int main(int argc, char *argv[])
/* Process command line into global variables and call blat. */
{
boolean tIsProtLike, qIsProtLike;
#ifdef DEBUG
{
char *cmd = "blat hCrea.geno hCrea.mrna foo.psl -t=dnax -q=rnax";
char *words[16];
printf("Debugging parameters\n");
cmd = cloneString(cmd);
argc = chopLine(cmd, words);
argv = words;
}
#endif /* DEBUG */
optionInit(&argc, argv, options);
if (argc != 4)
usage();
/* Get database and query sequence types and make sure they are
* legal and compatable. */
if (optionExists("prot"))
qType = tType = gftProt;
if (optionExists("t"))
tType = gfTypeFromName(optionVal("t", NULL));
trimA = optionExists("trimA") || optionExists("trima");
trimT = optionExists("trimT") || optionExists("trimt");
trimHardA = optionExists("trimHardA");
switch (tType)
{
case gftProt:
case gftDnaX:
tIsProtLike = TRUE;
break;
case gftDna:
tIsProtLike = FALSE;
break;
default:
tIsProtLike = FALSE;
errAbort("Illegal value for 't' parameter");
break;
}
if (optionExists("q"))
qType = gfTypeFromName(optionVal("q", NULL));
if (qType == gftRnaX || qType == gftRna)
trimA = TRUE;
if (optionExists("noTrimA"))
trimA = FALSE;
switch (qType)
{
case gftProt:
case gftDnaX:
case gftRnaX:
minIdentity = 25;
qIsProtLike = TRUE;
break;
default:
qIsProtLike = FALSE;
break;
}
if ((tIsProtLike ^ qIsProtLike) != 0)
errAbort("t and q must both be either protein or dna");
/* Set default tile size for protein-based comparisons. */
if (tIsProtLike)
{
tileSize = 5;
minMatch = 1;
oneOff = FALSE;
maxGap = 0;
}
/* Get tile size and related parameters from user and make sure
* they are within range. */
tileSize = optionInt("tileSize", tileSize);
stepSize = optionInt("stepSize", tileSize);
minMatch = optionInt("minMatch", minMatch);
oneOff = optionExists("oneOff");
fastMap = optionExists("fastMap");
minScore = optionInt("minScore", minScore);
maxGap = optionInt("maxGap", maxGap);
minRepDivergence = optionFloat("minRepDivergence", minRepDivergence);
minIdentity = optionFloat("minIdentity", minIdentity);
gfCheckTileSize(tileSize, tIsProtLike);
if (minMatch < 0)
errAbort("minMatch must be at least 1");
if (maxGap > 100)
errAbort("maxGap must be less than 100");
/* Set repMatch parameter from command line, or
* to reasonable value that depends on tile size. */
if (optionExists("repMatch"))
repMatch = optionInt("repMatch", repMatch);
else
repMatch = gfDefaultRepMatch(tileSize, stepSize, tIsProtLike);
/* Gather last few command line options. */
noHead = optionExists("noHead");
ooc = optionVal("ooc", NULL);
makeOoc = optionVal("makeOoc", NULL);
mask = optionVal("mask", NULL);
qMask = optionVal("qMask", NULL);
repeats = optionVal("repeats", NULL);
if (repeats != NULL && mask != NULL && differentString(repeats, mask))
errAbort("The -mask and -repeat settings disagree. "
"You can just omit -repeat if -mask is on");
if (mask != NULL) /* Mask setting will also set repeats. */
repeats = mask;
outputFormat = optionVal("out", outputFormat);
dotEvery = optionInt("dots", 0);
/* set global for fuzzy find functions */
setFfIntronMax(optionInt("maxIntron", ffIntronMaxDefault));
setFfExtendThroughN(optionExists("extendThroughN"));
/* Call routine that does the work. */
blat(argv[1], argv[2], argv[3]);
return 0;
}
void gfClient(char *hostName, char *portName, char *tSeqDir, char *inName,
char *outName, char *tTypeName, char *qTypeName)
/* gfClient - A client for the genomic finding program that produces a .psl file. */
{
struct lineFile *lf = lineFileOpen(inName, TRUE);
static bioSeq seq;
FILE *out = mustOpen(outName, "w");
enum gfType qType = gfTypeFromName(qTypeName);
enum gfType tType = gfTypeFromName(tTypeName);
int dotMod = 0;
char databaseName[256];
struct hash *tFileCache = gfFileCacheNew();
snprintf(databaseName, sizeof(databaseName), "%s:%s", hostName, portName);
gvo = gfOutputAny(outputFormat, round(minIdentity*10), qType == gftProt, tType == gftProt,
optionExists("nohead"), databaseName, 23, 3.0e9, minIdentity, out);
gfOutputHead(gvo, out);
while (faSomeSpeedReadNext(lf, &seq.dna, &seq.size, &seq.name, qType != gftProt))
{
int conn = gfConnect(hostName, portName);
if (dots != 0)
{
if (++dotMod >= dots)
{
dotMod = 0;
fputc('.', stdout);
fflush(stdout);
}
}
if (qType == gftProt && (tType == gftDnaX || tType == gftRnaX))
{
gvo->reportTargetStrand = TRUE;
gfAlignTrans(&conn, tSeqDir, &seq, minScore, tFileCache, gvo);
}
else if ((qType == gftRnaX || qType == gftDnaX) && (tType == gftDnaX || tType == gftRnaX))
{
gvo->reportTargetStrand = TRUE;
gfAlignTransTrans(&conn, tSeqDir, &seq, FALSE, minScore, tFileCache,
gvo, qType == gftRnaX);
if (qType == gftDnaX)
{
reverseComplement(seq.dna, seq.size);
close(conn);
conn = gfConnect(hostName, portName);
gfAlignTransTrans(&conn, tSeqDir, &seq, TRUE, minScore, tFileCache,
gvo, FALSE);
}
}
else if ((tType == gftDna || tType == gftRna) && (qType == gftDna || qType == gftRna))
{
gfAlignStrand(&conn, tSeqDir, &seq, FALSE, minScore, tFileCache, gvo);
conn = gfConnect(hostName, portName);
reverseComplement(seq.dna, seq.size);
gfAlignStrand(&conn, tSeqDir, &seq, TRUE, minScore, tFileCache, gvo);
}
else
{
errAbort("Comparisons between %s queries and %s databases not yet supported",
qTypeName, tTypeName);
}
gfOutputQuery(gvo, out);
}
if (out != stdout)
printf("Output is in %s\n", outName);
gfFileCacheFree(&tFileCache);
}
