boolean anyCdnaSeq(char *name, struct dnaSeq **retDna, struct wormCdnaInfo *retInfo)
/* Get a single cDNA sequence. Optionally (if retInfo is non-null) get additional
* info about the sequence. */
{
static FILE *cdnaFa;
static struct snof *cdnaSnof = NULL;
long offset;
char *faComment;
char **pFaComment = (retInfo == NULL ? NULL : &faComment);
if (cdnaSnof == NULL)
{
char buf[512];
cdnaSnof = snofMustOpen(cdnaName);
sprintf(buf, "%s%s", cdnaName, ".fa");
cdnaFa = mustOpen(buf, "rb");
}
if (!snofFindOffset(cdnaSnof, name, &offset))
return FALSE;
fseek(cdnaFa, offset, SEEK_SET);
if (!faReadNext(cdnaFa, name, TRUE, pFaComment, retDna))
return FALSE;
if (retInfo != NULL)
{
/* Kludge - only look up info if format is more or less right. */
int fieldCount = countChars(faComment, '|');
if (fieldCount >= 8)
wormFaCommentIntoInfo(faComment, retInfo);
else
zeroBytes(retInfo, sizeof(*retInfo));
}
return TRUE;
}
struct dnaSeq *readSeqFromFaPos(struct seqFilePos *sfp, FILE *f)
/* Read part of FA file. */
{
struct dnaSeq *seq;
fseek(f, sfp->pos, SEEK_SET);
if (!faReadNext(f, "", TRUE, NULL, &seq))
errAbort("Couldn't faReadNext on %s in %s\n", sfp->name, sfp->file);
return seq;
}
struct dnaSeq *faReadOneDnaSeq(FILE *f, char *defaultName, boolean mustStartWithComment)
/* Read sequence from FA file. Assumes positioned at or before
* the '>' at start of sequence. */
{
struct dnaSeq *seq;
if (!faReadNext(f, defaultName, mustStartWithComment, NULL, &seq))
return NULL;
else
return seq;
}
boolean nextWormCdnaAndInfo(struct wormCdnaIterator *it, struct dnaSeq **retSeq,
struct wormCdnaInfo *retInfo)
/* Return next sequence and associated info from database. */
{
char *faComment;
if (!faReadNext(it->faFile, "unknown", TRUE, &faComment, retSeq))
return FALSE;
wormFaCommentIntoInfo(faComment, retInfo);
return TRUE;
}
boolean wormCdnaSeq(char *name, struct dnaSeq **retDna, struct wormCdnaInfo *retInfo)
/* Get a single worm cDNA sequence. Optionally (if retInfo is non-null) get additional
* info about the sequence. */
{
long offset;
char *faComment;
char **pFaComment = (retInfo == NULL ? NULL : &faComment);
wormCdnaCache();
if (!snofFindOffset(cdnaSnof, name, &offset))
return FALSE;
fseek(cdnaFa, offset, SEEK_SET);
if (!faReadNext(cdnaFa, name, TRUE, pFaComment, retDna))
return FALSE;
wormFaCommentIntoInfo(faComment, retInfo);
return TRUE;
}
void chopFaLines(char *inName, char *outName)
/* chopFaLines - Read in FA file with long lines and rewrite it with shorter lines. */
{
FILE *in = mustOpen(inName, "r");
FILE *out = mustOpen(outName, "w");
char *commentLine;
struct dnaSeq *seq;
while (faReadNext(in, NULL, TRUE, &commentLine, &seq))
{
commentLine = trimSpaces(commentLine+1);
uglyf(">%s\n", commentLine);
mustWrite(uglyOut, seq->dna, 100);
uglyf("\n");
uglyAbort("All for now");
faWriteNext(out, commentLine, seq->dna, seq->size);
}
}
static struct traceInfo* parseFastaRecord(FILE* fh, char* fastaName)
/* read the next fasta record akd create a traceInfo object. This
* parses the sequence id and comment for the read and clone name. */
{
struct dnaSeq* dna;
char* comment;
struct traceInfo* traceInfo;
if (!faReadNext(fh, NULL, 0, &comment, &dna))
return NULL; /* EOF */
AllocVar(traceInfo);
traceInfo->ti = parseTraceId(dna->name, fastaName);
traceInfo->size = dna->size;
traceInfo->templateId = parseTemplateId(comment, fastaName);
freeMem(comment);
freeDnaSeq(&dna);
return traceInfo;
}
boolean flyCdnaSeq(char *name, struct dnaSeq **retDna, struct wormCdnaInfo *retInfo)
/* Get a single fly cDNA sequence. Optionally (if retInfo is non-null) get additional
* info about the sequence. */
{
long offset;
char *faComment;
char **pFaComment = (retInfo == NULL ? NULL : &faComment);
static struct snof *cdnaSnof = NULL;
static FILE *cdnaFa;
if (cdnaSnof == NULL)
cdnaSnof = snofMustOpen("c:/biodata/fly/cDna/allcdna");
if (cdnaFa == NULL)
cdnaFa = mustOpen("c:/biodata/fly/cDna/allcdna.fa", "rb");
if (!snofFindOffset(cdnaSnof, name, &offset))
return FALSE;
fseek(cdnaFa, offset, SEEK_SET);
if (!faReadNext(cdnaFa, name, TRUE, pFaComment, retDna))
return FALSE;
flyFaCommentIntoInfo(faComment, retInfo);
return TRUE;
}
int main(int argc, char *argv[])
{
char *estName, *targetName, *oocName;
FILE *estFile;
struct dnaSeq *target;
struct dnaSeq *est;
struct patSpace *ps;
struct patClump *clumpList, *clump;
int estIx = 0;
/* Check command line arguments and assign to local variables. */
if (argc != 4)
usage();
estName = argv[1];
estFile = mustOpen(estName, "rb");
targetName = argv[2];
oocName = argv[3];
/* Read in target DNA from fasta files and check not too big. */
fprintf(stderr, "Reading %s\n", targetName);
target = faReadAllDna(targetName);
if (totalSequenceSize(target) > 8000000)
{
errAbort("Can only handle 8000000 bases of genomic sequence at once, %s has %d.",
targetName, totalSequenceSize(target));
}
/* Make a pattern space index structure. */
fprintf(stderr, "Making Pattern Space index\n");
ps = makePatSpace(&target, 1, oocName, 4, 32000);
/* Loop through each EST in query list. */
printf("Searching for hits\n\n");
while (faReadNext(estFile, NULL, TRUE, NULL, &est))
{
boolean isRc; /* Reverse complemented? */
if (++estIx % 5000 == 0)
fprintf(stderr, "Processing EST %d\n", estIx);
if (est->size > 20000)
{
warn("Very large EST sequence %s.\n"
"Maybe you mixed up the EST and genomic parameters?", est->name);
usage();
}
for (isRc = 0; isRc <= 1; ++isRc) /* Search both strands. */
{
if (isRc)
reverseComplement(est->dna, est->size);
clumpList = patSpaceFindOne(ps, est->dna, est->size);
/* For each homology clump patSpace finds, do a fuzzyFinder
* alignment of it and print the results. */
for (clump = clumpList; clump != NULL; clump = clump->next)
{
struct ffAli *ali, *a;
boolean isRc;
int score;
struct dnaSeq *t = clump->seq;
DNA *tStart = t->dna + clump->start;
ali = ffFind(est->dna, est->dna+est->size, tStart, tStart + clump->size, ffCdna);
if (ali != NULL)
{
score = ffScoreCdna(ali);
printf("%s hits %s strand %c score %d\n",
est->name, t->name, (isRc ? '+' : '-'), score);
for (a = ali; a != NULL; a = a->right)
{
printf(" Q %4d - %4d\t T %4d -%4d\n",
a->nStart - est->dna, a->nEnd - est->dna,
a->hStart - t->dna, a->hEnd - t->dna);
}
printf("\n");
ffFreeAli(&ali);
}
else
{
printf("Couldn't align clump at %s %d-%d\n",
t->name, clump->start, clump->start + clump->size);
}
}
slFreeList(&clumpList);
}
freeDnaSeq(&est);
}
/* Clean up time. */
freePatSpace(&ps);
freeSeqList(&target);
return 0;
}