C++ reverseComplement函数代码示例

void tokenizePatternFile(std::ifstream& in) {
    // tokenize a line from the pattern file.  The first part will be the pattern and the second
    // part is the file to write to.

    std::string lineptr;

    while(in.good()) {
        std::getline(in, lineptr);
        if(lineptr.empty()) {
            continue;
        }
        std::vector<std::string> fields;
        split(fields, lineptr, "\t");
        switch(fields.size()) {
            case 0:
                break;
            case 1:
                manager.add(fields[0]);
                if(opts.r_flag) {
                    std::string rcpattern = fields[0];
                    reverseComplement(rcpattern);
                    manager.add(rcpattern);
                }
                break;
            default:
                manager.add(fields[0], fields[1]);
                if(opts.r_flag) {
                    std::string rcpattern = fields[0];
                    reverseComplement(rcpattern);
                    manager.add(rcpattern, fields[1]);
                }
                break;
        }
    }
}

static void showOverlap(const bam1_t *leftBam, const bam1_t *rightBam)
/* If the two reads overlap, show how. */
{
const bam1_core_t *leftCore = &(leftBam->core), *rightCore = &(rightBam->core);
int leftStart = leftCore->pos, rightStart = rightCore->pos;
int leftLen = bamGetTargetLength(leftBam), rightLen = bamGetTargetLength(rightBam);
char *leftSeq = bamGetQuerySequence(leftBam, useStrand);
char *rightSeq = bamGetQuerySequence(rightBam, useStrand);
if (useStrand && bamIsRc(leftBam))
    reverseComplement(leftSeq, strlen(leftSeq));
if (useStrand && bamIsRc(rightBam))
    reverseComplement(rightSeq, strlen(rightSeq));
if ((rightStart > leftStart && leftStart + leftLen > rightStart) ||
    (leftStart > rightStart && rightStart+rightLen > leftStart))
    {
    int leftClipLow, rightClipLow;
    bamGetSoftClipping(leftBam, &leftClipLow, NULL, NULL);
    bamGetSoftClipping(rightBam, &rightClipLow, NULL, NULL);
    leftStart -= leftClipLow;
    rightStart -= rightClipLow;
    printf("<B>Note: End read alignments overlap:</B><BR>\n<PRE><TT>");
    int i = leftStart - rightStart;
    while (i-- > 0)
	putc(' ', stdout);
    puts(leftSeq);
    i = rightStart - leftStart;
    while (i-- > 0)
	putc(' ', stdout);
    puts(rightSeq);
    puts("</TT></PRE>");
    }
}

void loadIfNewSeq(char *nibDir, char *newName, char strand, 
	char **pName, struct dnaSeq **pSeq, char *pStrand)
/* Load sequence unless it is already loaded.  Reverse complement
 * if necessary. */
{
struct dnaSeq *seq;
if (sameString(newName, *pName))
    {
    if (strand != *pStrand)
        {
	seq = *pSeq;
	reverseComplement(seq->dna, seq->size);
	*pStrand = strand;
	}
    }
else
    {
    char fileName[512];
    freeDnaSeq(pSeq);
    snprintf(fileName, sizeof(fileName), "%s/%s.nib", nibDir, newName);
    *pName = newName;
    *pSeq = seq = nibLoadAllMasked(NIB_MASK_MIXED, fileName);
    *pStrand = strand;
    if (strand == '-')
        reverseComplement(seq->dna, seq->size);
    uglyf("Loaded %d bases in %s\n", seq->size, fileName);
    }
}

void loadFaSeq(struct hash *faHash, char *newName, char strand, 
	char **pName, struct dnaSeq **pSeq, char *pStrand)
/* retrieve sequence from hash.  Reverse complement
 * if necessary. */
{
struct dnaSeq *seq;
if (sameString(newName, *pName))
    {
    if (strand != *pStrand)
        {
	seq = *pSeq;
	reverseComplement(seq->dna, seq->size);
	*pStrand = strand;
	}
    }
else
    {
    *pName = newName;
    *pSeq = seq = hashFindVal(faHash, newName);
    *pStrand = strand;
    if (strand == '-')
        reverseComplement(seq->dna, seq->size);
    verbose(1, "Loaded %d bases from %s fa\n", seq->size, newName);
    }
}

static void loadFaSeq(struct hash *faHash, char *newName, char strand,
	char **pName, struct dnaSeq **pSeq, char *pStrand, char *fastaFileName)
/* retrieve sequence from hash.  Reverse complement
 * if necessary. */
{
struct dnaSeq *seq;
if (sameString(newName, *pName))
    {
    if (strand != *pStrand)
        {
	seq = *pSeq;
	reverseComplement(seq->dna, seq->size);
	*pStrand = strand;
	}
    }
else
    {
    *pName = newName;
    *pSeq = seq = hashFindVal(faHash, newName);
    if (NULL == seq)
        errAbort("ERROR: can not find sequence name '%s' from fasta file '%s'\n", newName, fastaFileName);
    *pStrand = strand;
    if (strand == '-')
        reverseComplement(seq->dna, seq->size);
    verbose(1, "Loaded %d bases from %s fa\n", seq->size, newName);
    }
}

static void makeProfile(DNA *oligo, int oligoSize, int mismatchesAllowed, 
    struct seqList *seqList, boolean considerRc, double profile[16][4])
/* Scan through file counting up things that match oligo to within
 * mismatch tolerance, and use these counts to build up a profile. */
{
int counts[16][4];
int total = 0;
double invTotal;
int i,j;
int seqCount = 0;
struct seqList *seqEl;
DNA rcOligo[17];

if (considerRc)
    {
    assert(oligoSize < sizeof(rcOligo));
    memcpy(rcOligo, oligo, oligoSize);
    reverseComplement(rcOligo, oligoSize);
    }

zeroBytes(counts, sizeof(counts));
for (seqEl = seqList; seqEl != NULL; seqEl = seqEl->next)
    {
    struct dnaSeq *seq = seqEl->seq;
    DNA *dna = seq->dna;
    int size = seq->size;
    int endIx = size-oligoSize;

    ++seqCount;
    for (i=0; i<=endIx; ++i)
        {
        DNA *target = dna+i;
        if (allGoodBases(target, oligoSize))
            {
            if (mismatchCount(oligo, target, oligoSize) <= mismatchesAllowed)
                {
                ++total;
                for (j=0; j<oligoSize; ++j)
                    counts[j][ntVal[(int)target[j]]] += 1;
                }
            if (considerRc && mismatchCount(rcOligo, target, oligoSize) <= mismatchesAllowed)
                {
                ++total;
                reverseComplement(target, oligoSize);
                for (j=0; j<oligoSize; ++j)
                    counts[j][ntVal[(int)target[j]]] += 1;
                reverseComplement(target, oligoSize);
                }
            } 
        }
    }
invTotal = 1.0/total;
for (i=0; i<oligoSize; ++i)
    {
    for (j=0; j<4; ++j)
        {
        profile[i][j] = invTotal * counts[i][j];
        }
    }
}

struct dnaSeq *gfiExpandAndLoadCached(struct gfRange *range, 
	struct hash *tFileCache, char *tSeqDir, int querySize, 
	int *retTotalSeqSize, boolean respectFrame, boolean isRc, int expansion)
/* Expand range to cover an additional expansion bases on either side.
 * Load up target sequence and return. (Done together because don't
 * know target size before loading.) */
{
struct dnaSeq *target = NULL;
char fileName[PATH_LEN+256];

safef(fileName, sizeof(fileName), "%s/%s", tSeqDir, range->tName);
if (nibIsFile(fileName))
    {
    struct nibInfo *nib = hashFindVal(tFileCache, fileName);
    if (nib == NULL)
        {
	nib = nibInfoNew(fileName);
	hashAdd(tFileCache, fileName, nib);
	}
    if (isRc)
	reverseIntRange(&range->tStart, &range->tEnd, nib->size);
    gfiExpandRange(range, querySize, nib->size, respectFrame, isRc, expansion);
    target = nibLdPart(fileName, nib->f, nib->size, 
    	range->tStart, range->tEnd - range->tStart);
    if (isRc)
	{
	reverseComplement(target->dna, target->size);
	reverseIntRange(&range->tStart, &range->tEnd, nib->size);
	}
    *retTotalSeqSize = nib->size;
    }
else
    {
    struct twoBitFile *tbf = NULL;
    char *tSeqName = strchr(fileName, ':');
    int tSeqSize = 0;
    if (tSeqName == NULL)
        errAbort("No colon in .2bit response from gfServer");
    *tSeqName++ = 0;
    tbf = hashFindVal(tFileCache, fileName);
    if (tbf == NULL)
        {
	tbf = twoBitOpen(fileName);
	hashAdd(tFileCache, fileName, tbf);
	}
    tSeqSize = twoBitSeqSize(tbf, tSeqName);
    if (isRc)
	reverseIntRange(&range->tStart, &range->tEnd, tSeqSize);
    gfiExpandRange(range, querySize, tSeqSize, respectFrame, isRc, expansion);
    target = twoBitReadSeqFragLower(tbf, tSeqName, range->tStart, range->tEnd);
    if (isRc)
	{
	reverseComplement(target->dna, target->size);
	reverseIntRange(&range->tStart, &range->tEnd, tSeqSize);
	}
    *retTotalSeqSize = tSeqSize;
    }
return target;
}

void ReadsLayout::print(size_t index, ostream &out, bool dir, unsigned int start, unsigned int maxD, Pairing *P) {

    if (getNext(index) != 0) {
        cerr << "void ReadsLayout::print(size_t index) problem\n";
        sendBugReportPlease(cerr);
    }
    if (!dir)
        index = reverseComplement(index);

    size_t p = getBegin(index);
    size_t tmp;

    do {
        
        unsigned int position=getPosition(p);
        
        if (position > maxD)
            break;

        if (position < start)
        {
            tmp = p;
            p = getNext(p);
            continue;
        }

        unsigned int pairedRead=0;
        unsigned int pairedNode=0;
        int lib=0;

        if (P->getNLibrary() != 0)
        {
            pairedRead = P->getPairing(p);
            pairedNode = getNodeId(pairedRead);
            lib = P->getPeLibraryID(p);
        }
        
        if (getDirection(p))
            out << '>';
        else
            out << '<';

        for (int i = 0; i < getPosition(p) % 120; i++)
            out << " ";
        if (getDirection(p))
            out << getDirectRead(p) << " " << p << ' ' << lib << ' ' << pairedNode << '\n';
        else
            out << getReverseRead(p) << " " << p << ' ' << lib << ' ' << pairedNode << '\n';
        
        tmp = p;
        p = getNext(p);

    } while (tmp != index);

    out << flush;

    if (!dir) //back to initial direction
        index = reverseComplement(index);
}

static void makeDirFasta(char *regionsFile, char *hg18FastaFile, char *dir, int num) {
	FILE *fp, *sq;
	char buf[500], dirName[500], seqName[500], chr1[500], chr2[500];
	int b1, e1, b2, e2, i, len;
	char ori1, ori2;
	struct hash *seqHash = NULL;
	struct dnaSeq *seq1, *seq2;
	struct stat st;
	DNA *s1, *s2;

	seqHash = faReadAllIntoHash(hg18FastaFile, dnaUpper);
	if (stat(dir, &st) != 0)
		do_cmd("mkdir %s", dir);

	fp = mustOpen(regionsFile, "r");
	i = 0;
	while (fgets(buf, 500, fp)) {
		if (sscanf(buf, "%[^:]:%d-%d %[^:]:%d-%d [%c %c]", chr1, &b1, &e1, chr2, &b2, &e2, &ori1, &ori2) != 8)
			errAbort("error: %s", buf);
		++i;
		if (i != num) 
			continue;
		sprintf(dirName, "%s/R%d", dir, i);
		if (stat(dirName, &st) != 0)
			do_cmd("mkdir %s", dir);
		sprintf(seqName, "%s/ref.fa", dirName);
		sq = mustOpen(seqName, "w");
		fprintf(sq, ">%s:%d-%d+%s:%d-%d[%c%c]\n", chr1, b1, e1, chr2, b2, e2, ori1, ori2);
		seq1 = (struct dnaSeq *)hashFindVal(seqHash, chr1);
		assert(e1 <= seq1->size);
		len = e1 - b1 + 1;
		if (ori1 == '-') {
			s1 = cloneStringZExt(seq1->dna + b1 - 1, len, len+1);
			reverseComplement(s1, len);
			writeSeqWithBreaks(sq, s1, len, 80);
			freeMem(s1);
		}
		else
			writeSeqWithBreaks(sq, seq1->dna + b1 - 1, e1 - b1 + 1, 80);
		seq2 = (struct dnaSeq *)hashFindVal(seqHash, chr2);
		assert(e2 <= seq2->size);
		len = e2 - b2 + 1;
		if (ori2 == '-') {
			s2 = cloneStringZExt(seq2->dna + b2 - 1, len, len+1);
			reverseComplement(s2, len);
			writeSeqWithBreaks(sq, s2, len, 80);
			freeMem(s2);
		}
		else
			writeSeqWithBreaks(sq, seq2->dna + b2 - 1, e2 - b2 + 1, 80);
		fclose(sq);
	}
	fclose(fp);
	//FIXME: free space
} 

void CommonTest::test_reverseComplement() {
  // Case 1: upper case
  std::string nucs = "ACGATCGTGTCATGCNNACCACG";
  std::string rev = reverseComplement(nucs);
  CPPUNIT_ASSERT_MESSAGE("Incorrect reverse complement", rev == "CGTGGTNNGCATGACACGATCGT");

  // Case 2: lower case
  nucs = "acgaccacagctacgacnacgactan";
  rev = reverseComplement(nucs);
  CPPUNIT_ASSERT_MESSAGE("Incorrect reverse complement", rev == "NTAGTCGTNGTCGTAGCTGTGGTCGT");
}

void showTargetRange(struct xaAli *xa, int tOff, int tLen, char strand, boolean showSym)
/* Display a range of xa, indexed by target. */
{
char *hSym = xa->hSym, *qSym = xa->qSym, *tSym = xa->tSym;
int symCount = xa->symCount;
int tPos = 0;
int i = 0;
int j;
int maxLineLen = 50;
int lineLen;
int startIx;
int fullLen;
int endIx;

/* Figure out starting and ending positions taking inserts in target
 * into account. */
startIx = lenWithDashes(tSym, tOff);
fullLen = lenWithDashes(tSym+startIx, tLen);
endIx = startIx + fullLen;
if (strand == '-')
    {
    reverseComplement(qSym+startIx, fullLen);
    reverseComplement(tSym+startIx, fullLen);
    reverseBytes(hSym+startIx, fullLen);
    }
for (i=startIx; i<endIx; i += lineLen)
    {
    lineLen = endIx-i;
    if (lineLen > maxLineLen)
        lineLen = maxLineLen;
    mustWrite(stdout, qSym+i, lineLen);
    fputc('\n', stdout);
    for (j=0; j<lineLen; ++j)
        {
        char c = (toupper(qSym[i+j]) == toupper(tSym[i+j]) ? '|' : ' ');
        fputc(c, stdout);
        }
    fputc('\n', stdout);
    mustWrite(stdout, tSym+i, lineLen);
    fputc('\n', stdout);
    //if (showSym)
        {
        mustWrite(stdout, hSym+i, lineLen);
        fputc('\n', stdout);
        }
    fputc('\n', stdout);
    }
if (strand == '-')
    {
    reverseComplement(qSym+startIx, fullLen);
    reverseComplement(tSym+startIx, fullLen);
    reverseBytes(hSym+startIx, fullLen);
    }
}

bool HapgenUtil::makeFlankingHaplotypes(const HapgenAlignment& aln, 
                                        const ReadTable* pRefTable, 
                                        int flanking,
                                        const StringVector& inHaplotypes,
                                        StringVector& outFlankingHaplotypes,
                                        StringVector& outHaplotypes)
{
    std::string upstream;
    std::string referenceHaplotype;
    std::string downstream;

    extractReferenceSubstrings(aln, pRefTable, flanking, upstream, referenceHaplotype, downstream);

    // Flip reference strings to match the strand of the input haplotypes
    if(aln.isRC)
    {
        // reverse complement each string
        upstream = reverseComplement(upstream);
        referenceHaplotype = reverseComplement(referenceHaplotype);
        downstream = reverseComplement(downstream);

        // Swap up and downstream
        upstream.swap(downstream);
    }

    // Make the reference haplotype w/ flanking sequence
    std::string referenceFlanking = upstream + referenceHaplotype + downstream;
    outFlankingHaplotypes.push_back(referenceFlanking);
    outHaplotypes.push_back(referenceHaplotype);

    // Check that all sequences match the reference haplotype properly
    /*
    bool checkOk = checkAlignmentsAreConsistent(referenceFlanking, inHaplotypes);
    if(!checkOk)
    {
        outHaplotypes.clear();
        return false;
    }
    */

    // Make the flanking sequences for each haplotype
    for(size_t i = 0; i < inHaplotypes.size(); ++i)
    {
        // Skip if the input haplotype exactly matches the reference
        if(inHaplotypes[i] != referenceHaplotype)
        {
            outFlankingHaplotypes.push_back(upstream + inHaplotypes[i] + downstream);
            outHaplotypes.push_back(inHaplotypes[i]);
        }
    }

    return true;
}

void NGSReadSet::processReadWhileParsing(NGSRead &tempread) {

    //if (!tempread.flag) return;
    int i, id;

    if (!tempread.direction) {
        reverseComplement(tempread.scaff);
        reverseComplement(tempread.read);
    }
    tempread.convertStateStr(tempread.scaff, SEQ_DNA);
    tempread.convertStateStr(tempread.read, SEQ_DNA);
    assert(tempread.scaff.length() == tempread.read.length());

    int nstates = 4 + (!ngs_ignore_gaps);

    for (i = 0, id = 0; i < tempread.scaff.length(); i++) {
        int state1 = tempread.scaff[i];
        int state2 = tempread.read[i];
        if (state1 >= nstates || state2 >= nstates) continue;
        double *pair_pos, *state_pos;
        while (id >= state_freq.size()) {
            state_pos = new double[nstates];
            memset(state_pos, 0, sizeof(double)*(nstates));
            state_freq.push_back(state_pos);
        }
        state_pos = state_freq[id];
        state_pos[state2] += 1.0/tempread.times;
        while (id >= pair_freq.size()) {
            pair_pos = new double[(nstates) * (nstates)];
            memset(pair_pos, 0, sizeof(double)*(nstates) * (nstates));
            pair_freq.push_back(pair_pos);
        }
        pair_pos = pair_freq[id];
        pair_pos[state1*(nstates) + state2] += 1.0/tempread.times;
        id++;
    }

    if (tree) {
        ReadInfo read_info;
        tempread.homo_rate = homo_rate;
        tempread.computePairFreq();
        read_info.homo_distance = tempread.optimizeDist(1.0-tempread.identity);
        read_info.homo_logl = -tempread.computeFunction(read_info.homo_distance);
        tempread.homo_rate = 0.0;
        read_info.distance = tempread.optimizeDist(read_info.homo_distance);
        read_info.logl = -tempread.computeFunction(read_info.distance);
        read_info.id = tempread.id;
        read_info.identity = tempread.identity;
        push_back(read_info);
    }


}

boolean fastFind(DNA *needle, int needleSize, 
    struct patSpace *ps, struct ffAli **retAli, boolean *retRc, int *retScore)
/* Do fast alignment. */
{
struct patClump *clumpList, *clump;
boolean isRc;
struct aliList *aliList = NULL, *ali;

for (isRc = 0; isRc <= 1; ++isRc)
    {
    if (isRc)
        reverseComplement(needle, needleSize);
    if ((clumpList = patSpaceFindOne(ps, needle, needleSize)) != NULL)
        {
        for (clump = clumpList; clump != NULL; clump = clump->next)
            {
            struct dnaSeq *haySeq = clump->seq;
            DNA *haystack = haySeq->dna;
            int start = clump->start;
            struct ffAli *ffAli = ffFind(needle, needle+needleSize, 
                haystack+start, haystack+start+clump->size, ffCdna);
            if (ffAli != NULL)
                {
                AllocVar(ali);
                ali->ali = ffAli;
                ali->score = ffScoreCdna(ffAli);
                ali->isRc = isRc;
                slAddHead(&aliList, ali);
                }
            }
        slFreeList(&clumpList);
        }
    if (isRc)
        reverseComplement(needle, needleSize);
    }
if (aliList != NULL)
    {
    slSort(&aliList, cmpAliList);
    *retAli = aliList->ali;
    aliList->ali = NULL;
    *retRc = aliList->isRc;
    *retScore = aliList->score;
    for (ali = aliList->next; ali != NULL; ali = ali->next)
        ffFreeAli(&ali->ali);
    slFreeList(&aliList);
    return TRUE;
    }
else
    return FALSE;
}

// Merging ARBRC: R and B into RBR
// First, the sequence of the vertex is extended
// by the the content of the edge label
void Vertex::mergeTipVertex(Edge* pEdge)
{
    Edge* pTwin = pEdge->getTwin();
    //std::cout << "Adding label to " << getID() << " str: " << pSE->getLabel() << "\n";

    // Merge the sequence
    DNAEncodedString label1 = pEdge->getLabel();
    DNAEncodedString label2 = pTwin->getLabel();
    size_t RB_len = label1.length()+label2.length();

    //merge R and B into RBR
    if(pEdge->getDir() == ED_SENSE && pTwin->getComp()==EC_SAME)
    {
        m_seq.append(label1);
        m_seq.append(label2);
    }
    else if(pEdge->getDir() == ED_SENSE && pTwin->getComp()==EC_REVERSE)
    {
        m_seq.append(label1);
        DNAEncodedString tmp(reverseComplement(label2.toString()));
        m_seq.append(tmp);
    }
    else if(pEdge->getDir() == ED_ANTISENSE && pTwin->getComp()==EC_SAME)
    {
        label2.append(label1);
        label2.append(m_seq);
        m_seq=label2;
    }
    else
    {
        DNAEncodedString tmp(reverseComplement(label2.toString()));
        tmp.append(label1);
        tmp.append(m_seq);
        m_seq=tmp;
    }

    // All the SeqCoords for the edges must have their seqlen field updated
    // Also, if we prepended sequence to this edge, all the matches in the
    // SENSE direction must have their coordinates offset
    size_t newLen = m_seq.length();
    for(EdgePtrVecIter iter = m_edges.begin(); iter != m_edges.end(); ++iter)
    {
        Edge* pUpdateEdge = *iter;
        pUpdateEdge->updateSeqLen(newLen);
        //add offset RB to each sense edge
        if(pUpdateEdge->getDir() == ED_SENSE && pEdge != pUpdateEdge)
            pUpdateEdge->offsetMatch(RB_len);
    }
}