struct qaSeq *qacReadNext(FILE *f, boolean isSwapped)
/* Read in next record in .qac file. */
{
bits32 cSize, origSize;
struct qaSeq *qa;
signed char *buf;
char *s;
s = readString(f);
if (s == NULL)
return NULL;
AllocVar(qa);
qa->name = s;
mustReadOne(f, origSize);
if (isSwapped)
origSize = byteSwap32(origSize);
mustReadOne(f, cSize);
if (isSwapped)
cSize = byteSwap32(cSize);
qa->size = origSize;
qa->qa = needLargeMem(origSize);
buf = needLargeMem(cSize);
mustRead(f, buf, cSize);
rleUncompress(buf, cSize, qa->qa, origSize);
freeMem(buf);
return qa;
}
void chimpSuperQuals(char *agpFile, char *qacInName, char *qacOutName)
/* chimpSuperQuals - Map chimp quality scores from contig to supercontig.. */
{
struct hash *qacHash = qacReadToHash(qacInName);
struct scaffold *scaffold, *scaffoldList = readScaffoldsFromAgp(agpFile);
FILE *f = mustOpen(qacOutName, "w");
struct qaSeq qa;
struct agpFrag *frag;
struct qac *qac;
int bufSize = 0;
UBYTE *buf = NULL;
int qaMaxSize = 0;
int fragSize;
int count = 0;
qacWriteHead(f);
ZeroVar(&qa);
for (scaffold = scaffoldList; scaffold != NULL; scaffold = scaffold->next)
{
/* Set up qa to hold uncompressed quals for whole scaffold. */
qa.name = scaffold->list->chrom;
qa.size = scaffold->size;
if (qaMaxSize < qa.size)
{
freez(&qa.qa);
qa.qa = needHugeZeroedMem(qa.size);
qaMaxSize = qa.size;
}
/* Uncompress contig quality scores and copy into scaffold's quality buffer. */
for (frag = scaffold->list; frag != NULL; frag = frag->next)
{
qac = hashMustFindVal(qacHash, frag->frag);
if (bufSize < qac->uncSize)
{
freez(&buf);
bufSize = qac->uncSize;
buf = needMem(bufSize);
}
rleUncompress(qac->data, qac->compSize, buf, qac->uncSize);
fragSize = frag->fragEnd - frag->fragStart;
memcpy(qa.qa + frag->chromStart, buf + frag->fragStart, fragSize);
}
/* Compress and write it out. */
qacWriteNext(f, &qa);
++count;
}
carefulClose(&f);
}
int
RleCompressor::uncompress (const char *inPtr,
int inSize,
int minY,
const char *&outPtr)
{
//
// Special case - empty input buffer
//
if (inSize == 0)
{
outPtr = _outBuffer;
return 0;
}
//
// Decode the run-length encoded data
//
int outSize;
if (0 == (outSize = rleUncompress (inSize, _maxScanLineSize,
(const signed char *) inPtr,
_tmpBuffer)))
{
throw Iex::InputExc ("Data decoding (rle) failed.");
}
//
// Predictor.
//
{
unsigned char *t = (unsigned char *) _tmpBuffer + 1;
unsigned char *stop = (unsigned char *) _tmpBuffer + outSize;
while (t < stop)
{
int d = int (t[-1]) + int (t[0]) - 128;
t[0] = d;
++t;
}
}
//
// Reorder the pixel data.
//
{
const char *t1 = _tmpBuffer;
const char *t2 = _tmpBuffer + (outSize + 1) / 2;
char *s = _outBuffer;
char *stop = s + outSize;
while (true)
{
if (s < stop)
*(s++) = *(t1++);
else
break;
if (s < stop)
*(s++) = *(t2++);
else
break;
}
}
outPtr = _outBuffer;
return outSize;
}
void axtHiQualDiffs(char *axtFile, struct hash *qacHash, FILE *f)
/* Write out high quality diffs in axtFile to f. */
{
char *qName = cloneString("");
UBYTE *qQuals = NULL;
UBYTE *quals = NULL;
struct qac *qac = NULL;
struct axt *axt = NULL;
struct lineFile *lf = lineFileOpen(axtFile, TRUE);
int qStart, qDir, qPos, qWinStart, qWinEnd, tPos;
int qWinSize = optionInt("winSize", 11);
int qQualMin = optionInt("diffQualMin", 30);
int qWinQualMin = optionInt("winQualMin", 25);
int qWinMaxDiff = optionInt("winMaxDiff", 2);
boolean qIndelOk = optionExists("indelOk");
boolean qIgnore98 = optionExists("ignore98");
boolean chimpPos = optionExists("chimpPos");
int qHalfWinSize = qWinSize/2;
while ((axt = axtRead(lf)) != NULL)
{
char *qSym = axt->qSym, *tSym = axt->tSym;
int symIx, symCount = axt->symCount;
char qc,tc;
toUpperN(qSym, symCount);
toUpperN(tSym, symCount);
if (!sameString(axt->qName, qName))
{
freez(&qName);
qName = cloneString(axt->qName);
qac = hashMustFindVal(qacHash, qName);
freez(&qQuals);
qQuals = needHugeMem(qac->uncSize);
rleUncompress(qac->data, qac->compSize, qQuals, qac->uncSize);
}
if (axt->qStrand == '+')
{
qStart = axt->qStart;
qDir = 1;
}
else
{
qStart = qac->uncSize - axt->qStart - 1;
qDir = -1;
}
qPos = qStart;
tPos = axt->tStart;
for (symIx = 0; symIx < symCount; ++symIx)
{
qc = qSym[symIx];
tc = tSym[symIx];
if (qc == '-')
tPos += 1;
else if (tc == '-')
qPos += qDir;
else
{
if (qc != tc)
{
qWinStart = qPos - qHalfWinSize;
qWinEnd = qWinStart + qWinSize;
if (qWinStart >= 0 && qWinEnd < qac->uncSize)
{
if (qQuals[qPos] >= qQualMin)
{
int i;
boolean ok = TRUE;
for (i = qWinStart; i<qWinEnd; ++i)
if (qQuals[i] < qWinQualMin)
{
ok = FALSE;
break;
}
if (ok)
{
int diffCount = 0;
int symWinStart = symIx - qHalfWinSize;
int symWinEnd = symWinStart + qWinSize;
for (i=symWinStart; i < symWinEnd; ++i)
{
qc = qSym[i];
tc = tSym[i];
if (qc == '-' || tc == '-')
{
ok = FALSE;
break;
}
if (qc != tc)
++diffCount;
}
if (ok && diffCount <= qWinMaxDiff && (!qIgnore98 || qQuals[qPos] != 98) )
{
if (chimpPos)
fprintf(f, "%s\t%d\t%d\t%c\t%c\t%s\t%d\t%d\n",
axt->tName, tPos, tPos+1, tSym[symIx], qSym[symIx], axt->qName, qPos, qPos+1);
else
fprintf(f, "%s\t%d\t%d\t%c\t%c\n",
axt->tName, tPos, tPos+1, tSym[symIx], qSym[symIx]);
}
}
}
}
}
qPos += qDir;
tPos += 1;
}
}
axtFree(&axt);
}
lineFileClose(&lf);
}
void qacAgpLift(char *agpFile, char *qacInName, char *qacOutName)
/* qacAgpLift - Use AGP to combine per-scaffold qac into per-chrom qac. */
{
struct hash *qacHash = qacReadToHash(qacInName);
struct chrom *chrom, *chromList = readChromScaffoldsFromAgp(agpFile);
FILE *f = mustOpen(qacOutName, "w");
struct qaSeq qa;
struct agpFrag *frag;
struct qac *qac;
int bufSize = 0;
UBYTE *buf = NULL;
int qaMaxSize = 0;
int fragSize;
int count = 0;
qacWriteHead(f);
ZeroVar(&qa);
for (chrom = chromList; chrom != NULL; chrom = chrom->next)
{
/* Set up qa to hold uncompressed quals for whole chrom. */
qa.name = chrom->list->chrom;
verbose(1, " %s size=%d\n", chrom->list->chrom, chrom->size);
qa.size = chrom->size;
if (qaMaxSize < qa.size)
{
qa.qa = needHugeZeroedMem(qa.size);
qaMaxSize = qa.size;
}
else
{
zeroBytes(qa.qa, qa.size);
}
/* Uncompress contig quality scores and copy into chrom's quality buffer. */
for (frag = chrom->list; frag != NULL; frag = frag->next)
{
struct hashEl *hel;
fragSize = frag->fragEnd - frag->fragStart;
if ((hel = hashLookup(qacHash, frag->frag)) != NULL)
{
qac = (struct qac *) hel->val;
if (bufSize < qac->uncSize)
{
freez(&buf);
bufSize = qac->uncSize;
buf = needMem(bufSize);
}
rleUncompress(qac->data, qac->compSize, buf, qac->uncSize);
if (frag->strand[0] == '-')
reverseBytes((char*)buf, qac->uncSize);
memcpy(qa.qa + frag->chromStart, buf + frag->fragStart, fragSize);
}
else
{
/* agp frag not found in qac hash -- missing data */
if (mScore < 0)
errAbort("missing data: no quality scores for %s", frag->frag);
/* fill in missing data with specified score */
memset(qa.qa + frag->chromStart, mScore, fragSize);
}
}
/* Compress and write it out. */
qacWriteNext(f, &qa);
++count;
}
carefulClose(&f);
}
