static void computePartialReadToNodeMapping(IDnum nodeID,
ReadOccurence ** readNodes,
IDnum * readNodeCounts,
boolean * readMarker,
ReadSet * reads)
{
ShortReadMarker *shortMarker;
IDnum index, readIndex;
ReadOccurence *readArray, *readOccurence;
Node *node = getNodeInGraph(graph, nodeID);
ShortReadMarker *nodeArray = getNodeReads(node, graph);
IDnum nodeReadCount = getNodeReadCount(node, graph);
PassageMarkerI marker;
for (index = 0; index < nodeReadCount; index++) {
shortMarker = getShortReadMarkerAtIndex(nodeArray, index);
readIndex = getShortReadMarkerID(shortMarker);
readArray = readNodes[readIndex];
readOccurence = &readArray[readNodeCounts[readIndex]];
readOccurence->nodeID = nodeID;
readOccurence->position =
getShortReadMarkerPosition(shortMarker);
readOccurence->offset =
getShortReadMarkerOffset(shortMarker);
readNodeCounts[readIndex]++;
}
for (marker = getMarker(node); marker != NULL_IDX;
marker = getNextInNode(marker)) {
readIndex = getPassageMarkerSequenceID(marker);
if (readIndex <= 0 || reads->categories[readIndex - 1] == REFERENCE)
continue;
if (!readMarker[readIndex]) {
readArray = readNodes[readIndex];
readOccurence =
&readArray[readNodeCounts[readIndex]];
readOccurence->nodeID = nodeID;
readOccurence->position = getStartOffset(marker);
readOccurence->offset =
getPassageMarkerStart(marker);
readNodeCounts[readIndex]++;
readMarker[readIndex] = true;
} else {
readArray = readNodes[readIndex];
readOccurence =
&readArray[readNodeCounts[readIndex] - 1];
readOccurence->position = -1;
readOccurence->offset = -1;
}
}
for (marker = getMarker(node); marker != NULL_IDX;
marker = getNextInNode(marker)) {
readIndex = getPassageMarkerSequenceID(marker);
if (readIndex > 0)
readMarker[readIndex] = false;
}
}
static IDnum *computeReadToNodeCounts()
{
IDnum readIndex, nodeIndex;
IDnum maxNodeIndex = 2 * nodeCount(graph) + 1;
IDnum maxReadIndex = sequenceCount(graph) + 1;
IDnum *readNodeCounts = callocOrExit(maxReadIndex, IDnum);
boolean *readMarker = callocOrExit(maxReadIndex, boolean);
ShortReadMarker *nodeArray, *shortMarker;
PassageMarkerI marker;
Node *node;
IDnum nodeReadCount;
//puts("Computing read to node mapping array sizes");
for (nodeIndex = 0; nodeIndex < maxNodeIndex; nodeIndex++) {
node = getNodeInGraph(graph, nodeIndex - nodeCount(graph));
if (node == NULL)
continue;
// Short reads
if (readStartsAreActivated(graph)) {
nodeArray = getNodeReads(node, graph);
nodeReadCount = getNodeReadCount(node, graph);
for (readIndex = 0; readIndex < nodeReadCount; readIndex++) {
shortMarker =
getShortReadMarkerAtIndex(nodeArray,
readIndex);
readNodeCounts[getShortReadMarkerID
(shortMarker)]++;
}
}
// Long reads
for (marker = getMarker(node); marker != NULL_IDX;
marker = getNextInNode(marker)) {
readIndex = getPassageMarkerSequenceID(marker);
if (readIndex < 0)
continue;
if (readMarker[readIndex])
continue;
readNodeCounts[readIndex]++;
readMarker[readIndex] = true;
}
// Clean up marker array
for (marker = getMarker(node); marker != NULL_IDX;
marker = getNextInNode(marker)) {
readIndex = getPassageMarkerSequenceID(marker);
if (readIndex > 0)
readMarker[readIndex] = false;
}
}
free(readMarker);
return readNodeCounts;
}
static void projectFromNode(IDnum nodeID,
ReadOccurence ** readNodes,
IDnum * readNodeCounts,
IDnum * readPairs, Category * cats,
boolean * dubious, Coordinate * lengths)
{
IDnum index;
ShortReadMarker *nodeArray, *shortMarker;
PassageMarker *marker;
Node *node;
IDnum nodeReadCount;
node = getNodeInGraph(graph, nodeID);
if (node == NULL || !getUniqueness(node))
return;
nodeArray = getNodeReads(node, graph);
nodeReadCount = getNodeReadCount(node, graph);
for (index = 0; index < nodeReadCount; index++) {
shortMarker = getShortReadMarkerAtIndex(nodeArray, index);
if (dubious[getShortReadMarkerID(shortMarker) - 1])
continue;
projectFromShortRead(node, shortMarker, readPairs, cats,
readNodes, readNodeCounts, lengths);
}
for (marker = getMarker(node); marker != NULL;
marker = getNextInNode(marker)) {
if (getPassageMarkerSequenceID(marker) > 0)
projectFromLongRead(node, marker, readPairs, cats,
readNodes, readNodeCounts,
lengths);
}
}
static void projectFromLongRead(Node * node, PassageMarkerI marker,
IDnum * readPairs, Category * cats,
ReadOccurence ** readNodes,
IDnum * readNodeCounts,
ShortLength * lengths, boolean weight)
{
IDnum index;
IDnum readIndex = getPassageMarkerSequenceID(marker);
ReadOccurence *readArray;
IDnum readPairIndex;
Category cat;
Coordinate position = getStartOffset(marker);
Coordinate offset = getPassageMarkerStart(marker);
Coordinate length =
lengths[getPassageMarkerSequenceID(marker) - 1];
Coordinate insertLength;
double insertVariance;
// Going through single-read information
if (readNodeCounts[readIndex] > 1 && position >= 0) {
readArray = readNodes[readIndex];
for (index = 0; index < readNodeCounts[readIndex]; index++)
projectFromSingleRead(node, &readArray[index],
position, offset, length,
weight);
}
// Going through paired read information
if (readPairs == NULL)
return;
readPairIndex = readPairs[readIndex - 1] + 1;
if (readPairIndex == 0)
return;
cat = cats[readIndex - 1];
insertLength = getInsertLength(graph, cat);
insertVariance = getInsertLength_var(graph, cat);
readArray = readNodes[readPairIndex];
for (index = 0; index < readNodeCounts[readPairIndex]; index++)
projectFromReadPair(node, &readArray[index], position,
offset, insertLength, insertVariance,
weight);
}
TightString *expandPassageMarker(PassageMarkerI marker,
TightString ** sequences, int WORDLENGTH)
{
if (getPassageMarkerSequenceID(marker) > 0)
return readPositivePassageMarker(marker, sequences,
WORDLENGTH);
else
return readNegativePassageMarker(marker, sequences);
}
static TightString *readPositivePassageMarker(PassageMarkerI marker,
TightString ** seqs,
int WORDLENGTH)
{
Coordinate index;
Nucleotide nucleotide;
TightString *tString =
seqs[getPassageMarkerSequenceID(marker) - 1];
TightString *res = newTightString(getPassageMarkerLength(marker));
for (index = 0; index < getLength(tString); index++) {
nucleotide =
getNucleotide(getPassageMarkerStart(marker) + index +
WORDLENGTH - 1, tString);
writeNucleotideAtPosition(nucleotide, index, res);
}
return res;
}