Ejemplo n.º 1
0
static void convertInsertionMarkers(InsertionMarker * insertionMarkers,
				    InsertionMarker * veryLastMarker,
				    IDnum * chains)
{
	InsertionMarker *marker;
	Annotation *annot;

	for (marker = insertionMarkers; marker != veryLastMarker; marker++) {
		annot = marker->annot;

		if (getAnnotSequenceID(annot) > 0) {
			if (marker->isStart) {
				if (getStartID(annot) == 0)
					setStartID(annot,
						   chains
						   [getAnnotSequenceID
						    (annot)]);
				else
					setStartID(annot,
						   getStartID(annot) + 1);
			}
		} else {
			if (marker->isStart)
				setStartID(annot, -getStartID(annot));
			else {
				if (getFinishID(annot) == 0)
					setFinishID(annot,
						    -chains
						    [-getAnnotSequenceID
						     (annot)]);
				else
					setFinishID(annot,
						    -getFinishID(annot) -
						    1);
			}
		}
	}

	free(insertionMarkers);
}
Ejemplo n.º 2
0
static void threadSequenceThroughGraph(TightString * tString,
				       KmerOccurenceTable * kmerTable,
				       Graph * graph,
				       IDnum seqID, Category category,
				       boolean readTracking,
				       boolean double_strand,
				       ReferenceMapping * referenceMappings,
				       Coordinate referenceMappingCount,
				       IDnum refCount,
				       Annotation * annotations,
				       IDnum annotationCount,
				       boolean second_in_pair)
{
	Kmer word;
	Kmer antiWord;
	Coordinate readNucleotideIndex;
	Coordinate kmerIndex;
	KmerOccurence *kmerOccurence;
	int wordLength = getWordLength(graph);

	PassageMarkerI marker = NULL_IDX;
	PassageMarkerI previousMarker = NULL_IDX;
	Node *node = NULL;
	Node *previousNode = NULL;
	Coordinate coord = 0;
	Coordinate previousCoord = 0;
	Nucleotide nucleotide;
	boolean reversed;

	IDnum refID;
	Coordinate refCoord = 0;
	ReferenceMapping * refMap;
	Annotation * annotation = annotations;
	Coordinate index = 0;
	Coordinate uniqueIndex = 0;
	Coordinate annotIndex = 0;
	IDnum annotCount = 0;
	SmallNodeList * nodePile = NULL;

	// Neglect any string shorter than WORDLENGTH :
	if (getLength(tString) < wordLength)
		return;

	clearKmer(&word);
	clearKmer(&antiWord);

	// Fill in the initial word : 
	for (readNucleotideIndex = 0;
	     readNucleotideIndex < wordLength - 1; readNucleotideIndex++) {
		nucleotide = getNucleotide(readNucleotideIndex, tString);
		pushNucleotide(&word, nucleotide);
		if (double_strand || second_in_pair) {
#ifdef COLOR
			reversePushNucleotide(&antiWord, nucleotide);
#else
			reversePushNucleotide(&antiWord, 3 - nucleotide);
#endif
		}
	}

	// Go through sequence
	while (readNucleotideIndex < getLength(tString)) {
		nucleotide = getNucleotide(readNucleotideIndex++, tString);
		pushNucleotide(&word, nucleotide);
		if (double_strand || second_in_pair) {
#ifdef COLOR
			reversePushNucleotide(&antiWord, nucleotide);
#else
			reversePushNucleotide(&antiWord, 3 - nucleotide);
#endif
		}

		// Update annotation if necessary
		if (annotCount < annotationCount && annotIndex == getAnnotationLength(annotation)) {
			annotation = getNextAnnotation(annotation);
			annotCount++;
			annotIndex = 0;
		}

		// Search for reference mapping
		if (category == REFERENCE) {
			if (referenceMappings) 
				refMap = findReferenceMapping(seqID, index, referenceMappings, referenceMappingCount);
			else 
				refMap = NULL;

			if (refMap) {
				node = getNodeInGraph(graph, refMap->nodeID);
				if (refMap->nodeID > 0) {
					coord = refMap->nodeStart + (index - refMap->referenceStart);
				} else {
					coord = getNodeLength(node) - refMap->nodeStart - refMap->length + (index - refMap->referenceStart);
				}
			} else  {
				node = NULL;
				if (previousNode)
					break;
			}
		}
		// Search for reference-based mapping
		else if (annotCount < annotationCount && uniqueIndex >= getPosition(annotation) && getAnnotSequenceID(annotation) <= refCount && getAnnotSequenceID(annotation) >= -refCount) {
			refID = getAnnotSequenceID(annotation);
			if (refID > 0)
				refCoord = getStart(annotation) + annotIndex; 
			else
				refCoord = getStart(annotation) - annotIndex; 
			
			refMap = findReferenceMapping(refID, refCoord, referenceMappings, referenceMappingCount);
			// If success
			if (refMap) {
				if (refID > 0) {
					node = getNodeInGraph(graph, refMap->nodeID);
					if (refMap->nodeID > 0) {
						coord = refMap->nodeStart + (refCoord - refMap->referenceStart);
					} else {
						coord = getNodeLength(node) - refMap->nodeStart - refMap->length + (refCoord - refMap->referenceStart);
					}
				} else {
					node = getNodeInGraph(graph, -refMap->nodeID);
					if (refMap->nodeID > 0) {
						coord =  getNodeLength(node) - refMap->nodeStart - (refCoord - refMap->referenceStart) - 1;
					} else {
						coord = refMap->nodeStart + refMap->length - (refCoord - refMap->referenceStart) - 1;
					}
				}
			} else  {
				node = NULL;
				if (previousNode)
					break;
			}
		}		
		// Search in table
		else {
			reversed = false;
			if (double_strand) {
				if (compareKmers(&word, &antiWord) <= 0) {
					kmerOccurence =
					findKmerInKmerOccurenceTable(&word,
								       kmerTable);
				} else { 
					kmerOccurence =
					       findKmerInKmerOccurenceTable(&antiWord,
						kmerTable);
					reversed = true;
				}
			} else {
				if (!second_in_pair) {
					kmerOccurence =
					findKmerInKmerOccurenceTable(&word,
								       kmerTable);
				} else { 
					kmerOccurence =
					       findKmerInKmerOccurenceTable(&antiWord,
						kmerTable);
					reversed = true;
				}
			}
			
			if (kmerOccurence) {
				if (!reversed) {
					node = getNodeInGraph(graph, getKmerOccurenceNodeID(kmerOccurence));
					coord = getKmerOccurencePosition(kmerOccurence);
				} else {
					node = getNodeInGraph(graph, -getKmerOccurenceNodeID(kmerOccurence));
					coord = getNodeLength(node) - getKmerOccurencePosition(kmerOccurence) - 1;
				}
			} else {
				node = NULL;
				if (previousNode) 
					break;
			}
		}

		// Increment positions
		if (annotCount < annotationCount && uniqueIndex >= getPosition(annotation)) 
			annotIndex++;
		else
			uniqueIndex++;

		// Fill in graph
		if (node)
		{
#ifdef OPENMP
			lockNode(node);
#endif
			kmerIndex = readNucleotideIndex - wordLength;

			if (previousNode == node
			    && previousCoord == coord - 1) {
				if (category / 2 >= CATEGORIES) {
					setPassageMarkerFinish(marker,
							       kmerIndex +
							       1);
					setFinishOffset(marker,
							getNodeLength(node)
							- coord - 1);
				} else {
#ifndef SINGLE_COV_CAT
					incrementVirtualCoverage(node, category / 2, 1);
					incrementOriginalVirtualCoverage(node, category / 2, 1);
#else
					incrementVirtualCoverage(node, 1);
#endif
				}
#ifdef OPENMP
				unLockNode(node);
#endif
			} else {
				if (category / 2 >= CATEGORIES) {
					marker =
					    newPassageMarker(seqID,
							     kmerIndex,
							     kmerIndex + 1,
							     coord,
							     getNodeLength
							     (node) -
							     coord - 1);
					transposePassageMarker(marker,
							       node);
					connectPassageMarkers
					    (previousMarker, marker,
					     graph);
					previousMarker = marker;
				} else {
					if (readTracking) {
						if (!isNodeMemorized(node, nodePile)) {
							addReadStart(node,
								     seqID,
								     coord,
								     graph,
								     kmerIndex);
							memorizeNode(node, &nodePile);
						} else {
							blurLastShortReadMarker
							    (node, graph);
						}
					}

#ifndef SINGLE_COV_CAT
					incrementVirtualCoverage(node, category / 2, 1);
					incrementOriginalVirtualCoverage(node, category / 2, 1);
#else
					incrementVirtualCoverage(node, 1);
#endif
				}
#ifdef OPENMP
				lockTwoNodes(node, previousNode);
#endif
				createArc(previousNode, node, graph);
#ifdef OPENMP
				unLockTwoNodes(node, previousNode);
#endif
			}

			previousNode = node;
			previousCoord = coord;
		}
		index++;
	}

	if (readTracking && category / 2 < CATEGORIES)
		unMemorizeNodes(&nodePile);
}
Ejemplo n.º 3
0
// Creates insertion marker lists 
static void
setInsertionMarkers(RoadMapArray * rdmaps,
		    IDnum * markerCounters,
		    InsertionMarker ** veryLastMarker,
		    InsertionMarker ** insertionMarkers)
{
	IDnum sequenceCounter = rdmaps->length;
	IDnum sequenceIndex, sequenceIndex2;
	Coordinate totalCount = 0;
	RoadMap *rdmap;
	Annotation *annot = rdmaps->annotations;
	InsertionMarker *nextMarker, *newMarker;
	IDnum annotIndex, lastAnnotIndex;
	InsertionMarker **insMarkers =
	    callocOrExit(rdmaps->length + 1, InsertionMarker *);
	// Counting insertion markers
	for (sequenceIndex = 1; sequenceIndex < sequenceCounter + 1;
	     sequenceIndex++) {
		//velvetLog("Going through sequence %d\n", sequenceIndex);
		rdmap = getRoadMapInArray(rdmaps, sequenceIndex - 1);
		lastAnnotIndex = getAnnotationCount(rdmap);

		// Set insertion markers in previous sequences :

		for (annotIndex = 0; annotIndex < lastAnnotIndex;
		     annotIndex++) {
			if (getAnnotSequenceID(annot) > 0) {
				markerCounters[getAnnotSequenceID(annot)]
				    += 2;
			} else {
				markerCounters[-getAnnotSequenceID(annot)]
				    += 2;
			}
			totalCount += 2;
			annot = getNextAnnotation(annot);
		}
	}

	// Allocating space
	*insertionMarkers = callocOrExit(totalCount, InsertionMarker);
	*veryLastMarker = *insertionMarkers + totalCount;

	// Pointing each node to its space      
	nextMarker = *insertionMarkers;
	for (sequenceIndex = 1; sequenceIndex < sequenceCounter + 1;
	     sequenceIndex++) {
		insMarkers[sequenceIndex] = nextMarker;
		nextMarker = nextMarker + markerCounters[sequenceIndex];
		markerCounters[sequenceIndex] = 0;
	}

	// Filling up space with data
	annot = rdmaps->annotations;
	for (sequenceIndex = 1; sequenceIndex < sequenceCounter + 1;
	     sequenceIndex++) {
		//velvetLog("Going through sequence %d\n", sequenceIndex);
		rdmap = getRoadMapInArray(rdmaps, sequenceIndex - 1);
		lastAnnotIndex = getAnnotationCount(rdmap);

		// Set insertion markers in previous sequences :

		for (annotIndex = 0; annotIndex < lastAnnotIndex;
		     annotIndex++) {
			sequenceIndex2 = getAnnotSequenceID(annot);
			if (sequenceIndex2 > 0) {
				newMarker =
				    insMarkers[sequenceIndex2] +
				    (markerCounters[sequenceIndex2])++;
				newMarker->annot = annot;
				newMarker->isStart = true;

				newMarker =
				    insMarkers[sequenceIndex2] +
				    (markerCounters[sequenceIndex2])++;
				newMarker->annot = annot;
				newMarker->isStart = false;
			} else {
				incrementAnnotationCoordinates(annot);

				newMarker =
				    insMarkers[-sequenceIndex2] +
				    (markerCounters[-sequenceIndex2])++;
				newMarker->annot = annot;
				newMarker->isStart = true;

				newMarker =
				    insMarkers[-sequenceIndex2] +
				    (markerCounters[-sequenceIndex2])++;
				newMarker->annot = annot;
				newMarker->isStart = false;
			}
			annot = getNextAnnotation(annot);
		}
	}

	orderInsertionMarkers(insMarkers, markerCounters, rdmaps);
	free(insMarkers);
}
Ejemplo n.º 4
0
static void ghostThreadSequenceThroughGraph(TightString * tString,
					    KmerOccurenceTable *
					    kmerTable, Graph * graph,
					    IDnum seqID, Category category,
					    boolean readTracking,
					    boolean double_strand,
					    ReferenceMapping * referenceMappings,
					    Coordinate referenceMappingCount,
					    IDnum refCount,
					    Annotation * annotations,
					    IDnum annotationCount,
					    boolean second_in_pair)
{
	Kmer word;
	Kmer antiWord;
	Coordinate readNucleotideIndex;
	KmerOccurence *kmerOccurence;
	int wordLength = getWordLength(graph);
	Nucleotide nucleotide;
	IDnum refID;
	Coordinate refCoord;
	ReferenceMapping * refMap = NULL;
	Coordinate uniqueIndex = 0;
	Coordinate annotIndex = 0;
	IDnum annotCount = 0;
	boolean reversed;
	SmallNodeList * nodePile = NULL;
	Annotation * annotation = annotations;

	Node *node;
	Node *previousNode = NULL;

	// Neglect any read which will not be short paired
	if ((!readTracking && category % 2 == 0)
	    || category / 2 >= CATEGORIES)
		return;

	// Neglect any string shorter than WORDLENGTH :
	if (getLength(tString) < wordLength)
		return;

	// Verify that all short reads are reasonnably short
	if (getLength(tString) > USHRT_MAX) {
		velvetLog("Short read of length %lli, longer than limit %i\n",
			  (long long) getLength(tString), SHRT_MAX);
		velvetLog("You should better declare this sequence as long, because it genuinely is!\n");
		exit(1);
	}

	clearKmer(&word);
	clearKmer(&antiWord);

	// Fill in the initial word :
	for (readNucleotideIndex = 0;
	     readNucleotideIndex < wordLength - 1; readNucleotideIndex++) {
		nucleotide = getNucleotide(readNucleotideIndex, tString);
		pushNucleotide(&word, nucleotide);
		if (double_strand || second_in_pair) {
#ifdef COLOR
			reversePushNucleotide(&antiWord, nucleotide);
#else
			reversePushNucleotide(&antiWord, 3 - nucleotide);
#endif
		}
	}

	// Go through sequence
	while (readNucleotideIndex < getLength(tString)) {
		// Shift word:
		nucleotide = getNucleotide(readNucleotideIndex++, tString);
		pushNucleotide(&word, nucleotide);
		if (double_strand || second_in_pair) {
#ifdef COLOR
			reversePushNucleotide(&antiWord, nucleotide);
#else
			reversePushNucleotide(&antiWord, 3 - nucleotide);
#endif
		}

		// Update annotation if necessary
		if (annotCount < annotationCount && annotIndex == getAnnotationLength(annotation)) {
			annotation = getNextAnnotation(annotation);
			annotCount++;
			annotIndex = 0;
		}

		// Search for reference mapping
 		if (annotCount < annotationCount && uniqueIndex >= getPosition(annotation) && getAnnotSequenceID(annotation) <= refCount && getAnnotSequenceID(annotation) >= -refCount) {
			refID = getAnnotSequenceID(annotation);
			if (refID > 0)
				refCoord = getStart(annotation) + annotIndex;
			else
				refCoord = getStart(annotation) - annotIndex;
			
			refMap = findReferenceMapping(refID, refCoord, referenceMappings, referenceMappingCount);
			// If success
			if (refMap) {
				if (refID > 0) 
					node = getNodeInGraph(graph, refMap->nodeID);
				else
					node = getNodeInGraph(graph, -refMap->nodeID);
			} else  {
				node = NULL;
				if (previousNode)
					break;
			}
		}
		// if not.. look in table
		else {
			reversed = false;
			if (double_strand) {
				if (compareKmers(&word, &antiWord) <= 0) {
					kmerOccurence =
					findKmerInKmerOccurenceTable(&word,
								       kmerTable);
				} else { 
					kmerOccurence =
					       findKmerInKmerOccurenceTable(&antiWord,
						kmerTable);
					reversed = true;
				}
			} else {
				if (!second_in_pair) {
					kmerOccurence =
					findKmerInKmerOccurenceTable(&word,
								       kmerTable);
				} else { 
					kmerOccurence =
					       findKmerInKmerOccurenceTable(&antiWord,
						kmerTable);
					reversed = true;
				}
			}
			
			if (kmerOccurence) {
				if (!reversed) 
					node = getNodeInGraph(graph, getKmerOccurenceNodeID(kmerOccurence));
				else
					node = getNodeInGraph(graph, -getKmerOccurenceNodeID(kmerOccurence));
			} else {
				node = NULL;
				if (previousNode) 
					break;
			}

		}

		if (annotCount < annotationCount && uniqueIndex >= getPosition(annotation))
			annotIndex++;
		else
			uniqueIndex++;

		previousNode = node;

		// Fill in graph
		if (node && !isNodeMemorized(node, nodePile))
		{
#ifdef OPENMP
			lockNode(node);
#endif
			incrementReadStartCount(node, graph);
#ifdef OPENMP
			unLockNode(node);
#endif
			memorizeNode(node, &nodePile);
		}
	}

	unMemorizeNodes(&nodePile);
}