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();
}
static void setAdjacencyLength(Cap *cap, Cap *adjacentCap, int64_t adjacencyLength) {
//Set the coordinates of the caps to the adjacency size
cap_setCoordinates(cap, adjacencyLength, cap_getStrand(cap), NULL);
cap_setCoordinates(adjacentCap, adjacencyLength, cap_getStrand(adjacentCap),
NULL);
assert(cap_getCoordinate(cap) == cap_getCoordinate(adjacentCap));
}
static int64_t traceThreadLength(Cap *cap, Cap **terminatingCap) {
/*
* Gets the length in bases of the thread in the flower, starting from a given attached stub cap.
* The thread length includes the lengths of adjacencies that it contains.
*
* Terminating cap is initialised with the final cap on the thread from cap.
*/
assert(end_isStubEnd(cap_getEnd(cap)));
int64_t threadLength = 0;
while (1) {
assert(cap_getCoordinate(cap) != INT64_MAX);
int64_t adjacencyLength = cap_getCoordinate(cap);
threadLength += adjacencyLength;
Cap *adjacentCap = cap_getAdjacency(cap);
assert(adjacentCap != NULL);
assert(adjacencyLength == cap_getCoordinate(adjacentCap));
//Traverse any block..
if (cap_getSegment(adjacentCap) != NULL) {
threadLength += segment_getLength(cap_getSegment(adjacentCap));
cap = cap_getOtherSegmentCap(adjacentCap);
assert(cap != NULL);
} else {
assert(end_isStubEnd(cap_getEnd(adjacentCap)));
*terminatingCap = adjacentCap;
return threadLength;
}
}
return 1;
}
void printPositions(stList *positions, const char *substitutionType, FILE *fileHandle) {
for (int64_t i = 0; i < stList_length(positions); i++) {
SegmentHolder *segmentHolder = stList_get(positions, i);
int64_t j = segment_getStart(segmentHolder->segment);
if (segment_getStrand(segmentHolder->segment)) {
j += segmentHolder->offset;
assert(
cap_getCoordinate(segment_get5Cap(segmentHolder->segment)) == segment_getStart(
segmentHolder->segment));
assert(
segment_getStart(segmentHolder->segment) + segment_getLength(segmentHolder->segment) - 1
== cap_getCoordinate(segment_get3Cap(segmentHolder->segment)));
} else {
j -= segmentHolder->offset;
assert(
segment_getStart(segmentHolder->segment) - segment_getLength(segmentHolder->segment) + 1
== cap_getCoordinate(segment_get3Cap(segmentHolder->segment)));
}
fprintf(fileHandle, "%s: %s_%" PRIi64 " %" PRIi64 " %c %c %c\n", substitutionType,
event_getHeader(segment_getEvent(segmentHolder->segment)),
sequence_getLength(segment_getSequence(segmentHolder->segment)), j,
segmentHolder->base1, segmentHolder->base2, segmentHolder->base3);
getMAFBlock(segment_getBlock(segmentHolder->segment), fileHandle);
}
}
static int64_t setCoordinates(Flower *flower, MetaSequence *metaSequence, Cap *cap, int64_t coordinate) {
/*
* Sets the coordinates of the reference thread and sets the bases of the actual sequence
* that of the consensus.
*/
Sequence *sequence = flower_getSequence(flower, metaSequence_getName(metaSequence));
if (sequence == NULL) {
sequence = sequence_construct(metaSequence, flower);
}
while (1) {
assert(cap_getStrand(cap));
assert(!cap_getSide(cap));
Cap *adjacentCap = cap_getAdjacency(cap);
assert(adjacentCap != NULL);
assert(cap_getStrand(adjacentCap));
assert(cap_getSide(adjacentCap));
int64_t adjacencyLength = cap_getCoordinate(cap);
assert(adjacencyLength == cap_getCoordinate(adjacentCap));
assert(adjacencyLength != INT64_MAX);
assert(adjacencyLength >= 0);
cap_setCoordinates(cap, coordinate, 1, sequence);
coordinate += adjacencyLength + 1;
cap_setCoordinates(adjacentCap, coordinate, 1, sequence);
//Traverse any block..
if ((cap = cap_getOtherSegmentCap(adjacentCap)) == NULL) {
break;
}
coordinate += segment_getLength(cap_getSegment(adjacentCap)) - 1;
}
return coordinate;
}
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();
}
static void setAdjacencyLengthsAndRecoverNewCapsAndBrokenAdjacencies(Cap *cap, stList *recoveredCaps) {
/*
* Sets the coordinates of the caps to be equal to the length of the adjacency sequence between them.
* Used to build the reference sequence bottom up.
*
* One complexity is that a reference thread between the two caps
* in each flower f may be broken into two in the children of f.
* Therefore, for each flower f first identify attached stub ends present in the children of f that are
* not present in f and copy them into f, reattaching the reference caps as needed.
*/
while (1) {
Cap *adjacentCap = cap_getAdjacency(cap);
assert(adjacentCap != NULL);
assert(cap_getCoordinate(cap) == INT64_MAX);
assert(cap_getCoordinate(adjacentCap) == INT64_MAX);
assert(cap_getStrand(cap) == cap_getStrand(adjacentCap));
assert(cap_getSide(cap) != cap_getSide(adjacentCap));
Group *group = end_getGroup(cap_getEnd(cap));
assert(group != NULL);
if (!group_isLeaf(group)) { //Adjacency is not terminal, so establish its sequence.
Flower *nestedFlower = group_getNestedFlower(group);
Cap *nestedCap = flower_getCap(nestedFlower, cap_getName(cap));
assert(nestedCap != NULL);
Cap *nestedAdjacentCap = flower_getCap(nestedFlower, cap_getName(adjacentCap));
assert(nestedAdjacentCap != NULL);
Cap *breakerCap;
int64_t adjacencyLength = traceThreadLength(nestedCap, &breakerCap);
assert(cap_getOrientation(nestedAdjacentCap));
if (cap_getPositiveOrientation(breakerCap) != nestedAdjacentCap) { //The thread is broken at the lower level.
//Copy cap into higher level graph.
breakerCap = copyCapToParent(breakerCap, recoveredCaps);
assert(cap_getSide(breakerCap));
cap_makeAdjacent(cap, breakerCap);
setAdjacencyLength(cap, breakerCap, adjacencyLength);
adjacencyLength = traceThreadLength(nestedAdjacentCap, &breakerCap);
assert(cap_getPositiveOrientation(breakerCap) != cap);
breakerCap = copyCapToParent(breakerCap, recoveredCaps);
assert(!cap_getSide(breakerCap));
cap_makeAdjacent(breakerCap, adjacentCap);
setAdjacencyLength(adjacentCap, breakerCap, adjacencyLength);
} else { //The thread is not broken at the lower level
setAdjacencyLength(cap, adjacentCap, adjacencyLength);
}
} else {
//Set the coordinates of the caps to the adjacency size
setAdjacencyLength(cap, adjacentCap, 0);
}
if ((cap = cap_getOtherSegmentCap(adjacentCap)) == NULL) {
break;
}
}
}
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 testCap_setCoordinate(CuTest* testCase) {
cactusCapTestSetup();
CuAssertTrue(testCase, cap_getCoordinate(rootCap) == INT64_MAX);
CuAssertTrue(testCase, cap_getStrand(rootCap));
CuAssertTrue(testCase, cap_getSequence(rootCap) == NULL);
cap_setCoordinates(rootCap, 5, 0, NULL);
CuAssertTrue(testCase, cap_getCoordinate(rootCap) == 5);
CuAssertTrue(testCase, !cap_getStrand(rootCap));
CuAssertTrue(testCase, cap_getSequence(rootCap) == NULL);
cap_setCoordinates(rootCap, INT64_MAX, 1, NULL);
CuAssertTrue(testCase, cap_getCoordinate(rootCap) == INT64_MAX);
CuAssertTrue(testCase, cap_getStrand(rootCap));
CuAssertTrue(testCase, cap_getSequence(rootCap) == NULL);
cactusCapTestTeardown();
}
void mapBlockToExon(Cap *cap, int level, FILE *fileHandle){
fprintf(fileHandle, "\t\t\t<block>\n");
Block *block = end_getBlock(cap_getEnd(cap));
Chain *chain = block_getChain(block);
int start = cap_getCoordinate(cap);
int end = cap_getCoordinate(cap_getOtherSegmentCap(cap)) +1;
fprintf(fileHandle, "\t\t\t\t<blockName>%s</blockName>\n", cactusMisc_nameToString(block_getName(block)));
if(chain != NULL){
fprintf(fileHandle, "\t\t\t\t<chainName>%s</chainName>\n", cactusMisc_nameToString(chain_getName(chain)));
}else{
fprintf(fileHandle, "\t\t\t\t<chainName>NA</chainName>\n");
}
fprintf(fileHandle, "\t\t\t\t<level>%d</level>\n", level);
fprintf(fileHandle, "\t\t\t\t<start>%d</start>\n", start);
fprintf(fileHandle, "\t\t\t\t<end>%d</end>\n", end);
fprintf(fileHandle, "\t\t\t</block>\n");
st_logInfo("mapBlockToExon: start: %d, end: %d\n", start, end);
}
static int addAdjacenciesPP(Cap *cap1, Cap *cap2) {
assert(cap_getStrand(cap1) && cap_getStrand(cap2));
Sequence *sequence1 = cap_getSequence(cap1);
Sequence *sequence2 = cap_getSequence(cap2);
int64_t i = cactusMisc_nameCompare(sequence_getName(sequence1), sequence_getName(sequence2));
if (i == 0) {
int64_t j = cap_getCoordinate(cap1);
int64_t k = cap_getCoordinate(cap2);
i = j > k ? 1 : (j < k ? -1 : 0);
if (i == 0) {
assert(cap_getSegment(cap1) == cap_getSegment(cap2));
j = cap_getSide(cap1);
k = cap_getSide(cap2);
assert((j && !k) || (!j && k));
i = j ? -1 : 1;
}
}
return i;
}
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);
}
}
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;
}
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;
}
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);
}
int main(int argc, char *argv[]) {
st_setLogLevelFromString(argv[1]);
st_logDebug("Set up logging\n");
stKVDatabaseConf *kvDatabaseConf = stKVDatabaseConf_constructFromString(argv[2]);
CactusDisk *cactusDisk = cactusDisk_construct(kvDatabaseConf, 0);
stKVDatabaseConf_destruct(kvDatabaseConf);
st_logDebug("Set up the flower disk\n");
Name flowerName = cactusMisc_stringToName(argv[3]);
Flower *flower = cactusDisk_getFlower(cactusDisk, flowerName);
int64_t totalBases = flower_getTotalBaseLength(flower);
int64_t totalEnds = flower_getEndNumber(flower);
int64_t totalFreeEnds = flower_getFreeStubEndNumber(flower);
int64_t totalAttachedEnds = flower_getAttachedStubEndNumber(flower);
int64_t totalCaps = flower_getCapNumber(flower);
int64_t totalBlocks = flower_getBlockNumber(flower);
int64_t totalGroups = flower_getGroupNumber(flower);
int64_t totalChains = flower_getChainNumber(flower);
int64_t totalLinkGroups = 0;
int64_t maxEndDegree = 0;
int64_t maxAdjacencyLength = 0;
int64_t totalEdges = 0;
Flower_EndIterator *endIt = flower_getEndIterator(flower);
End *end;
while((end = flower_getNextEnd(endIt)) != NULL) {
assert(end_getOrientation(end));
if(end_getInstanceNumber(end) > maxEndDegree) {
maxEndDegree = end_getInstanceNumber(end);
}
stSortedSet *ends = stSortedSet_construct();
End_InstanceIterator *capIt = end_getInstanceIterator(end);
Cap *cap;
while((cap = end_getNext(capIt)) != NULL) {
if(cap_getSequence(cap) != NULL) {
Cap *adjacentCap = cap_getAdjacency(cap);
assert(adjacentCap != NULL);
End *adjacentEnd = end_getPositiveOrientation(cap_getEnd(adjacentCap));
stSortedSet_insert(ends, adjacentEnd);
int64_t adjacencyLength = cap_getCoordinate(cap) - cap_getCoordinate(adjacentCap);
if(adjacencyLength < 0) {
adjacencyLength *= -1;
}
assert(adjacencyLength >= 1);
if(adjacencyLength >= maxAdjacencyLength) {
maxAdjacencyLength = adjacencyLength;
}
}
}
end_destructInstanceIterator(capIt);
totalEdges += stSortedSet_size(ends);
if(stSortedSet_search(ends, end) != NULL) { //This ensures we count self edges twice, so that the division works.
totalEdges += 1;
}
stSortedSet_destruct(ends);
}
assert(totalEdges % 2 == 0);
flower_destructEndIterator(endIt);
Flower_GroupIterator *groupIt = flower_getGroupIterator(flower);
Group *group;
while((group = flower_getNextGroup(groupIt)) != NULL) {
if(group_getLink(group) != NULL) {
totalLinkGroups++;
}
}
flower_destructGroupIterator(groupIt);
printf("flower name: %" PRIi64 " total bases: %" PRIi64 " total-ends: %" PRIi64 " total-caps: %" PRIi64 " max-end-degree: %" PRIi64 " max-adjacency-length: %" PRIi64 " total-blocks: %" PRIi64 " total-groups: %" PRIi64 " total-edges: %" PRIi64 " total-free-ends: %" PRIi64 " total-attached-ends: %" PRIi64 " total-chains: %" PRIi64 " total-link groups: %" PRIi64 "\n",
flower_getName(flower), totalBases, totalEnds, totalCaps, maxEndDegree, maxAdjacencyLength, totalBlocks, totalGroups, totalEdges/2, totalFreeEnds, totalAttachedEnds, totalChains, totalLinkGroups);
return 0;
}
int mapGene(Cap *cap, int level, int exon, struct bed *gene, FILE *fileHandle){
/*
*Following cactus adjacencies, starting from 'cap', find regions that overlap with
*exons of input gene. Report chain relations of these regions with the exons.
*cap: current cap. Level = chain level. exon = exon number. gene = bed record of gene
*/
int64_t exonStart, exonEnd;
if(isStubCap(cap)){
Group *group = end_getGroup(cap_getEnd(cap));
Flower *nestedFlower = group_getNestedFlower(group);
if(nestedFlower != NULL){//recursive call
Cap *childCap = flower_getCap(nestedFlower, cap_getName(cap));
assert(childCap != NULL);
exon = mapGene(childCap, level + 1, exon, gene, fileHandle);
exonStart = gene->chromStarts->list[exon] + gene->chromStart;
exonEnd = exonStart + gene->blockSizes->list[exon];
}
}
cap = cap_getAdjacency(cap);
Cap *nextcap;
int64_t capCoor;
exonStart = gene->chromStarts->list[exon] + gene->chromStart;
exonEnd = exonStart + gene->blockSizes->list[exon];
Block *block = end_getBlock(cap_getEnd(cap));
if(block == NULL){
moveCapToNextBlock(&cap);
}
while(!isStubCap(cap) && exon < gene->blockCount){
End *cend = cap_getEnd(cap);
capCoor = cap_getCoordinate(cap);//Cap coordinate is always the coordinate on + strand
nextcap = cap_getAdjacency(cap_getOtherSegmentCap(cap));
st_logInfo("capCoor: %d, nextCap: %d, eStart: %d, eEnd: %d. Exon: %d\n",
capCoor, cap_getCoordinate(nextcap), exonStart, exonEnd, exon);
//keep moving if nextBlock Start is still upstream of current exon
if(cap_getCoordinate(nextcap) <= exonStart){
moveCapToNextBlock(&cap);
st_logInfo("Still upstream, nextcap <= exonStart. Move to next chainBlock\n");
}else if(capCoor >= exonEnd){//Done with current exon, move to next
st_logInfo("Done with current exon, move to next one\n\n");
fprintf(fileHandle, "\t\t</exon>\n");//end previous exon
exon++;
if(exon < gene->blockCount){
exonStart = gene->chromStarts->list[exon] + gene->chromStart;
exonEnd = exonStart + gene->blockSizes->list[exon];
fprintf(fileHandle, "\t\t<exon id=\"%d\" start=\"%" PRIi64 "\" end=\"%" PRIi64 "\">\n", exon, exonStart, exonEnd);
}
}else{//current exon overlaps with current block Or with lower level flower
Cap *oppcap = cap_getOtherSegmentCap(cap);
st_logInfo("Current exon overlaps with current block or with lower flower\n");
if(cap_getCoordinate(oppcap) >= exonStart && exonEnd > capCoor){
mapBlockToExon(cap, level, fileHandle);
if(exonEnd <= cap_getCoordinate(oppcap) + 1){
st_logInfo("Done with current exon, move to next one\n\n");
fprintf(fileHandle, "\t\t</exon>\n");//end previous exon
exon++;
if(exon < gene->blockCount){
exonStart = gene->chromStarts->list[exon] + gene->chromStart;
exonEnd = exonStart + gene->blockSizes->list[exon];
fprintf(fileHandle, "\t\t<exon id=\"%d\" start=\"%" PRIi64 "\" end=\"%" PRIi64 "\">\n", exon, exonStart, exonEnd);
}
continue;
}
}
//Traverse lower level flowers if exists
Group *group = end_getGroup(end_getOtherBlockEnd(cend));
Flower *nestedFlower = group_getNestedFlower(group);
if(nestedFlower != NULL){//recursive call
Cap *childCap = flower_getCap(nestedFlower, cap_getName(cap_getOtherSegmentCap(cap)));
assert(childCap != NULL);
exon = mapGene(childCap, level + 1, exon, gene, fileHandle);
exonStart = gene->chromStarts->list[exon] + gene->chromStart;
exonEnd = exonStart + gene->blockSizes->list[exon];
}
moveCapToNextBlock(&cap);
}
}
return exon;
}