void TopSegmentIsGapTest::createCallBack(Alignment *alignment) {
size_t numSequences = 3;
vector<Sequence::Info> seqVec(numSequences);
BottomSegmentIteratorPtr bi;
BottomSegmentStruct bs;
TopSegmentIteratorPtr ti;
TopSegmentStruct ts;
Genome *parent1 = alignment->addRootGenome("parent1");
Genome *child1 = alignment->addLeafGenome("child1", "parent1", 1);
// set up two genomes. each with three sequences. each sequence
// with 5 segments of length two. start with segment i in parent
// aligned with segment i in child.
for (size_t i = 0; i < numSequences; ++i) {
string name = "Sequence" + std::to_string(i);
seqVec[i] = Sequence::Info(name, 10, 5, 5);
}
parent1->setDimensions(seqVec);
child1->setDimensions(seqVec);
for (bi = parent1->getBottomSegmentIterator(); not bi->atEnd(); bi->toRight()) {
bs.set(bi->getBottomSegment()->getArrayIndex() * 2, 2);
bs._children.clear();
bs._children.push_back(pair<hal_size_t, bool>(bi->getBottomSegment()->getArrayIndex(), false));
bs.applyTo(bi);
}
for (ti = child1->getTopSegmentIterator(); not ti->atEnd(); ti->toRight()) {
ts.set(ti->getTopSegment()->getArrayIndex() * 2, 2, ti->getTopSegment()->getArrayIndex());
ts.applyTo(ti);
}
// insertion in middle (8th top segment)
bi = parent1->getBottomSegmentIterator(8);
ti = child1->getTopSegmentIterator(8);
assert(bi->getBottomSegment()->getChildIndex(0) == 8 && ti->getTopSegment()->getParentIndex() == 8);
bi->getBottomSegment()->setChildIndex(0, 9);
ti->getTopSegment()->setParentIndex(NULL_INDEX);
ti->toRight();
ti->getTopSegment()->setParentIndex(8);
// insertion at begining (10th top segment)
bi = parent1->getBottomSegmentIterator(10);
ti = child1->getTopSegmentIterator(10);
assert(bi->getBottomSegment()->getChildIndex(0) == 10 && ti->getTopSegment()->getParentIndex() == 10);
bi->getBottomSegment()->setChildIndex(0, 11);
ti->getTopSegment()->setParentIndex(NULL_INDEX);
ti->toRight();
ti->getTopSegment()->setParentIndex(10);
// just having a null parent is not enough for an insertion
bi = parent1->getBottomSegmentIterator(2);
ti = child1->getTopSegmentIterator(2);
assert(bi->getBottomSegment()->getChildIndex(0) == 2 && ti->getTopSegment()->getParentIndex() == 2);
ti->getTopSegment()->setParentIndex(NULL_INDEX);
}
void MappedSegmentMapDupeTest::createCallBack(AlignmentPtr alignment)
{
vector<Sequence::Info> seqVec(1);
BottomSegmentIteratorPtr bi;
BottomSegmentStruct bs;
TopSegmentIteratorPtr ti;
TopSegmentStruct ts;
// setup simple case were there is an edge from a parent to
// child and it is reversed
Genome* parent = alignment->addRootGenome("parent");
Genome* child1 = alignment->addLeafGenome("child1", "parent", 1);
Genome* child2 = alignment->addLeafGenome("child2", "parent", 1);
seqVec[0] = Sequence::Info("Sequence", 3, 0, 1);
parent->setDimensions(seqVec);
seqVec[0] = Sequence::Info("Sequence", 9, 3, 0);
child1->setDimensions(seqVec);
seqVec[0] = Sequence::Info("Sequence", 9, 3, 0);
child2->setDimensions(seqVec);
parent->setString("CCC");
child1->setString("CCCTACGTG");
child2->setString("CCCTACGTG");
bi = parent->getBottomSegmentIterator();
bs.set(0, 3);
bs._children.push_back(pair<hal_size_t, bool>(0, true));
bs._children.push_back(pair<hal_size_t, bool>(0, false));
bs.applyTo(bi);
ti = child1->getTopSegmentIterator();
ts.set(0, 3, 0, true, NULL_INDEX, 1);
ts.applyTo(ti);
ti->toRight();
ts.set(3, 3, 0, true, NULL_INDEX, 2);
ts.applyTo(ti);
ti->toRight();
ts.set(6, 3, 0, true, NULL_INDEX, 0);
ts.applyTo(ti);
ti = child2->getTopSegmentIterator();
ts.set(0, 3, 0, false);
ts.applyTo(ti);
ti->toRight();
ts.set(3, 3, NULL_INDEX, true);
ts.applyTo(ti);
ti->toRight();
ts.set(6, 3, NULL_INDEX, false);
ts.applyTo(ti);
}
void TopSegmentIteratorReverseTest::createCallBack(AlignmentPtr alignment)
{
vector<Sequence::Info> seqVec(1);
BottomSegmentIteratorPtr bi;
BottomSegmentStruct bs;
TopSegmentIteratorPtr ti;
TopSegmentStruct ts;
// setup simple case were there is an edge from a parent to
// child and it is reversed
Genome* parent1 = alignment->addRootGenome("parent1");
Genome* child1 = alignment->addLeafGenome("child1", "parent1", 1);
seqVec[0] = Sequence::Info("Sequence", 10, 2, 2);
parent1->setDimensions(seqVec);
seqVec[0] = Sequence::Info("Sequence", 10, 2, 2);
child1->setDimensions(seqVec);
parent1->setString("CCCTACGTGC");
child1->setString("CCCTACGTGC");
bi = parent1->getBottomSegmentIterator();
bs.set(0, 10, 0);
bs._children.push_back(pair<hal_size_t, bool>(0, true));
bs.applyTo(bi);
ti = child1->getTopSegmentIterator();
ts.set(0, 10, 0, true, 0);
ts.applyTo(ti);
bi = child1->getBottomSegmentIterator();
bs.set(0, 5, 0);
bs._children.clear();
bs.applyTo(bi);
bi->toRight();
bs.set(5, 5, 0);
bs.applyTo(bi);
}
void TopSegmentIteratorParseTest::createCallBack(AlignmentPtr alignment)
{
vector<Sequence::Info> seqVec(1);
BottomSegmentIteratorPtr bi;
BottomSegmentStruct bs;
TopSegmentIteratorPtr ti;
TopSegmentStruct ts;
// case 1: bottom segment aligns perfectly with top segment
Genome* case1 = alignment->addRootGenome("case1");
seqVec[0] = Sequence::Info("Sequence", 10, 2, 2);
case1->setDimensions(seqVec);
ti = case1->getTopSegmentIterator();
ts.set(0, 10, NULL_INDEX, false, 0, NULL_INDEX);
ts.applyTo(ti);
bi = case1->getBottomSegmentIterator();
bs.set(0, 10, 0);
bs.applyTo(bi);
// case 2: bottom segment is completely contained in top segment
Genome* case2 = alignment->addRootGenome("case2");
seqVec[0] = Sequence::Info("Sequence", 10, 2, 3);
case2->setDimensions(seqVec);
ti = case2->getTopSegmentIterator();
ts.set(0, 9, NULL_INDEX, false, 0, NULL_INDEX);
ts.applyTo(ti);
bi = case2->getBottomSegmentIterator();
bs.set(0, 3, 0);
bs.applyTo(bi);
bi->toRight();
bs.set(3, 4, 0);
bs.applyTo(bi);
bi->toRight();
bs.set(7, 3, 0);
bs.applyTo(bi);
// case 3 top segment is completely contained in bottom segment
Genome* case3 = alignment->addRootGenome("case3");
seqVec[0] = Sequence::Info("Sequence", 10, 3, 2);
case3->setDimensions(seqVec);
ti = case3->getTopSegmentIterator();
ts.set(0, 3, NULL_INDEX, false, 0);
ts.applyTo(ti);
ti->toRight();
ts.set(3, 4, NULL_INDEX, false, 0);
ts.applyTo(ti);
ti->toRight();
ts.set(7, 3, NULL_INDEX, false, 0);
ts.applyTo(ti);
bi = case3->getBottomSegmentIterator();
bs.set(0, 9, 0);
bs.applyTo(bi);
// case 4: top segment overhangs bottom segment on the left
Genome* case4 = alignment->addRootGenome("case4");
seqVec[0] = Sequence::Info("Sequence", 10, 2, 2);
case4->setDimensions(seqVec);
ti = case4->getTopSegmentIterator();
ts.set(0, 9, NULL_INDEX, false, 0);
ts.applyTo(ti);
bi = case4->getBottomSegmentIterator();
bs.set(0, 5, 0);
bs.applyTo(bi);
bi->toRight();
bs.set(5, 5, 0);
bs.applyTo(bi);
}
void MappedSegmentMapExtraParalogsTest::createCallBack(AlignmentPtr alignment)
{
vector<Sequence::Info> seqVec(1);
BottomSegmentIteratorPtr bi;
BottomSegmentStruct bs;
TopSegmentIteratorPtr ti;
TopSegmentStruct ts;
// Set up a case where all the segments of grandChild1 coalesce with
// the first segment of grandChild2, but only if using the root as
// the coalescence limit. Otherwise only the first segments map to
// each other.
Genome* root = alignment->addRootGenome("root");
Genome* parent = alignment->addLeafGenome("parent", "root", 1);
Genome* grandChild1 = alignment->addLeafGenome("grandChild1", "parent", 1);
Genome* grandChild2 = alignment->addLeafGenome("grandChild2", "parent", 1);
seqVec[0] = Sequence::Info("Sequence", 3, 0, 1);
root->setDimensions(seqVec);
seqVec[0] = Sequence::Info("Sequence", 9, 3, 3);
parent->setDimensions(seqVec);
seqVec[0] = Sequence::Info("Sequence", 9, 3, 0);
grandChild1->setDimensions(seqVec);
seqVec[0] = Sequence::Info("Sequence", 9, 3, 0);
grandChild2->setDimensions(seqVec);
root->setString("CCC");
parent->setString("CCCTACGTG");
grandChild1->setString("CCCTACGTG");
grandChild2->setString("CCCTACGTG");
bi = root->getBottomSegmentIterator();
bs.set(0, 3);
bs._children.push_back(pair<hal_size_t, bool>(0, false));
bs.applyTo(bi);
ti = parent->getTopSegmentIterator();
ts.set(0, 3, 0, false, NULL_INDEX, 1);
ts.applyTo(ti);
ti->toRight();
ts.set(3, 3, 0, false, NULL_INDEX, 2);
ts.applyTo(ti);
ti->toRight();
ts.set(6, 3, 0, false, NULL_INDEX, 0);
ts.applyTo(ti);
bi = parent->getBottomSegmentIterator();
bs.set(0, 3);
bs._children.clear();
bs._children.push_back(pair<hal_size_t, bool>(0, true));
bs._children.push_back(pair<hal_size_t, bool>(0, false));
bs.applyTo(bi);
bi->toRight();
bs.set(3, 3);
bs._children.clear();
bs._children.push_back(pair<hal_size_t, bool>(1, true));
bs._children.push_back(pair<hal_size_t, bool>(NULL_INDEX, true));
bs.applyTo(bi);
bi->toRight();
bs.set(6, 3);
bs._children.clear();
bs._children.push_back(pair<hal_size_t, bool>(2, true));
bs._children.push_back(pair<hal_size_t, bool>(NULL_INDEX, false));
bs.applyTo(bi);
ti = grandChild1->getTopSegmentIterator();
ts.set(0, 3, 0, true);
ts.applyTo(ti);
ti->toRight();
ts.set(3, 3, 1, true);
ts.applyTo(ti);
ti->toRight();
ts.set(6, 3, 2, true);
ts.applyTo(ti);
ti = grandChild2->getTopSegmentIterator();
ts.set(0, 3, 0, false);
ts.applyTo(ti);
ti->toRight();
ts.set(3, 3, NULL_INDEX, true);
ts.applyTo(ti);
ti->toRight();
ts.set(6, 3, NULL_INDEX, false);
ts.applyTo(ti);
parent->fixParseInfo();
}
void MappedSegmentMapUpTest::createCallBack(AlignmentPtr alignment)
{
vector<Sequence::Info> seqVec(1);
BottomSegmentIteratorPtr bi;
BottomSegmentStruct bs;
TopSegmentIteratorPtr ti;
TopSegmentStruct ts;
// setup simple case were there is an edge from a parent to
// child1 and it is reversed and nonreversed to child2
Genome* parent = alignment->addRootGenome("parent");
Genome* child1 = alignment->addLeafGenome("child1", "parent", 1);
Genome* child2 = alignment->addLeafGenome("child2", "parent", 1);
// add a bunch of grandchildren with no rearrangemnts to test
// simple parsing
Genome* g1 = alignment->addLeafGenome("g1", "child2", 1);
Genome* g2 = alignment->addLeafGenome("g2", "g1", 1);
Genome* g3 = alignment->addLeafGenome("g3", "g2", 1);
Genome* g4 = alignment->addLeafGenome("g4", "g3", 1);
Genome* g5 = alignment->addLeafGenome("g5", "g4", 1);
// add some with random inversions
Genome* gi1 = alignment->addLeafGenome("gi1", "child1", 1);
Genome* gi2 = alignment->addLeafGenome("gi2", "gi1", 1);
Genome* gi3 = alignment->addLeafGenome("gi3", "gi2", 1);
Genome* gi4 = alignment->addLeafGenome("gi4", "gi3", 1);
Genome* gi5 = alignment->addLeafGenome("gi5", "gi4", 1);
Genome* gs[] = {g1, g2, g3, g4, g5};
Genome* gis[] = {gi1, gi2, gi3, gi4, gi5};
seqVec[0] = Sequence::Info("Sequence", 12, 0, 1);
parent->setDimensions(seqVec);
seqVec[0] = Sequence::Info("Sequence", 12, 1, 6);
child1->setDimensions(seqVec);
seqVec[0] = Sequence::Info("Sequence", 12, 1, 6);
child2->setDimensions(seqVec);
seqVec[0] = Sequence::Info("Sequence", 12, 6, 4);
g1->setDimensions(seqVec);
seqVec[0] = Sequence::Info("Sequence", 12, 4, 3);
g2->setDimensions(seqVec);
seqVec[0] = Sequence::Info("Sequence", 12, 3, 2);
g3->setDimensions(seqVec);
seqVec[0] = Sequence::Info("Sequence", 12, 2, 12);
g4->setDimensions(seqVec);
seqVec[0] = Sequence::Info("Sequence", 12, 12, 0);
g5->setDimensions(seqVec);
seqVec[0] = Sequence::Info("Sequence", 12, 6, 4);
gi1->setDimensions(seqVec);
seqVec[0] = Sequence::Info("Sequence", 12, 4, 3);
gi2->setDimensions(seqVec);
seqVec[0] = Sequence::Info("Sequence", 12, 3, 2);
gi3->setDimensions(seqVec);
seqVec[0] = Sequence::Info("Sequence", 12, 2, 12);
gi4->setDimensions(seqVec);
seqVec[0] = Sequence::Info("Sequence", 12, 12, 0);
gi5->setDimensions(seqVec);
parent->setString("CCCTACTTGTGC");
child1->setString("CCCTACTTGTGC");
child2->setString("CCCTACTTGTGC");
for (size_t i = 0; i < 5; ++i)
{
gs[i]->setString("TCCTACTTGTGC");
gis[i]->setString("TCCTACTTGTGC");
}
bi = parent->getBottomSegmentIterator();
bs.set(0, 12);
bs._children.push_back(pair<hal_size_t, bool>(0, true));
bs._children.push_back(pair<hal_size_t, bool>(0, false));
bs.applyTo(bi);
ti = child1->getTopSegmentIterator();
ts.set(0, 12, 0, true, 0);
ts.applyTo(ti);
ti = child2->getTopSegmentIterator();
ts.set(0, 12, 0, false, 0);
ts.applyTo(ti);
for (size_t i = 0; i < 6; ++i)
{
bi = child2->getBottomSegmentIterator(i);
bs.set(i * 2, 2, 0);
bs._children.clear();
bs._children.push_back(pair<hal_size_t, bool>(i, false));
bs.applyTo(bi);
ti = g1->getTopSegmentIterator(i);
ts.set(i * 2, 2, i, false);
ts.applyTo(ti);
}
for (size_t i = 0; i < 6; ++i)
{
bi = child1->getBottomSegmentIterator(i);
bs.set(i * 2, 2, 0);
bs._children.clear();
bs._children.push_back(pair<hal_size_t, bool>(i, false));
bs.applyTo(bi);
ti = gi1->getTopSegmentIterator(i);
ts.set(i * 2, 2, i, false);
ts.applyTo(ti);
}
for (size_t i = 0; i < 5; ++i)
{
const Genome* g = gs[i];
const Genome* parent = g->getParent();
const Genome* child = i == 4 ? NULL : g->getChild(0);
hal_size_t segLen = g->getSequenceLength() / g->getNumTopSegments();
hal_size_t psegLen = parent->getSequenceLength() /
parent->getNumTopSegments();
hal_size_t csegLen = 0;
if (child)
{
csegLen = child->getSequenceLength() / child->getNumTopSegments();
}
for (size_t j = 0; j < g->getNumTopSegments(); ++j)
{
bool inv = false;
bi = parent->getBottomSegmentIterator(j);
bs.set(j * segLen, segLen, (j * segLen) / psegLen);
bs._children.clear();
bs._children.push_back(pair<hal_size_t, bool>(j, inv));
bs.applyTo(bi);
hal_index_t bparse = NULL_INDEX;
if (child != NULL)
{
bparse = (j * segLen) / csegLen;
}
ti = g->getTopSegmentIterator(j);
ts.set(j * segLen, segLen, j, inv, bparse);
ts.applyTo(ti);
}
}
for (size_t i = 0; i < 5; ++i)
{
const Genome* g = gis[i];
const Genome* parent = g->getParent();
const Genome* child = i == 4 ? NULL : g->getChild(0);
hal_size_t segLen = g->getSequenceLength() / g->getNumTopSegments();
hal_size_t psegLen = parent->getSequenceLength() /
parent->getNumTopSegments();
hal_size_t csegLen = 0;
if (child)
{
csegLen = child->getSequenceLength() / child->getNumTopSegments();
}
for (size_t j = 0; j < g->getNumTopSegments(); ++j)
{
bool inv = rand() % 4 == 0;
bi = parent->getBottomSegmentIterator(j);
bs.set(j * segLen, segLen, (j * segLen) / psegLen);
bs._children.clear();
bs._children.push_back(pair<hal_size_t, bool>(j, inv));
bs.applyTo(bi);
hal_index_t bparse = NULL_INDEX;
if (child != NULL)
{
bparse = (j * segLen) / csegLen;
}
ti = g->getTopSegmentIterator(j);
ts.set(j * segLen, segLen, j, inv, bparse);
ts.applyTo(ti);
}
}
}
