bool capsAreAdjacent(Cap *cap1, Cap *cap2, int64_t *separationDistance) {
if (cap_getName(cap2) != cap_getName(cap1) && cap_getCoordinate(cap1) != cap_getCoordinate(cap2)) { //This can happen if end1 == end2
if (sequence_getMetaSequence(cap_getSequence(cap1)) == sequence_getMetaSequence(cap_getSequence(cap2))) {
assert(strcmp(event_getHeader(cap_getEvent(cap1)), event_getHeader(
cap_getEvent(cap2))) == 0);
assert(cap_getPositiveOrientation(cap1)
!= cap_getPositiveOrientation(cap2));
assert(cap_getName(cap1) != cap_getName(cap2));
assert(sequence_getMetaSequence(cap_getSequence(cap1))
== sequence_getMetaSequence(cap_getSequence(cap2)));
if (!cap_getStrand(cap1)) {
cap1 = cap_getReverse(cap1);
}
if (!cap_getStrand(cap2)) {
cap2 = cap_getReverse(cap2);
}
assert(cap_getStrand(cap1));
assert(cap_getStrand(cap2));
if (cap_getCoordinate(cap1) < cap_getCoordinate(cap2)) {
if (!cap_getSide(cap1) && cap_getSide(cap2)) {
*separationDistance = cap_getCoordinate(cap2) - cap_getCoordinate(cap1) - 1; //The minus 1, to give the length of the sequence between the two caps.
return 1;
}
} else {
if (cap_getSide(cap1) && !cap_getSide(cap2)) {
*separationDistance = cap_getCoordinate(cap1) - cap_getCoordinate(cap2) - 1;
return 1;
}
}
}
}
return 0;
}
void testEnd_getInstance(CuTest* testCase) {
cactusEndTestSetup();
CuAssertTrue(testCase, end_getInstance(end, cap_getName(rootCap)) == cap_getReverse(rootCap));
CuAssertTrue(testCase, end_getInstance(end, cap_getName(leaf1Cap)) == cap_getReverse(leaf1Cap));
CuAssertTrue(testCase, end_getInstance(end, cap_getName(leaf2Cap)) == leaf2Cap);
cactusEndTestTeardown();
}
void testCap_getCoordinate(CuTest* testCase) {
cactusCapTestSetup();
CuAssertTrue(testCase, cap_getCoordinate(rootCap) == INT64_MAX);
CuAssertTrue(testCase, cap_getCoordinate(cap_getReverse(rootCap)) == INT64_MAX);
CuAssertTrue(testCase, cap_getCoordinate(leaf1Cap) == 4);
CuAssertTrue(testCase, cap_getCoordinate(cap_getReverse(leaf1Cap)) == 4);
cactusCapTestTeardown();
}
void testCap_getSequence(CuTest* testCase) {
cactusCapTestSetup();
CuAssertTrue(testCase, cap_getSequence(rootCap) == NULL);
CuAssertTrue(testCase, cap_getSequence(cap_getReverse(rootCap)) == NULL);
CuAssertTrue(testCase, cap_getSequence(leaf1Cap) == sequence);
CuAssertTrue(testCase, cap_getSequence(cap_getReverse(leaf1Cap)) == sequence);
cactusCapTestTeardown();
}
void testCap_getEnd(CuTest* testCase) {
cactusCapTestSetup();
CuAssertTrue(testCase, cap_getEnd(rootCap) == end_getReverse(end));
CuAssertTrue(testCase, cap_getEnd(cap_getReverse(rootCap)) == end);
CuAssertTrue(testCase, cap_getEnd(leaf2Cap) == end);
CuAssertTrue(testCase, cap_getEnd(cap_getReverse(leaf2Cap)) == end_getReverse(end));
cactusCapTestTeardown();
}
void testCap_getSide(CuTest* testCase) {
cactusCapTestSetup();
CuAssertTrue(testCase, !cap_getSide(rootCap));
CuAssertTrue(testCase, cap_getSide(cap_getReverse(rootCap)));
CuAssertTrue(testCase, !cap_getSide(leaf1Cap));
CuAssertTrue(testCase, cap_getSide(cap_getReverse(leaf1Cap)));
CuAssertTrue(testCase, cap_getSide(leaf2Cap));
CuAssertTrue(testCase, !cap_getSide(cap_getReverse(leaf2Cap)));
cactusCapTestTeardown();
}
void testCap_adjacent(CuTest* testCase) {
cactusCapTestSetup();
CuAssertTrue(testCase, cap_getAdjacency(leaf1Cap) == NULL);
CuAssertTrue(testCase, cap_getAdjacency(leaf3Cap) == NULL);
cap_makeAdjacent(leaf1Cap, leaf3Cap);
CuAssertTrue(testCase, cap_getAdjacency(leaf1Cap) == cap_getReverse(leaf3Cap));
CuAssertTrue(testCase, cap_getAdjacency(leaf3Cap) == cap_getReverse(leaf1Cap));
CuAssertTrue(testCase, cap_getAdjacency(cap_getReverse(leaf1Cap)) == leaf3Cap);
CuAssertTrue(testCase, cap_getAdjacency(cap_getReverse(leaf3Cap)) == leaf1Cap);
cactusCapTestTeardown();
}
void testCap_getSegment(CuTest* testCase) {
cactusCapTestSetup();
Block *block = block_construct(2, flower);
Segment *segment = segment_construct(block, rootEvent);
CuAssertTrue(testCase, cap_getSegment(segment_get5Cap(segment)) == segment);
CuAssertTrue(testCase, cap_getSegment(segment_get3Cap(segment)) == segment);
CuAssertTrue(testCase, cap_getOrientation(segment_get5Cap(segment)) == segment_getOrientation(segment));
CuAssertTrue(testCase, cap_getOrientation(segment_get3Cap(segment)) == segment_getOrientation(segment));
CuAssertTrue(testCase, cap_getSegment(cap_getReverse(segment_get5Cap(segment))) == segment_getReverse(segment));
CuAssertTrue(testCase, cap_getSegment(cap_getReverse(segment_get3Cap(segment))) == segment_getReverse(segment));
cactusCapTestTeardown();
}
void testCap_getChild(CuTest* testCase) {
cactusCapTestSetup();
CuAssertTrue(testCase, cap_getChild(rootCap, 0) == leaf1Cap);
if (!nestedTest) {
CuAssertTrue(testCase, cap_getChild(rootCap, 1) == cap_getReverse(leaf2Cap));
} else {
// leaf2Cap is at the end of the child list when it's been
// serialized, deleted, and unserialized.
CuAssertTrue(testCase, cap_getChild(rootCap, 2) == cap_getReverse(leaf2Cap));
}
cactusCapTestTeardown();
}
void testCap_getOrientation(CuTest* testCase) {
cactusCapTestSetup();
CuAssertTrue(testCase, cap_getOrientation(rootCap) == end_getOrientation(cap_getEnd(rootCap)));
CuAssertTrue(testCase, cap_getOrientation(leaf1Cap) == end_getOrientation(cap_getEnd(leaf1Cap)));
CuAssertTrue(testCase, cap_getOrientation(leaf2Cap) == end_getOrientation(cap_getEnd(leaf2Cap)));
CuAssertTrue(testCase, cap_getOrientation(cap_getReverse(rootCap)) == end_getOrientation(end_getReverse(cap_getEnd(rootCap))));
CuAssertTrue(testCase, cap_getOrientation(cap_getReverse(leaf1Cap)) == end_getOrientation(end_getReverse(cap_getEnd(leaf1Cap))));
CuAssertTrue(testCase, cap_getOrientation(cap_getReverse(leaf2Cap)) == end_getOrientation(end_getReverse(cap_getEnd(leaf2Cap))));
CuAssertTrue(testCase, cap_getOrientation(leaf1Cap) == cap_getOrientation(rootCap));
CuAssertTrue(testCase, cap_getOrientation(leaf1Cap) != cap_getOrientation(leaf2Cap));
cactusCapTestTeardown();
}
static void recoverBrokenAdjacencies(Flower *flower, stList *recoveredCaps, Name referenceEventName) {
/*
* Find reference intervals that are book-ended by stubs created in a child flower.
*/
Flower_GroupIterator *groupIt = flower_getGroupIterator(flower);
Group *group;
while((group = flower_getNextGroup(groupIt)) != NULL) {
Flower *nestedFlower;
if((nestedFlower = group_getNestedFlower(group)) != NULL) {
Flower_EndIterator *endIt = flower_getEndIterator(nestedFlower);
End *childEnd;
while((childEnd = flower_getNextEnd(endIt)) != NULL) {
if(end_isStubEnd(childEnd) && flower_getEnd(flower, end_getName(childEnd)) == NULL) { //We have a thread we need to promote
Cap *childCap = getCapForReferenceEvent(childEnd, referenceEventName); //The cap in the reference
assert(childCap != NULL);
assert(!end_isAttached(childEnd));
childCap = cap_getStrand(childCap) ? childCap : cap_getReverse(childCap);
if (!cap_getSide(childCap)) {
Cap *adjacentChildCap = NULL;
int64_t adjacencyLength = traceThreadLength(childCap, &adjacentChildCap);
Cap *cap = copyCapToParent(childCap, recoveredCaps);
assert(adjacentChildCap != NULL);
assert(!end_isAttached(cap_getEnd(adjacentChildCap)));
assert(!cap_getSide(cap));
Cap *adjacentCap = copyCapToParent(adjacentChildCap, recoveredCaps);
cap_makeAdjacent(cap, adjacentCap);
setAdjacencyLength(cap, adjacentCap, adjacencyLength);
}
}
}
flower_destructEndIterator(endIt);
}
}
flower_destructGroupIterator(groupIt);
}
void stCaf_addAdjacencies(Flower *flower) {
//Build a list of caps.
stList *list = stList_construct();
Flower_EndIterator *endIterator = flower_getEndIterator(flower);
End *end;
while ((end = flower_getNextEnd(endIterator)) != NULL) {
End_InstanceIterator *instanceIterator = end_getInstanceIterator(end);
Cap *cap;
while ((cap = end_getNext(instanceIterator)) != NULL) {
if (!cap_getStrand(cap)) {
cap = cap_getReverse(cap);
}
stList_append(list, cap);
}
end_destructInstanceIterator(instanceIterator);
}
flower_destructEndIterator(endIterator);
assert(stList_length(list) % 2 == 0);
//Sort the list of caps.
stList_sort(list, (int(*)(const void *, const void *)) addAdjacenciesPP);
//Now make the adjacencies.
for (int64_t i = 1; i < stList_length(list); i += 2) {
Cap *cap = stList_get(list, i - 1);
Cap *cap2 = stList_get(list, i);
cap_makeAdjacent(cap, cap2);
}
//Clean up.
stList_destruct(list);
}
void testCap_getParent(CuTest* testCase) {
cactusCapTestSetup();
CuAssertTrue(testCase, cap_getParent(rootCap) == NULL);
CuAssertTrue(testCase, cap_getParent(leaf1Cap) == rootCap);
CuAssertTrue(testCase, cap_getParent(leaf2Cap) == cap_getReverse(rootCap));
cactusCapTestTeardown();
}
void testCap_getTopFace(CuTest* testCase) {
cactusCapTestSetup();
Face *face = face_construct(flower);
cap_setTopFace(rootCap, face);
CuAssertTrue(testCase, cap_getTopFace(rootCap) == face);
CuAssertTrue(testCase, cap_getTopFace(cap_getReverse(rootCap)) == face);
cactusCapTestTeardown();
}
Cap *getTerminalCap(Cap *cap) {
Flower *nestedFlower = group_getNestedFlower(end_getGroup(cap_getEnd(cap)));
if (nestedFlower != NULL) {
Cap *nestedCap = flower_getCap(nestedFlower, cap_getName(cap));
assert(nestedCap != NULL);
return getTerminalCap(cap_getOrientation(cap) ? nestedCap : cap_getReverse(nestedCap));
}
return cap;
}
void testCap_getOtherSegmentCap(CuTest *testCase) {
cactusCapTestSetup();
Block *block = block_construct(3, flower);
Segment *segment = segment_construct2(block, 2, 1, sequence);
Cap *_5Cap = segment_get5Cap(segment);
Cap *_3Cap = segment_get3Cap(segment);
CuAssertTrue(testCase, cap_getOtherSegmentCap(leaf1Cap) == NULL);
CuAssertTrue(testCase, cap_getOtherSegmentCap(cap_getReverse(leaf1Cap)) == NULL);
CuAssertTrue(testCase, cap_getOtherSegmentCap(_5Cap) == _3Cap);
CuAssertTrue(testCase, cap_getOtherSegmentCap(_3Cap) == _5Cap);
CuAssertTrue(testCase, cap_getOtherSegmentCap(cap_getReverse(_5Cap)) == cap_getReverse(_3Cap));
CuAssertTrue(testCase, cap_getOtherSegmentCap(cap_getReverse(_3Cap)) == cap_getReverse(_5Cap));
cactusCapTestTeardown();
}
void testEnd_getCapForEvent(CuTest* testCase) {
cactusEndTestSetup();
CuAssertPtrEquals(testCase, end_getCapForEvent(end_getReverse(end), event_getName(rootEvent)), rootCap);
Cap *cap = end_getCapForEvent(end, event_getName(leafEvent));
CuAssertTrue(testCase, cap == cap_getReverse(leaf1Cap) || cap == leaf2Cap || cap == cap_getReverse(leaf3Cap));
CuAssertTrue(testCase, end_getCapForEvent(end, NULL_NAME) == NULL);
cactusEndTestTeardown();
}
void testEnd_getSetRootInstance(CuTest* testCase) {
cactusEndTestSetup();
CuAssertTrue(testCase, end_getRootInstance(end) == cap_getReverse(rootCap));
CuAssertTrue(testCase, end_getRootInstance(end_getReverse(end)) == rootCap);
End *end2 = end_construct(0, flower);
CuAssertTrue(testCase, end_getRootInstance(end2) == NULL);
CuAssertTrue(testCase, end_getRootInstance(end_getReverse(end2)) == NULL);
cactusEndTestTeardown();
}
static void testAdjacencySequence_4(CuTest *testCase) {
setup();
AdjacencySequence *adjacencySequence = adjacencySequence_construct(cap_getReverse(cap2), 0);
CuAssertTrue(testCase, adjacencySequence->subsequenceIdentifier == cap_getName(cap1)); //sequence_getName(sequence1));
CuAssertIntEquals(testCase, adjacencySequence->start, 4);
CuAssertIntEquals(testCase, adjacencySequence->strand, 0);
CuAssertIntEquals(testCase, adjacencySequence->length, 0);
CuAssertStrEquals(testCase, "", adjacencySequence->string);
adjacencySequence_destruct(adjacencySequence);
teardown();
}
void testEnd_serialisation(CuTest* testCase) {
cactusEndTestSetup();
Name rootInstanceName = cap_getName(rootCap);
Name leaf1InstanceName = cap_getName(leaf1Cap);
Name leaf2InstanceName = cap_getName(leaf2Cap);
Name leaf3InstanceName = cap_getName(leaf3Cap);
int64_t i;
void *vA = binaryRepresentation_makeBinaryRepresentation(end,
(void(*)(void *, void(*)(const void *, size_t, size_t))) end_writeBinaryRepresentation, &i);
CuAssertTrue(testCase, i > 0);
end_destruct(end);
void *vA2 = vA;
end = end_loadFromBinaryRepresentation(&vA2, flower);
rootCap = cap_getReverse(end_getInstance(end, rootInstanceName));
leaf1Cap = cap_getReverse(end_getInstance(end, leaf1InstanceName));
leaf2Cap = end_getInstance(end, leaf2InstanceName);
leaf3Cap = cap_getReverse(end_getInstance(end, leaf3InstanceName));
CuAssertTrue(testCase, leaf3Cap != NULL);
free(vA);
nestedTest = 1;
testEnd_copyConstruct(testCase);
testEnd_getName(testCase);
testEnd_getOrientation(testCase);
testEnd_getReverse(testCase);
testEnd_getSide(testCase);
testEnd_getFlower(testCase);
testEnd_getBlock(testCase);
testEnd_getOtherBlockEnd(testCase);
testEnd_getGroup(testCase);
testEnd_setGroup(testCase);
testEnd_getInstanceNumber(testCase);
testEnd_getInstance(testCase);
testEnd_getFirst(testCase);
testEnd_getSetRootInstance(testCase);
testEnd_instanceIterator(testCase);
testEnd_isBlockOrStubEnd(testCase);
testEnd_isAttachedOrFree(testCase);
testEnd_getCapForEvent(testCase);
nestedTest = 0;
cactusEndTestTeardown();
}
void testCap_segmentCoordinatesReverseStrand(CuTest* testCase) {
/*
* Tests the coordinates of an segment and its 5 and 3 prime caps.
*/
cactusCapTestSetup();
Block *block = block_construct(3, flower);
Segment *segment = segment_construct2(block, 2, 0, sequence);
Cap *_5Cap = segment_get5Cap(segment);
Cap *_3Cap = segment_get3Cap(segment);
CuAssertTrue(testCase, cap_getSide(_5Cap));
CuAssertTrue(testCase, !cap_getSide(_3Cap));
CuAssertTrue(testCase, !cap_getStrand(_5Cap));
CuAssertTrue(testCase, !cap_getStrand(_3Cap));
CuAssertIntEquals(testCase, 4, cap_getCoordinate(_5Cap));
CuAssertIntEquals(testCase, 2, cap_getCoordinate(_3Cap));
CuAssertTrue(testCase, !segment_getStrand(segment));
CuAssertIntEquals(testCase, 4, segment_getStart(segment));
CuAssertIntEquals(testCase, 3, segment_getLength(segment));
CuAssertTrue(testCase, !cap_getSide(cap_getReverse(_5Cap)));
CuAssertTrue(testCase, cap_getSide(cap_getReverse(_3Cap)));
CuAssertTrue(testCase, cap_getStrand(cap_getReverse(_5Cap)));
CuAssertTrue(testCase, cap_getStrand(cap_getReverse(_3Cap)));
CuAssertIntEquals(testCase, 4, cap_getCoordinate(cap_getReverse(_5Cap)));
CuAssertIntEquals(testCase, 2, cap_getCoordinate(cap_getReverse(_3Cap)));
CuAssertTrue(testCase, segment_getStrand(segment_getReverse(segment)));
CuAssertIntEquals(testCase, 2, segment_getStart(segment_getReverse(segment)));
CuAssertIntEquals(testCase, 3, segment_getLength(segment_getReverse(segment)));
cactusCapTestTeardown();
}
void topDown(Flower *flower, Name referenceEventName) {
/*
* Run on each flower, top down. Sets the coordinates of each reference cap to the correct
* sequence, and sets the bases of the reference sequence to be consensus bases.
*/
Flower_EndIterator *endIt = flower_getEndIterator(flower);
End *end;
while ((end = flower_getNextEnd(endIt)) != NULL) {
Cap *cap = getCapForReferenceEvent(end, referenceEventName); //The cap in the reference
if (cap != NULL) {
cap = cap_getStrand(cap) ? cap : cap_getReverse(cap);
if (!cap_getSide(cap)) {
assert(cap_getCoordinate(cap) != INT64_MAX);
Sequence *sequence = cap_getSequence(cap);
assert(sequence != NULL);
Group *group = end_getGroup(end);
if (!group_isLeaf(group)) {
Flower *nestedFlower = group_getNestedFlower(group);
Cap *nestedCap = flower_getCap(nestedFlower, cap_getName(cap));
assert(nestedCap != NULL);
nestedCap = cap_getStrand(nestedCap) ? nestedCap : cap_getReverse(nestedCap);
assert(cap_getStrand(nestedCap));
assert(!cap_getSide(nestedCap));
int64_t endCoordinate = setCoordinates(nestedFlower, sequence_getMetaSequence(sequence),
nestedCap, cap_getCoordinate(cap));
(void) endCoordinate;
assert(endCoordinate == cap_getCoordinate(cap_getAdjacency(cap)));
assert(endCoordinate
== cap_getCoordinate(
flower_getCap(nestedFlower, cap_getName(cap_getAdjacency(cap)))));
}
}
}
}
flower_destructEndIterator(endIt);
}
void testCap_getTopCap(CuTest* testCase) {
cactusCapTestSetup();
End *end1 = end_construct(0, flower);
End *end2 = end_construct(0, flower);
End *end3 = end_construct(0, flower);
Event *intermediateEvent = event_construct4(NULL, 0.0, rootEvent, leafEvent, eventTree);
Cap *cap1T = cap_construct(end1, rootEvent);
Cap *cap1I = cap_construct(end1, intermediateEvent);
Cap *cap1L1 = cap_construct(end1, leafEvent);
Cap *cap1L2 = cap_construct(end1, leafEvent);
cap_makeParentAndChild(cap1I, cap1L1);
cap_makeParentAndChild(cap1I, cap1L2);
cap_makeParentAndChild(cap1T, cap1I);
end_setRootInstance(end1, cap1T);
assert(end_getRootInstance(end1) == cap1T);
CuAssertTrue(testCase, cap_getTopCap(cap1L1) == NULL);
CuAssertTrue(testCase, cap_getTopCap(cap_getReverse(cap1L1)) == NULL);
CuAssertTrue(testCase, cap_getTopCap(cap1L2) == NULL);
CuAssertTrue(testCase, cap_getTopCap(cap1I) == NULL);
Cap *cap2T = cap_construct(end2, rootEvent);
Cap *cap2L = cap_construct(end2, leafEvent);
cap_makeParentAndChild(cap2T, cap2L);
end_setRootInstance(end2, cap2T);
cap_makeAdjacent(cap1L1, cap2L);
CuAssertTrue(testCase, cap_getTopCap(cap1L1) == cap1T);
CuAssertTrue(testCase, cap_getTopCap(cap_getReverse(cap1L1)) == cap_getReverse(cap1T));
CuAssertTrue(testCase, cap_getTopCap(cap1I) == NULL);
Cap *cap3T = cap_construct(end3, rootEvent);
Cap *cap3I = cap_construct(end3, intermediateEvent);
cap_makeParentAndChild(cap3T, cap3I);
end_setRootInstance(end3, cap3T);
cap_makeAdjacent(cap1I, cap3I);
cap_makeAdjacent(cap1T, cap3T);
CuAssertTrue(testCase, cap_getTopCap(cap1L1) == cap1I);
CuAssertTrue(testCase, cap_getTopCap(cap_getReverse(cap1L1)) == cap_getReverse(cap1I));
CuAssertTrue(testCase, cap_getTopCap(cap1I) == cap1T);
CuAssertTrue(testCase, cap_getTopCap(cap_getReverse(cap1I)) == cap_getReverse(cap1T));
CuAssertTrue(testCase, cap_getTopCap(cap1T) == NULL);
cactusCapTestTeardown();
}
static Cap *copyCapToParent(Cap *cap, stList *recoveredCaps) {
/*
* Get the adjacent stub end by looking at the reference adjacency in the parent.
*/
End *end = cap_getEnd(cap);
assert(end != NULL);
Group *parentGroup = flower_getParentGroup(end_getFlower(end));
assert(parentGroup != NULL);
End *copiedEnd = end_copyConstruct(end, group_getFlower(parentGroup));
end_setGroup(copiedEnd, parentGroup); //Set group
Cap *copiedCap = end_getInstance(copiedEnd, cap_getName(cap));
assert(copiedCap != NULL);
copiedCap = cap_getStrand(copiedCap) ? copiedCap : cap_getReverse(copiedCap);
if (!cap_getSide(copiedCap)) {
stList_append(recoveredCaps, copiedCap);
}
return copiedCap;
}
char *getTerminalAdjacencySubString(Cap *cap) {
if(getTerminalAdjacencyLength_ignoreAdjacencies) {
return stString_copy("");
}
cap = getTerminalCap(cap);
cap = cap_getStrand(cap) ? cap : cap_getReverse(cap); //This ensures the asserts are as expected.
Cap *adjacentCap = cap_getAdjacency(cap);
int64_t i = cap_getCoordinate(cap) - cap_getCoordinate(adjacentCap);
assert(i != 0);
if (i > 0) {
assert(cap_getSide(cap));
assert(!cap_getSide(adjacentCap));
return sequence_getString(cap_getSequence(cap),
cap_getCoordinate(adjacentCap) + 1, i - 1, 1);
} else {
assert(cap_getSide(adjacentCap));
assert(!cap_getSide(cap));
return sequence_getString(cap_getSequence(cap), cap_getCoordinate(cap) + 1, -i - 1, 1);
}
}
static stList *getCaps(stList *flowers, Name referenceEventName) {
stList *caps = stList_construct();
for (int64_t i = 0; i < stList_length(flowers); i++) {
Flower *flower = stList_get(flowers, i);
//Get list of caps
Flower_EndIterator *endIt = flower_getEndIterator(flower);
End *end;
while ((end = flower_getNextEnd(endIt)) != NULL) {
if (end_isStubEnd(end)) {
Cap *cap = getCapForReferenceEvent(end, referenceEventName); //The cap in the reference
if(cap != NULL) {
cap = cap_getStrand(cap) ? cap : cap_getReverse(cap);
if (!cap_getSide(cap)) {
stList_append(caps, cap);
}
}
}
}
flower_destructEndIterator(endIt);
}
return caps;
}
int64_t flower_getTotalBaseLength(Flower *flower) {
/*
* The implementation of this function is very like that in group_getTotalBaseLength, with a few differences. Consider merging them.
*/
Flower_EndIterator *endIterator = flower_getEndIterator(flower);
End *end;
int64_t totalLength = 0;
while ((end = flower_getNextEnd(endIterator)) != NULL) {
if (!end_isBlockEnd(end)) {
End_InstanceIterator *instanceIterator = end_getInstanceIterator(end);
Cap *cap;
while ((cap = end_getNext(instanceIterator)) != NULL) {
cap = cap_getStrand(cap) ? cap : cap_getReverse(cap);
if (!cap_getSide(cap) && cap_getSequence(cap) != NULL) {
Cap *cap2 = cap_getAdjacency(cap);
assert(cap2 != NULL);
while (end_isBlockEnd(cap_getEnd(cap2))) {
Segment *segment = cap_getSegment(cap2);
assert(segment != NULL);
assert(segment_get5Cap(segment) == cap2);
cap2 = cap_getAdjacency(segment_get3Cap(segment));
assert(cap2 != NULL);
assert(cap_getStrand(cap2));
assert(cap_getSide(cap2));
}
assert(cap_getStrand(cap2));
assert(cap_getSide(cap2));
int64_t length = cap_getCoordinate(cap2) - cap_getCoordinate(cap) - 1;
assert(length >= 0);
totalLength += length;
}
}
end_destructInstanceIterator(instanceIterator);
}
}
flower_destructEndIterator(endIterator);
return totalLength;
}
Segment *getCapsSegment(Cap *cap) {
if (cap_getSegment(cap) != NULL) {
return cap_getSegment(cap);
}
assert(!end_isBlockEnd(cap_getEnd(cap)));
assert(end_isStubEnd(cap_getEnd(cap)));
//Walk up to get the next adjacency.
Group *parentGroup = flower_getParentGroup(end_getFlower(cap_getEnd(cap)));
if (parentGroup != NULL) {
Cap *parentCap = flower_getCap(group_getFlower(parentGroup), cap_getName(cap));
if (parentCap != NULL) {
assert(cap_getOrientation(parentCap));
if (!cap_getOrientation(cap)) {
parentCap = cap_getReverse(parentCap);
}
return getCapsSegment(parentCap);
} else { //Cap must be a free stub end.
assert(0); //Not in the current alignments.
assert(end_isFree(cap_getEnd(cap)));
}
}
return NULL;
}
static stList *getSubstringsForFlowers(stList *flowers) {
/*
* Get the set of substrings for sequence intervals in the given set of flowers.
*/
stList *substrings = stList_construct3(0, (void (*)(void *)) substring_destruct);
for (int64_t i = 0; i < stList_length(flowers); i++) {
Flower *flower = stList_get(flowers, i);
Flower_EndIterator *endIt = flower_getEndIterator(flower);
End *end;
while ((end = flower_getNextEnd(endIt)) != NULL) {
if (end_isStubEnd(end)) {
End_InstanceIterator *instanceIt = end_getInstanceIterator(end);
Cap *cap;
while ((cap = end_getNext(instanceIt)) != NULL) {
Sequence *sequence;
if ((sequence = cap_getSequence(cap)) != NULL) {
cap = cap_getStrand(cap) ? cap : cap_getReverse(cap);
if (!cap_getSide(cap)) { //We have a sequence interval of interest
Cap *adjacentCap = cap_getAdjacency(cap);
assert(adjacentCap != NULL);
int64_t length = cap_getCoordinate(adjacentCap) - cap_getCoordinate(cap) - 1;
assert(length >= 0);
if (length > 0) {
stList_append(substrings,
substring_construct(sequence_getMetaSequence(sequence)->stringName,
cap_getCoordinate(cap) + 1 - sequence_getStart(sequence), length));
}
}
}
}
end_destructInstanceIterator(instanceIt);
}
}
flower_destructEndIterator(endIt);
}
return substrings;
}
stSortedSet *makeEndAlignment(StateMachine *sM, End *end, int64_t spanningTrees, int64_t maxSequenceLength,
bool useProgressiveMerging, float gapGamma,
PairwiseAlignmentParameters *pairwiseAlignmentBandingParameters) {
//Make an alignment of the sequences in the ends
//Get the adjacency sequences to be aligned.
Cap *cap;
End_InstanceIterator *it = end_getInstanceIterator(end);
stList *sequences = stList_construct3(0, (void (*)(void *))adjacencySequence_destruct);
stList *seqFrags = stList_construct3(0, (void (*)(void *))seqFrag_destruct);
stHash *endInstanceNumbers = stHash_construct2(NULL, free);
while((cap = end_getNext(it)) != NULL) {
if(cap_getSide(cap)) {
cap = cap_getReverse(cap);
}
AdjacencySequence *adjacencySequence = adjacencySequence_construct(cap, maxSequenceLength);
stList_append(sequences, adjacencySequence);
assert(cap_getAdjacency(cap) != NULL);
End *otherEnd = end_getPositiveOrientation(cap_getEnd(cap_getAdjacency(cap)));
stList_append(seqFrags, seqFrag_construct(adjacencySequence->string, 0, end_getName(otherEnd)));
//Increase count of seqfrags with a given end.
int64_t *c = stHash_search(endInstanceNumbers, otherEnd);
if(c == NULL) {
c = st_calloc(1, sizeof(int64_t));
assert(*c == 0);
stHash_insert(endInstanceNumbers, otherEnd, c);
}
(*c)++;
}
end_destructInstanceIterator(it);
//Get the alignment.
MultipleAlignment *mA = makeAlignment(sM, seqFrags, spanningTrees, 100000000, useProgressiveMerging, gapGamma, pairwiseAlignmentBandingParameters);
//Build an array of weights to reweight pairs in the alignment.
int64_t *pairwiseAlignmentsPerSequenceNonCommonEnds = st_calloc(stList_length(seqFrags), sizeof(int64_t));
int64_t *pairwiseAlignmentsPerSequenceCommonEnds = st_calloc(stList_length(seqFrags), sizeof(int64_t));
//First build array on number of pairwise alignments to each sequence, distinguishing alignments between sequences sharing
//common ends.
for(int64_t i=0; i<stList_length(mA->chosenPairwiseAlignments); i++) {
stIntTuple *pairwiseAlignment = stList_get(mA->chosenPairwiseAlignments, i);
int64_t seq1 = stIntTuple_get(pairwiseAlignment, 1);
int64_t seq2 = stIntTuple_get(pairwiseAlignment, 2);
assert(seq1 != seq2);
SeqFrag *seqFrag1 = stList_get(seqFrags, seq1);
SeqFrag *seqFrag2 = stList_get(seqFrags, seq2);
int64_t *pairwiseAlignmentsPerSequence = seqFrag1->rightEndId == seqFrag2->rightEndId
? pairwiseAlignmentsPerSequenceCommonEnds : pairwiseAlignmentsPerSequenceNonCommonEnds;
pairwiseAlignmentsPerSequence[seq1]++;
pairwiseAlignmentsPerSequence[seq2]++;
}
//Now calculate score adjustments.
double *scoreAdjustmentsNonCommonEnds = st_malloc(stList_length(seqFrags) * sizeof(double));
double *scoreAdjustmentsCommonEnds = st_malloc(stList_length(seqFrags) * sizeof(double));
for(int64_t i=0; i<stList_length(seqFrags); i++) {
SeqFrag *seqFrag = stList_get(seqFrags, i);
End *otherEnd = flower_getEnd(end_getFlower(end), seqFrag->rightEndId);
assert(otherEnd != NULL);
assert(stHash_search(endInstanceNumbers, otherEnd) != NULL);
int64_t commonInstanceNumber = *(int64_t *)stHash_search(endInstanceNumbers, otherEnd);
int64_t nonCommonInstanceNumber = stList_length(seqFrags) - commonInstanceNumber;
assert(commonInstanceNumber > 0 && nonCommonInstanceNumber >= 0);
assert(pairwiseAlignmentsPerSequenceNonCommonEnds[i] <= nonCommonInstanceNumber);
assert(pairwiseAlignmentsPerSequenceNonCommonEnds[i] >= 0);
assert(pairwiseAlignmentsPerSequenceCommonEnds[i] < commonInstanceNumber);
assert(pairwiseAlignmentsPerSequenceCommonEnds[i] >= 0);
//scoreAdjustmentsNonCommonEnds[i] = ((double)nonCommonInstanceNumber + commonInstanceNumber - 1)/(pairwiseAlignmentsPerSequenceNonCommonEnds[i] + pairwiseAlignmentsPerSequenceCommonEnds[i]);
//scoreAdjustmentsCommonEnds[i] = scoreAdjustmentsNonCommonEnds[i];
if(pairwiseAlignmentsPerSequenceNonCommonEnds[i] > 0) {
scoreAdjustmentsNonCommonEnds[i] = ((double)nonCommonInstanceNumber)/pairwiseAlignmentsPerSequenceNonCommonEnds[i];
assert(scoreAdjustmentsNonCommonEnds[i] >= 1.0);
assert(scoreAdjustmentsNonCommonEnds[i] <= nonCommonInstanceNumber);
}
else {
scoreAdjustmentsNonCommonEnds[i] = INT64_MIN;
}
if(pairwiseAlignmentsPerSequenceCommonEnds[i] > 0) {
scoreAdjustmentsCommonEnds[i] = ((double)commonInstanceNumber-1)/pairwiseAlignmentsPerSequenceCommonEnds[i];
assert(scoreAdjustmentsCommonEnds[i] >= 1.0);
assert(scoreAdjustmentsCommonEnds[i] <= commonInstanceNumber-1);
}
else {
scoreAdjustmentsCommonEnds[i] = INT64_MIN;
}
}
//Convert the alignment pairs to an alignment of the caps..
stSortedSet *sortedAlignment =
stSortedSet_construct3((int (*)(const void *, const void *))alignedPair_cmpFn,
(void (*)(void *))alignedPair_destruct);
while(stList_length(mA->alignedPairs) > 0) {
stIntTuple *alignedPair = stList_pop(mA->alignedPairs);
assert(stIntTuple_length(alignedPair) == 5);
int64_t seqIndex1 = stIntTuple_get(alignedPair, 1);
int64_t seqIndex2 = stIntTuple_get(alignedPair, 3);
AdjacencySequence *i = stList_get(sequences, seqIndex1);
AdjacencySequence *j = stList_get(sequences, seqIndex2);
assert(i != j);
int64_t offset1 = stIntTuple_get(alignedPair, 2);
int64_t offset2 = stIntTuple_get(alignedPair, 4);
int64_t score = stIntTuple_get(alignedPair, 0);
if(score <= 0) { //Happens when indel probs are included
score = 1; //This is the minimum
}
assert(score > 0 && score <= PAIR_ALIGNMENT_PROB_1);
SeqFrag *seqFrag1 = stList_get(seqFrags, seqIndex1);
SeqFrag *seqFrag2 = stList_get(seqFrags, seqIndex2);
assert(seqFrag1 != seqFrag2);
double *scoreAdjustments = seqFrag1->rightEndId == seqFrag2->rightEndId ? scoreAdjustmentsCommonEnds : scoreAdjustmentsNonCommonEnds;
assert(scoreAdjustments[seqIndex1] != INT64_MIN);
assert(scoreAdjustments[seqIndex2] != INT64_MIN);
AlignedPair *alignedPair2 = alignedPair_construct(
i->subsequenceIdentifier, i->start + (i->strand ? offset1 : -offset1), i->strand,
j->subsequenceIdentifier, j->start + (j->strand ? offset2 : -offset2), j->strand,
score*scoreAdjustments[seqIndex1], score*scoreAdjustments[seqIndex2]); //Do the reweighting here.
assert(stSortedSet_search(sortedAlignment, alignedPair2) == NULL);
assert(stSortedSet_search(sortedAlignment, alignedPair2->reverse) == NULL);
stSortedSet_insert(sortedAlignment, alignedPair2);
stSortedSet_insert(sortedAlignment, alignedPair2->reverse);
stIntTuple_destruct(alignedPair);
}
//Cleanup
stList_destruct(seqFrags);
stList_destruct(sequences);
free(pairwiseAlignmentsPerSequenceNonCommonEnds);
free(pairwiseAlignmentsPerSequenceCommonEnds);
free(scoreAdjustmentsNonCommonEnds);
free(scoreAdjustmentsCommonEnds);
multipleAlignment_destruct(mA);
stHash_destruct(endInstanceNumbers);
return sortedAlignment;
}