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 ggcChrom(struct chromGenes *chrom, char *axtFile,
struct ggcInfo *g, struct hash *restrictHash,
FILE *fParts)
/* Tabulate matches on chromosome. */
{
struct lineFile *lf = lineFileOpen(axtFile, TRUE);
bool *hits, *covers;
int hitCount = 0, coverCount = 0;
struct axt *axt;
struct genePred *gp;
int closeSize = g->closeSize;
int closeHalf = closeSize/2;
/* Build up array of booleans - one per base - which are
* 1's where mouse/human align and bases match, zero
* elsewhere. */
AllocArray(hits, chrom->size);
AllocArray(covers, chrom->size);
printf("%s (%d bases)\n", chrom->name, chrom->size);
while ((axt = axtRead(lf)) != NULL)
{
int tPos = axt->tStart;
int symCount = axt->symCount, i;
char t, q, *tSym = axt->tSym, *qSym = axt->qSym;
if (axt->tEnd > chrom->size)
errAbort("tEnd %d, chrom size %d in %s",
axt->tEnd, chrom->size, axtFile);
if (axt->tStrand == '-')
errAbort("Can't handle minus strand on target in %s", axtFile);
for (i=0; i<symCount; ++i)
{
t = tSym[i];
if (t != '-')
{
q = qSym[i];
if (toupper(t) == toupper(q))
{
hits[tPos] = TRUE;
++hitCount;
}
if (q == '-')
covers[tPos] = 1;
else
covers[tPos] = 2;
++tPos;
}
}
axtFree(&axt);
}
for (gp = chrom->geneList; gp != NULL; gp = gp->next)
{
int exonIx;
int utr3Size = 0, utr5Size = 0, cdsAllSize = 0;
int utr3Pos = 0, utr5Pos = 0, cdsAllPos = 0;
bool *utr3Hits = NULL, *utr3Covers = NULL;
bool *utr5Hits = NULL, *utr5Covers = NULL;
bool *cdsAllHits = NULL, *cdsAllCovers = NULL;
bool isRev = (gp->strand[0] == '-');
/* Filter out genes not in restrict hash if any. */
++totalGenes;
if (restrictHash != NULL)
if (!hashLookup(restrictHash, gp->name))
continue;
++reviewedGenes;
/* Filter out genes without meaningful UTRs */
if (gp->cdsStart - gp->txStart < g->closeSize/2 ||
gp->txEnd - gp->cdsEnd < g->closeSize/2)
continue;
++genesUsed;
/* Total up UTR and CDS sizes. */
for (exonIx=0; exonIx<gp->exonCount; ++exonIx)
{
int eStart = gp->exonStarts[exonIx];
int eEnd = gp->exonEnds[exonIx];
int eSize = eEnd - eStart;
int oneUtr, oneCds;
oneCds = rangeIntersection(gp->cdsStart, gp->cdsEnd, eStart, eEnd);
if (oneCds > 0)
{
cdsAllSize += oneCds;
}
if (eStart < gp->cdsStart)
{
int utrStart = eStart;
int utrEnd = min(gp->cdsStart, eEnd);
int utrSize = utrEnd - utrStart;
if (isRev)
utr3Size += utrSize;
else
utr5Size += utrSize;
}
if (eEnd > gp->cdsEnd)
{
int utrStart = max(gp->cdsEnd, eStart);
int utrEnd = eEnd;
int utrSize = utrEnd - utrStart;
if (isRev)
utr5Size += utrSize;
else
utr3Size += utrSize;
}
}
/* Condense hits from UTRs and CDSs */
if (utr5Size > 0)
{
AllocArray(utr5Hits, utr5Size);
AllocArray(utr5Covers, utr5Size);
}
if (utr3Size > 0)
{
AllocArray(utr3Hits, utr3Size);
AllocArray(utr3Covers, utr3Size);
}
if (cdsAllSize > 0)
{
AllocArray(cdsAllHits, cdsAllSize);
AllocArray(cdsAllCovers, cdsAllSize);
}
for (exonIx=0; exonIx<gp->exonCount; ++exonIx)
{
int eStart = gp->exonStarts[exonIx];
int eEnd = gp->exonEnds[exonIx];
int eSize = eEnd - eStart;
int oneUtr, oneCds;
oneCds = rangeIntersection(gp->cdsStart, gp->cdsEnd, eStart, eEnd);
if (oneCds > 0)
{
int cdsStart = eStart;
int cdsEnd = gp->cdsEnd;
if (cdsStart < gp->cdsStart)
cdsStart = gp->cdsStart;
memcpy(cdsAllHits + cdsAllPos, hits + cdsStart, oneCds * sizeof(*hits));
memcpy(cdsAllCovers + cdsAllPos, covers + cdsStart, oneCds * sizeof(*covers));
cdsAllPos += oneCds;
}
if (eStart < gp->cdsStart)
{
int utrStart = eStart;
int utrEnd = min(gp->cdsStart, eEnd);
int utrSize = utrEnd - utrStart;
if (isRev)
{
memcpy(utr3Hits + utr3Pos, hits + utrStart, utrSize * sizeof(*hits));
memcpy(utr3Covers + utr3Pos, covers + utrStart, utrSize * sizeof(*covers));
utr3Pos += utrSize;
}
else
{
memcpy(utr5Hits + utr5Pos, hits + utrStart, utrSize * sizeof(*hits));
memcpy(utr5Covers + utr5Pos, covers + utrStart, utrSize * sizeof(*covers));
utr5Pos += utrSize;
}
}
if (eEnd > gp->cdsEnd)
{
int utrStart = max(gp->cdsEnd, eStart);
int utrEnd = eEnd;
int utrSize = utrEnd - utrStart;
if (isRev)
{
memcpy(utr5Hits + utr5Pos, hits + utrStart, utrSize * sizeof(*hits));
memcpy(utr5Covers + utr5Pos, covers + utrStart, utrSize * sizeof(*covers));
utr5Pos += utrSize;
}
else
{
memcpy(utr3Hits + utr3Pos, hits + utrStart, utrSize * sizeof(*hits));
memcpy(utr3Covers + utr3Pos, covers + utrStart, utrSize * sizeof(*covers));
utr3Pos += utrSize;
}
}
}
assert(utr3Pos == utr3Size);
assert(utr5Pos == utr5Size);
assert(cdsAllPos == cdsAllSize);
tallyHits(&g->utr5, utr5Hits, utr5Covers, utr5Size, isRev);
tallyHits(&g->utr3, utr3Hits, utr3Covers, utr3Size, isRev);
tallyHits(&g->cdsAll, cdsAllHits, cdsAllCovers, cdsAllSize, isRev);
/* Optionally write out file with gene by gene info. */
if (fParts != NULL)
{
/* Write header line first time through. */
static boolean firstTime = TRUE;
if (firstTime)
{
firstTime = FALSE;
fprintf(fParts, "#accession\tsize_5\tali_5\tmatch_5\tsize_c\tali_c\tmatch_c\tsize_3\tali_3\tmatch_3\n");
}
fprintf(fParts, "%s\t", gp->name);
fprintf(fParts, "%d\t%d\t%d\t", utr5Size,
countBools(utr5Covers, utr5Size),
countBools(utr5Hits, utr5Size));
fprintf(fParts, "%d\t%d\t%d\t", cdsAllSize,
countBools(cdsAllCovers, cdsAllSize),
countBools(cdsAllHits, cdsAllSize));
fprintf(fParts, "%d\t%d\t%d\n", utr3Size,
countBools(utr3Covers, utr3Size),
countBools(utr3Hits, utr3Size));
}
/* Tally upstream/downstream hits. */
{
int s1 = gp->txStart - closeHalf;
int e1 = s1 + closeSize;
int s2 = gp->txEnd - closeHalf;
int e2 = s2 + closeSize;
if (isRev)
{
tallyInRange(&g->down, hits, covers, chrom->size, gp->txStart - g->baseDown,
gp->txStart, isRev);
tallyInRange(&g->up, hits, covers, chrom->size, gp->txEnd,
gp->txEnd + g->baseUp, isRev);
tallyInRange(&g->txEnd, hits, covers, chrom->size, s1, e1, isRev);
tallyInRange(&g->txStart, hits, covers, chrom->size, s2, e2, isRev);
}
else
{
tallyInRange(&g->up, hits, covers, chrom->size, gp->txStart - g->baseUp,
gp->txStart, isRev);
tallyInRange(&g->down, hits, covers, chrom->size, gp->txEnd,
gp->txEnd + g->baseDown, isRev);
tallyInRange(&g->txStart, hits, covers, chrom->size, s1, e1, isRev);
tallyInRange(&g->txEnd, hits, covers, chrom->size, s2, e2, isRev);
}
}
/* Tally hits in coding exons */
for (exonIx=0; exonIx < gp->exonCount; ++exonIx)
{
int eStart = gp->exonStarts[exonIx];
int eEnd = gp->exonEnds[exonIx];
/* Single coding exon. */
if (eStart <= gp->cdsStart && eEnd >= gp->cdsEnd)
{
eStart = gp->cdsStart;
eEnd = gp->cdsEnd;
tallyInRange(&g->cdsSingle, hits, covers, chrom->size,
eStart, eEnd, isRev);
}
/* Initial coding exon */
else if (eStart < gp->cdsStart && eEnd > gp->cdsStart)
{
int cs = gp->cdsStart - closeHalf;
int ce = cs + closeSize;
eStart = gp->cdsStart;
if (isRev)
{
tallyInRange(&g->tlEnd, hits, covers, chrom->size, cs, ce, isRev);
tallyInRange(&g->cdsLast, hits, covers, chrom->size,
eStart, eEnd, isRev);
}
else
{
tallyInRange(&g->tlStart, hits, covers, chrom->size, cs, ce, isRev);
tallyInRange(&g->cdsFirst, hits, covers, chrom->size,
eStart, eEnd, isRev);
}
}
/* Final coding exon */
else if (eStart < gp->cdsEnd && eEnd > gp->cdsEnd)
{
int cs = gp->cdsEnd - closeHalf;
int ce = cs + closeSize;
eEnd = gp->cdsEnd;
if (isRev)
{
tallyInRange(&g->tlStart, hits, covers, chrom->size, cs, ce, isRev);
tallyInRange(&g->cdsFirst, hits, covers, chrom->size,
eStart, eEnd, isRev);
}
else
{
tallyInRange(&g->tlEnd, hits, covers, chrom->size, cs, ce, isRev);
tallyInRange(&g->cdsLast, hits, covers, chrom->size,
eStart, eEnd, isRev);
}
}
/* Middle (but not only) coding exon */
else if (eStart >= gp->cdsStart && eEnd <= gp->cdsEnd)
{
tallyInRange(&g->cdsMiddle, hits, covers, chrom->size, eStart, eEnd, isRev);
}
else
{
}
}
/* Tally hits in introns and splice sites. */
for (exonIx=1; exonIx<gp->exonCount; ++exonIx)
{
int iStart = gp->exonEnds[exonIx-1];
int iEnd = gp->exonStarts[exonIx];
int s1 = iStart - closeHalf;
int e1 = s1 + closeSize;
int s2 = iEnd - closeHalf;
int e2 = s2 + closeSize;
if (isRev)
{
tallyInRange(&g->splice3, hits, covers, chrom->size,
s1, e1, isRev);
tallyInRange(&g->splice5, hits, covers, chrom->size,
s2, e2, isRev);
}
else
{
tallyInRange(&g->splice5, hits, covers, chrom->size,
s1, e1, isRev);
tallyInRange(&g->splice3, hits, covers, chrom->size,
s2, e2, isRev);
}
tallyInRange(&g->intron, hits, covers, chrom->size, iStart, iEnd, isRev);
}
freez(&utr5Hits);
freez(&utr3Hits);
freez(&cdsAllHits);
freez(&utr5Covers);
freez(&utr3Covers);
freez(&cdsAllCovers);
}
freez(&hits);
freez(&covers);
lineFileClose(&lf);
}
