator();
n = copyLeafGenome->getNumTopSegments();
for (; topIt->getArrayIndex() < n; topIt->toRight()) {
CuAssertTrue(_testCase, topIt->getStartPosition() == topIt->getArrayIndex());
CuAssertTrue(_testCase, topIt->getLength() == 1);
CuAssertTrue(_testCase, topIt->tseg()->getParentIndex() == 3);
CuAssertTrue(_testCase, topIt->tseg()->getParentReversed() == true);
CuAssertTrue(_testCase, topIt->tseg()->getBottomParseIndex() == 5);
if (topIt->getArrayIndex() != 6) {
CuAssertTrue(_testCase, topIt->tseg()->getNextParalogyIndex() == 6);
} else {
CuAssertTrue(_testCase, topIt->tseg()->getNextParalogyIndex() == 7);
}
}
botIt = copyRootGenome->getBottomSegmentIterator();
n = copyRootGenome->getNumBottomSegments();
for (; botIt->getArrayIndex() < n; botIt->toRight()) {
CuAssertTrue(_testCase, botIt->getStartPosition() == botIt->getArrayIndex());
CuAssertTrue(_testCase, botIt->getLength() == 1);
CuAssertTrue(_testCase, botIt->bseg()->getChildIndex(0) == 3);
CuAssertTrue(_testCase, botIt->bseg()->getChildReversed(0) == true);
CuAssertTrue(_testCase, botIt->bseg()->getTopParseIndex() == 5);
}
_secondAlignment->close();
remove(_path.c_str());
}
void GenomeCopyTest::createCallBack(Alignment *alignment) {
hal_size_t alignmentSize = alignment->getNumGenomes();
CuAssertTrue(_testCase, alignmentSize == 0);
// Hacky: Need a different alignment to test copying the bottom
// segments correctly. (the names of a node's children are used
// when copying bottom segments, and two genomes can't have the same
// name in the same alignment)
_path = getTempFile();
_secondAlignment =
AlignmentPtr(getTestAlignmentInstances(alignment->getStorageFormat(), _path, WRITE_ACCESS | CREATE_ACCESS));
Genome *ancGenome = alignment->addRootGenome("AncGenome", 0);
Genome *leafGenome = alignment->addLeafGenome("LeafGenome1", "AncGenome", 0);
// This genome will test copyDimensions, copyTopSegments,
// copyBottomSegments, copySequence, copyMetadata
Genome *copyRootGenome = _secondAlignment->addRootGenome("copyRootGenome", 0);
Genome *copyLeafGenome = _secondAlignment->addLeafGenome("LeafGenome1", "copyRootGenome", 0);
MetaData *ancMeta = ancGenome->getMetaData();
ancMeta->set("Young", "Jeezy");
vector<Sequence::Info> seqVec(1);
seqVec[0] = Sequence::Info("Sequence", 1000000, 0, 700000);
ancGenome->setDimensions(seqVec);
seqVec[0] = Sequence::Info("Sequence", 1000000, 5000, 0);
leafGenome->setDimensions(seqVec);
string ancSeq = "CAT";
hal_index_t n = ancGenome->getSequenceLength();
DnaIteratorPtr dnaIt = ancGenome->getDnaIterator();
for (; dnaIt->getArrayIndex() < n; dnaIt->toRight()) {
size_t i = dnaIt->getArrayIndex() % ancSeq.size();
dnaIt->setBase(ancSeq[i]);
}
dnaIt->flush();
n = leafGenome->getSequenceLength();
dnaIt = leafGenome->getDnaIterator();
for (; dnaIt->getArrayIndex() < n; dnaIt->toRight()) {
size_t i = dnaIt->getArrayIndex() % ancSeq.size();
dnaIt->setBase(ancSeq[i]);
}
dnaIt->flush();
TopSegmentIteratorPtr topIt = leafGenome->getTopSegmentIterator();
n = leafGenome->getNumTopSegments();
for (; topIt->getArrayIndex() < n; topIt->toRight()) {
topIt->setCoordinates(topIt->getArrayIndex(), 1);
topIt->tseg()->setParentIndex(3);
topIt->tseg()->setParentReversed(true);
topIt->tseg()->setBottomParseIndex(5);
if (topIt->getArrayIndex() != 6) {
topIt->tseg()->setNextParalogyIndex(6);
} else {
topIt->tseg()->setNextParalogyIndex(7);
}
}
BottomSegmentIteratorPtr botIt = ancGenome->getBottomSegmentIterator();
n = ancGenome->getNumBottomSegments();
for (; botIt->getArrayIndex() < n; botIt->toRight()) {
botIt->setCoordinates(botIt->getArrayIndex(), 1);
botIt->bseg()->setChildIndex(0, 3);
botIt->bseg()->setChildReversed(0, true);
botIt->bseg()->setTopParseIndex(5);
}
seqVec[0] = Sequence::Info("Sequence", 3300, 0, 1100);
copyRootGenome->setDimensions(seqVec);
seqVec[0] = Sequence::Info("Sequence", 3300, 2200, 0);
copyLeafGenome->setDimensions(seqVec);
string copySeq = "TAG";
dnaIt = copyRootGenome->getDnaIterator();
n = copyRootGenome->getSequenceLength();
for (; dnaIt->getArrayIndex() < n; dnaIt->toRight()) {
size_t i = dnaIt->getArrayIndex() % copySeq.size();
dnaIt->setBase(copySeq[i]);
}
dnaIt->flush();
dnaIt = copyLeafGenome->getDnaIterator();
n = copyLeafGenome->getSequenceLength();
for (; dnaIt->getArrayIndex() < n; dnaIt->toRight()) {
size_t i = dnaIt->getArrayIndex() % copySeq.size();
dnaIt->setBase(copySeq[i]);
}
dnaIt->flush();
topIt = copyLeafGenome->getTopSegmentIterator();
n = copyLeafGenome->getNumTopSegments();
for (; topIt->getArrayIndex() < n; topIt->toRight()) {
topIt->setCoordinates(7, 8);
topIt->tseg()->setParentIndex(9);
topIt->tseg()->setParentReversed(false);
topIt->tseg()->setBottomParseIndex(11);
if (topIt->getArrayIndex() != 12) {
topIt->tseg()->setNextParalogyIndex(12);
} else {
topIt->tseg()->setNextParalogyIndex(7);
}
}
botIt = copyRootGenome->getBottomSegmentIterator();
n = copyRootGenome->getNumBottomSegments();
for (; botIt->getArrayIndex() < n; botIt->toRight()) {
botIt->setCoordinates(6, 7);
botIt->bseg()->setChildIndex(0, 8);
botIt->bseg()->setChildReversed(0, false);
botIt->bseg()->setTopParseIndex(10);
}
ancGenome->copy(copyRootGenome);
leafGenome->copy(copyLeafGenome);
_secondAlignment->close();
}
// note: takes smart pointer as it maybe added to the results
static hal_size_t mapSelf(MappedSegmentPtr mappedSeg, list<MappedSegmentPtr> &results, hal_size_t minLength) {
hal_size_t added = 0;
if (mappedSeg->isTop() == true) {
SegmentIteratorPtr target = mappedSeg->getTargetIteratorPtr();
SegmentIteratorPtr source = mappedSeg->getSourceIteratorPtr();
TopSegmentIteratorPtr top = std::dynamic_pointer_cast<TopSegmentIterator>(target);
TopSegmentIteratorPtr topCopy = top->clone();
do {
// FIXME: why isn't clone() polymorphic?
SegmentIteratorPtr newSource;
if (source->isTop()) {
newSource = std::dynamic_pointer_cast<TopSegmentIterator>(source)->clone();
} else {
newSource = std::dynamic_pointer_cast<BottomSegmentIterator>(source)->clone();
}
TopSegmentIteratorPtr newTop = topCopy->clone();
MappedSegmentPtr newMappedSeg(new MappedSegment(newSource, newTop));
assert(newMappedSeg->getGenome() == mappedSeg->getGenome());
assert(newMappedSeg->getSource()->getGenome() == mappedSeg->getSource()->getGenome());
results.push_back(newMappedSeg);
++added;
if (topCopy->tseg()->hasNextParalogy()) {
topCopy->toNextParalogy();
}
} while (topCopy->tseg()->hasNextParalogy() == true && topCopy->getLength() >= minLength &&
topCopy->getArrayIndex() != top->getArrayIndex());
} else if (mappedSeg->getGenome()->getParent() != NULL) {
hal_index_t rightCutoff = mappedSeg->getEndPosition();
BottomSegmentIteratorPtr bottom = mappedSeg->targetAsBottom();
hal_index_t startOffset = (hal_index_t)bottom->getStartOffset();
hal_index_t endOffset = (hal_index_t)bottom->getEndOffset();
TopSegmentIteratorPtr top = mappedSeg->getGenome()->getTopSegmentIterator();
top->toParseUp(bottom);
do {
TopSegmentIteratorPtr topNew = top->clone();
// we map the new target back to see how the offsets have
// changed. these changes are then applied to the source segment
// as deltas
BottomSegmentIteratorPtr bottomBack = bottom->clone();
bottomBack->toParseDown(topNew);
hal_index_t startBack = (hal_index_t)bottomBack->getStartOffset();
hal_index_t endBack = (hal_index_t)bottomBack->getEndOffset();
assert(startBack >= startOffset);
assert(endBack >= endOffset);
SegmentIteratorPtr newSource = mappedSeg->sourceClone();
hal_index_t startDelta = startBack - startOffset;
hal_index_t endDelta = endBack - endOffset;
assert((hal_index_t)newSource->getLength() > startDelta + endDelta);
newSource->slice(newSource->getStartOffset() + startDelta, newSource->getEndOffset() + endDelta);
MappedSegmentPtr newMappedSeg(new MappedSegment(newSource, topNew));
assert(newMappedSeg->isTop() == true);
assert(newMappedSeg->getSource()->getGenome() == mappedSeg->getSource()->getGenome());
added += mapSelf(newMappedSeg, results, minLength);
// stupid that we have to make this check but odn't want to
// make fundamental api change now
if (top->getEndPosition() != rightCutoff) {
top->toRight(rightCutoff);
} else {
break;
}
} while (true);
}
return added;
}