void faLowerToN(char *inName, char *outName)
/* faLowerToN - Convert lower case bases to N.. */
{
struct lineFile *lf = lineFileOpen(inName, TRUE);
FILE *f = mustOpen(outName, "w");
char *line;
while (lineFileNext(lf, &line, NULL))
{
if (line[0] != '>')
lowerToN(line);
fprintf(f, "%s\n", line);
}
carefulClose(&f);
}
void seqFromPsl(char *inPsl, char *inTwoBit, char *outFa)
/* seqFromPsl - Extract masked sequence from database corresponding to psl file. */
{
struct twoBitFile *tbf = twoBitOpen(inTwoBit);
struct lineFile *lf = pslFileOpen(inPsl);
FILE *f = mustOpen(outFa, "w");
struct psl *psl;
while ((psl = pslNext(lf)) != NULL)
{
char faHead[512];
struct dnaSeq *seq = twoBitReadSeqFrag(tbf, psl->tName,
psl->tStart, psl->tEnd);
if (psl->strand[0] == '-')
reverseComplement(seq->dna, seq->size);
safef(faHead, sizeof(faHead), "%s (%s:%d-%d)",
psl->qName, psl->tName, psl->tStart+1, psl->tEnd);
if (hardMask)
lowerToN(seq->dna, seq->size);
faWriteNext(f, faHead, seq->dna, seq->size);
}
carefulClose(&f);
}
static void hgSeqConcatRegionsDb(char *db, char *chrom, int chromSize, char strand, char *name,
int rCount, unsigned *rStarts, unsigned *rSizes,
boolean *exonFlags, boolean *cdsFlags)
/* Concatenate and print out dna for a series of regions. */
{
// Note: this code use to generate different sequence ids if the global
// database in hdb was different than the db parameter. This functionality
// has been removed since the global database was removed and it didn't
// appear to be used.
struct dnaSeq *rSeq = NULL;
struct dnaSeq *cSeq = NULL;
char recName[256];
int seqStart, seqEnd;
int offset, cSize;
int i;
boolean isRc = (strand == '-') || cgiBoolean("hgSeq.revComp");
boolean maskRep = cgiBoolean("hgSeq.maskRepeats");
int padding5 = cgiOptionalInt("hgSeq.padding5", 0);
int padding3 = cgiOptionalInt("hgSeq.padding3", 0);
char *casing = cgiString("hgSeq.casing");
char *repMasking = cgiString("hgSeq.repMasking");
char *granularity = cgiOptionalString("hgSeq.granularity");
boolean concatRegions = granularity && sameString("gene", granularity);
if (rCount < 1)
return;
/* Don't support padding if granularity is gene (i.e. concat'ing all). */
if (concatRegions)
{
padding5 = padding3 = 0;
}
i = rCount - 1;
seqStart = rStarts[0] - (isRc ? padding3 : padding5);
seqEnd = rStarts[i] + rSizes[i] + (isRc ? padding5 : padding3);
/* Padding might push us off the edge of the chrom; if so, truncate: */
if (seqStart < 0)
{
if (isRc)
padding3 += seqStart;
else
padding5 += seqStart;
seqStart = 0;
}
/* if we know the chromSize, don't pad out beyond it */
if ((chromSize > 0) && (seqEnd > chromSize))
{
if (isRc)
padding5 += (chromSize - seqEnd);
else
padding3 += (chromSize - seqEnd);
seqEnd = chromSize;
}
if (seqEnd <= seqStart)
{
printf("# Null range for %s_%s (range=%s:%d-%d 5'pad=%d 3'pad=%d) (may indicate a query-side insert)\n",
db,
name,
chrom, seqStart+1, seqEnd,
padding5, padding3);
return;
}
if (maskRep)
{
rSeq = hDnaFromSeq(db, chrom, seqStart, seqEnd, dnaMixed);
if (sameString(repMasking, "N"))
lowerToN(rSeq->dna, strlen(rSeq->dna));
if (!sameString(casing, "upper"))
tolowers(rSeq->dna);
}
else if (sameString(casing, "upper"))
rSeq = hDnaFromSeq(db, chrom, seqStart, seqEnd, dnaUpper);
else
rSeq = hDnaFromSeq(db, chrom, seqStart, seqEnd, dnaLower);
/* Handle casing and compute size of concatenated sequence */
cSize = 0;
for (i=0; i < rCount; i++)
{
if ((sameString(casing, "exon") && exonFlags[i]) ||
(sameString(casing, "cds") && cdsFlags[i]))
{
int rStart = rStarts[i] - seqStart;
toUpperN(rSeq->dna+rStart, rSizes[i]);
}
cSize += rSizes[i];
}
cSize += (padding5 + padding3);
AllocVar(cSeq);
cSeq->dna = needLargeMem(cSize+1);
cSeq->size = cSize;
offset = 0;
for (i=0; i < rCount; i++)
{
int start = rStarts[i] - seqStart;
int size = rSizes[i];
if (i == 0)
{
start -= (isRc ? padding3 : padding5);
assert(start == 0);
size += (isRc ? padding3 : padding5);
}
if (i == rCount-1)
{
size += (isRc ? padding5 : padding3);
}
memcpy(cSeq->dna+offset, rSeq->dna+start, size);
offset += size;
}
assert(offset == cSeq->size);
cSeq->dna[offset] = 0;
freeDnaSeq(&rSeq);
if (isRc)
reverseComplement(cSeq->dna, cSeq->size);
safef(recName, sizeof(recName),
"%s_%s range=%s:%d-%d 5'pad=%d 3'pad=%d "
"strand=%c repeatMasking=%s",
db,
name,
chrom, seqStart+1, seqEnd,
padding5, padding3,
(isRc ? '-' : '+'),
(maskRep ? repMasking : "none"));
faWriteNext(stdout, recName, cSeq->dna, cSeq->size);
freeDnaSeq(&cSeq);
}
