int
main(int argc, char **argv) {
char *gkpStoreName = NULL;
char *outPrefix = NULL;
char *clrName = NULL;
uint32 libToDump = 0;
uint32 bgnID = 1;
uint32 endID = UINT32_MAX;
bool dumpAllReads = false;
bool dumpAllBases = false;
bool dumpOnlyDeleted = false;
bool dumpFASTQ = true;
bool dumpFASTA = false;
bool withLibName = true;
argc = AS_configure(argc, argv);
int arg = 1;
int err = 0;
while (arg < argc) {
if (strcmp(argv[arg], "-G") == 0) {
gkpStoreName = argv[++arg];
} else if (strcmp(argv[arg], "-o") == 0) {
outPrefix = argv[++arg];
} else if (strcmp(argv[arg], "-c") == 0) {
clrName = argv[++arg];
} else if (strcmp(argv[arg], "-l") == 0) {
libToDump = atoi(argv[++arg]);
} else if (strcmp(argv[arg], "-b") == 0) {
bgnID = atoi(argv[++arg]);
} else if (strcmp(argv[arg], "-e") == 0) {
endID = atoi(argv[++arg]);
} else if (strcmp(argv[arg], "-r") == 0) {
bgnID = atoi(argv[++arg]);
endID = bgnID;
} else if (strcmp(argv[arg], "-allreads") == 0) {
dumpAllReads = true;
} else if (strcmp(argv[arg], "-allbases") == 0) {
dumpAllBases = true;
} else if (strcmp(argv[arg], "-onlydeleted") == 0) {
dumpOnlyDeleted = true;
dumpAllReads = true; // Otherwise we won't report the deleted reads!
} else if (strcmp(argv[arg], "-fastq") == 0) {
dumpFASTQ = true;
dumpFASTA = false;
} else if (strcmp(argv[arg], "-fasta") == 0) {
dumpFASTQ = false;
dumpFASTA = true;
} else if (strcmp(argv[arg], "-nolibname") == 0) {
withLibName = false;
} else {
err++;
fprintf(stderr, "ERROR: unknown option '%s'\n", argv[arg]);
}
arg++;
}
if (gkpStoreName == NULL)
err++;
if (outPrefix == NULL)
err++;
if (err) {
fprintf(stderr, "usage: %s [...] -o fastq-prefix -g gkpStore\n", argv[0]);
fprintf(stderr, " -G gkpStore\n");
fprintf(stderr, " -o fastq-prefix write files fastq-prefix.(libname).fastq, ...\n");
fprintf(stderr, "\n");
fprintf(stderr, " -l libToDump output only read in library number libToDump (NOT IMPLEMENTED)\n");
fprintf(stderr, " -b id output starting at read 'id'\n");
fprintf(stderr, " -e id output stopping after read 'id'\n");
fprintf(stderr, "\n");
fprintf(stderr, " -c clearFile clear range file from OBT modules\n");
fprintf(stderr, " -allreads if a clear range file, lower case mask the deleted reads\n");
fprintf(stderr, " -allbases if a clear range file, lower case mask the non-clear bases\n");
fprintf(stderr, " -onlydeleted if a clear range file, only output deleted reads (the entire read)\n");
fprintf(stderr, "\n");
fprintf(stderr, " -r id output only the single read 'id'\n");
fprintf(stderr, "\n");
fprintf(stderr, " -fastq output is FASTQ format (with extension .fastq, default)\n");
fprintf(stderr, " -fasta output is FASTA format (with extension .fasta)\n");
fprintf(stderr, "\n");
fprintf(stderr, " -nolibname don't include the library name in the output file name\n");
fprintf(stderr, "\n");
if (gkpStoreName == NULL)
fprintf(stderr, "ERROR: no gkpStore (-G) supplied.\n");
if (outPrefix == NULL)
fprintf(stderr, "ERROR: no output prefix (-o) supplied.\n");
exit(1);
}
gkStore *gkpStore = new gkStore(gkpStoreName);
uint32 numReads = gkpStore->gkStore_getNumReads();
uint32 numLibs = gkpStore->gkStore_getNumLibraries();
clearRangeFile *clrRange = (clrName == NULL) ? NULL : new clearRangeFile(clrName, gkpStore);
if (bgnID < 1)
bgnID = 1;
if (numReads < endID)
endID = numReads;
if (endID < bgnID)
fprintf(stderr, "No reads to dump; reversed ranges make no sense: bgn="F_U32" end="F_U32"??\n", bgnID, endID);
fprintf(stderr, "Dumping reads from %u to %u (inclusive).\n", bgnID, endID);
libOutput **out = new libOutput * [numLibs + 1];
// Allocate outputs. If withLibName == false, all reads will artificially be in lib zero, the
// other files won't ever be created. Otherwise, the zeroth file won't ever be created.
out[0] = new libOutput(outPrefix, NULL);
for (uint32 i=1; i<=numLibs; i++)
out[i] = new libOutput(outPrefix, gkpStore->gkStore_getLibrary(i)->gkLibrary_libraryName());
// Grab a new readData, and iterate through reads to dump.
gkReadData *readData = new gkReadData;
for (uint32 rid=bgnID; rid<=endID; rid++) {
gkRead *read = gkpStore->gkStore_getRead(rid);
uint32 libID = (withLibName == false) ? 0 : read->gkRead_libraryID();
uint32 lclr = 0;
uint32 rclr = read->gkRead_sequenceLength();
bool ignore = false;
//fprintf(stderr, "READ %u claims id %u length %u in lib %u\n", rid, read->gkRead_readID(), read->gkRead_sequenceLength(), libID);
// If a clear range file is supplied, grab the clear range. If it hasn't been set, the default
// is the entire read.
if (clrRange) {
lclr = clrRange->bgn(rid);
rclr = clrRange->end(rid);
ignore = clrRange->isDeleted(rid);
}
// Abort if we're not dumping anything from this read
// - not in a library we care about
// - deleted, and not dumping all reads
// - not deleted, but only reporting deleted reads
if (((libToDump != 0) && (libID == libToDump)) ||
((dumpAllReads == false) && (ignore == true)) ||
((dumpOnlyDeleted == true) && (ignore == false)))
continue;
// And if we're told to ignore the read, and here, then the read was deleted and we're printing
// all reads. Reset the clear range to the whole read, the clear range is invalid.
if (ignore) {
rclr = read->gkRead_sequenceLength();
lclr = 0;
}
// Grab the sequence and quality.
gkpStore->gkStore_loadReadData(read, readData);
char *seq = readData->gkReadData_getSequence();
char *qlt = readData->gkReadData_getQualities();
uint32 len = rclr - lclr;
// Soft mask not-clear bases
if (dumpAllBases == true) {
for (uint32 i=0; i<lclr; i++)
seq[i] += (seq[i] >= 'A') ? 'a' - 'A' : 0;
for (uint32 i=lclr; i<rclr; i++)
seq[i] += (seq[i] >= 'A') ? 0 : 'A' - 'a';
for (uint32 i=rclr; seq[i]; i++)
seq[i] += (seq[i] >= 'A') ? 'a' - 'A' : 0;
rclr = read->gkRead_sequenceLength();
lclr = 0;
}
// Chop off the ends we're not printing.
seq += lclr;
seq[len] = 0;
qlt[len] = 0;
// And print the read.
if (dumpFASTQ)
AS_UTL_writeFastQ(out[libID]->getFASTQ(), seq, len, qlt, len,
"@"F_U32" clr="F_U32","F_U32"\n",
rid, lclr, rclr);
if (dumpFASTA)
AS_UTL_writeFastA(out[libID]->getFASTA(), seq, len, 0,
">"F_U32" clr="F_U32","F_U32"\n",
rid, lclr, rclr);
}
delete readData;
for (uint32 i=1; i<=numLibs; i++)
delete out[i];
delete [] out;
delete gkpStore;
exit(0);
}
int
main( int argc, char **argv) {
int ckptNum = NULLINDEX;
int makeMiniScaffolds = 1;
uint64 uidStart = 1230000;
UIDserver *uids = NULL;
GlobalData = new Globals_CGW();
argc = AS_configure(argc, argv);
int err=0;
int arg=1;
while (arg < argc) {
if (strcmp(argv[arg], "-p") == 0) {
ckptNum = GlobalData->setPrefix(argv[++arg]);
} else if (strcmp(argv[arg], "-c") == 0) {
strcpy(GlobalData->outputPrefix, argv[++arg]);
} else if (strcmp(argv[arg], "-g") == 0) {
strcpy(GlobalData->gkpStoreName, argv[++arg]);
} else if (strcmp(argv[arg], "-t") == 0) {
strcpy(GlobalData->tigStoreName, argv[++arg]);
} else if (strcmp(argv[arg], "-n") == 0) {
ckptNum = atoi(argv[++arg]);
} else if (strcmp(argv[arg], "-U") == 0) {
uidStart = 0;
} else if (strcmp(argv[arg], "-S") == 0) {
makeMiniScaffolds = 0;
} else {
fprintf(stderr, "unknown option '%s'\n", argv[arg]);
err = 1;
}
arg++;
}
if ((GlobalData->outputPrefix[0] == 0) ||
(GlobalData->gkpStoreName[0] == 0)) {
fprintf(stderr, "usage: %s [[-p prefix] | [-c name -g gkpstore -n ckptNum]] [-U] [-S]\n", argv[0]);
fprintf(stderr, " -p Attempt to locate the last checkpoint in directory 7-CGW.\n");
fprintf(stderr, " -c Look for checkpoints in 'name'\n");
fprintf(stderr, " -g Path to gkpStore\n");
fprintf(stderr, " -n Checkpoint number to load\n");
fprintf(stderr, " -U Use real UIDs for miniscaffolds, otherwise, UIDs start at 1230000\n");
fprintf(stderr, " -S Do NOT make mini scaffolds.\n");
exit(1);
}
uids = UIDserverInitialize(256, uidStart);
char *toprint = (char *)safe_malloc(sizeof(char) * (AS_READ_MAX_NORMAL_LEN + 51 + AS_READ_MAX_NORMAL_LEN + 2));
LoadScaffoldGraphFromCheckpoint(GlobalData->outputPrefix, ckptNum, FALSE);
int ifrag;
for (ifrag=0; ifrag < GetNumVA_CIFragT(ScaffoldGraph->CIFrags); ifrag++) {
CIFragT *frag = GetCIFragT(ScaffoldGraph->CIFrags, ifrag);
CIFragT *mate = NULL;
if (frag->flags.bits.isDeleted)
continue;
assert(frag->cid != NULLINDEX);
assert((frag->flags.bits.hasMate == 0) || (frag->mate_iid != 0));
// Fix for missing mates -- OBT used to not delete mate links, leaving
// dangling mates. Somebody else seems to be doing this too.
//
if (frag->flags.bits.hasMate) {
mate = GetCIFragT(ScaffoldGraph->CIFrags, frag->mate_iid);
if (mate == NULL)
frag->flags.bits.hasMate = 0;
}
// If this fragment is not chaff, we have nothing to do here.
//
if (GetGraphNode(ScaffoldGraph->CIGraph,frag->cid)->flags.bits.isChaff == 0)
continue;
// Print a singleton if there is no mate, the mate isn't chaff,
// or we were told to not make miniscaffolds.
//
if ((mate == NULL) ||
(mate->flags.bits.isChaff == 0) ||
(makeMiniScaffolds == 0)) {
AS_UID fUID = getFragmentClear(frag->read_iid, 0, toprint);
AS_UTL_writeFastA(stdout,
toprint, strlen(toprint), 0,
">%s /type=singleton\n", AS_UID_toString(fUID));
} else if ((mate != NULL) &&
(mate->flags.bits.isChaff == 1) &&
(makeMiniScaffolds == 1) &&
(frag->read_iid < mate->read_iid)) {
// make sure the following chain of Ns is divisible by three;
// the exact length is arbitrary but Doug Rusch points out that
// by making it divisible by 3, we can get lucky and maintain
// the phase of a protein ... which helps in the
// auto-annotation of environmental samples
AS_UID fUID = getFragmentClear(frag->read_iid, 0, toprint);
strcat(toprint, "NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN");
AS_UID mUID = getFragmentClear(mate->read_iid, 1, toprint + strlen(toprint));
AS_UTL_writeFastA(stdout,
toprint, strlen(toprint), 0,
">"F_U64" /type=mini_scaffold /frgs=(%s,%s)\n",
getUID(uids),
AS_UID_toString(fUID),
AS_UID_toString(mUID));
}
}
delete GlobalData;
exit(0);
}
int32
main(int32 argc, char **argv) {
int32 arg = 1;
int32 err = 0;
int32 hlp = 0;
char * gkpStoreName = NULL;
int32 gkpStorePart = 0;
char * msgFile = NULL;
char * outputFileName= NULL;
char * seqAn = NULL;
char * wrkDir = NULL;
char * seqStoreName = NULL;
int32 seqStoreVer = 0;
int32 seqStorePart = 0;
argc = AS_configure(argc, argv);
while (arg < argc) {
if (strcmp(argv[arg], "-c") == 0) {
msgFile = argv[++arg];
} else if (strcmp(argv[arg], "-G") == 0) {
gkpStoreName = argv[++arg];
} else if (strcmp(argv[arg], "-S") == 0) {
gkpStorePart = atoi(argv[++arg]);
} else if (strcmp(argv[arg], "-o") == 0) {
outputFileName = argv[++arg];
} else if (strcmp(argv[arg], "-s") == 0) {
seqAn = argv[++arg];
} else if (strcmp(argv[arg], "-w") == 0) {
wrkDir = argv[++arg];
} else if (strcmp(argv[arg], "-u") == 0) {
seqStoreName = argv[++arg];
} else if (strcmp(argv[arg], "-V") == 0) {
seqStoreVer = atoi(argv[++arg]);
} else if (strcmp(argv[arg], "-p") == 0) {
seqStorePart = atoi(argv[++arg]);
} else {
err++;
}
arg++;
}
if ((err) || (gkpStoreName == NULL) || (msgFile == NULL) || (outputFileName == NULL) || seqAn == NULL) {
fprintf(stderr, "USAGE: SeqAn_CNS -G <gkpStore> -c <input.cgb> -o <output.cgi> -s <seqan_executable> [-u seqstore, required for contig consensus] [-w working directory]\n");
exit(1);
}
gkStore *gkpStore = new gkStore(gkpStoreName, FALSE, FALSE);
gkpStore->gkStore_loadPartition(gkpStorePart);
gkFragment fr;
GenericMesg *pmesg;
tSequenceDB *sequenceDB = NULL;
FILE *infp = fopen(msgFile,"r");
FILE *tempReads;
FILE *outfp = fopen(outputFileName, "w");
char fileName[AS_SEQAN_MAX_BUFFER_LENGTH];
char *prefix = outputFileName;
getFileName(prefix, wrkDir, AS_SEQAN_INPUT_NAME, fileName);
int32 i = 0;
while ((EOF != ReadProtoMesg_AS(infp, &pmesg))) {
int32 freeMem = 0;
if (pmesg->t == MESG_IUM) {
IntUnitigMesg *ium_mesg = (IntUnitigMesg *)pmesg->m;
if (strlen(ium_mesg->consensus) == 0) {
tempReads = fopen(fileName,"w");
for (i =0; i < ium_mesg->num_frags; i++) {
// get the fragment sequence
gkpStore->gkStore_getFragment(ium_mesg->f_list[i].ident, &fr, GKFRAGMENT_QLT);
uint32 clrBeg = fr.gkFragment_getClearRegionBegin();
uint32 clrEnd = fr.gkFragment_getClearRegionEnd ();
char *seqStart = fr.gkFragment_getSequence();
char *seq = seqStart+clrBeg;
seq[clrEnd] = 0;
AS_UTL_writeFastA(tempReads,
seq, clrEnd-clrBeg,
">"F_IID","F_IID"\n", ium_mesg->f_list[i].position.bgn, ium_mesg->f_list[i].position.end);
}
fclose(tempReads);
updateRecord(ium_mesg, fileName, seqAn, prefix, wrkDir);
freeMem = 1;
}
WriteProtoMesg_AS(outfp, pmesg);
if (freeMem) {
safe_free(ium_mesg->consensus);
safe_free(ium_mesg->quality);
}
}
else if (pmesg->t == MESG_ICM) {
IntConConMesg *icm_mesg = (IntConConMesg *)pmesg->m;
if (seqStoreName == NULL) {
fprintf(stderr, "USAGE: The -u option is required for contig consensus\n");
exit(1);
}
if (sequenceDB == NULL) {
sequenceDB = openSequenceDB(seqStoreName, FALSE, seqStoreVer);
openSequenceDBPartition(sequenceDB, seqStorePart);
}
if (strlen(icm_mesg->consensus) == 0) {
tempReads = fopen(fileName,"w");
for (i =0; i < icm_mesg->num_pieces; i++) {
// get the fragment sequence
gkpStore->gkStore_getFragment(icm_mesg->pieces[i].ident, &fr, GKFRAGMENT_QLT);
uint32 clrBeg = fr.gkFragment_getClearRegionBegin();
uint32 clrEnd = fr.gkFragment_getClearRegionEnd ();
char *seqStart = fr.gkFragment_getSequence();
char *seq = seqStart+clrBeg;
seq[clrEnd] = 0;
AS_UTL_writeFastA(tempReads,
seq, clrEnd-clrBeg,
">"F_IID","F_IID"\n", icm_mesg->pieces[i].position.bgn, icm_mesg->pieces[i].position.end);
}
// now handle the unitig messages
for (i =0; i < icm_mesg->num_unitigs; i++) {
VA_TYPE(char) *ungappedSequence = CreateVA_char(0);
VA_TYPE(char) *ungappedQuality = CreateVA_char(0);
MultiAlignT *uma = loadMultiAlignTFromSequenceDB(sequenceDB, icm_mesg->unitigs[i].ident, 1);
assert(uma != NULL);
GetMultiAlignUngappedConsensus(uma, ungappedSequence, ungappedQuality);
char * seq = Getchar(ungappedSequence,0);
AS_UTL_writeFastA(tempReads,
seq, strlen(seq),
">"F_IID","F_IID"\n", icm_mesg->unitigs[i].position.bgn, icm_mesg->unitigs[i].position.end);
}
fclose(tempReads);
updateICMRecord(icm_mesg, fileName, seqAn, prefix, wrkDir);
freeMem = 1;
}
WriteProtoMesg_AS(outfp, pmesg);
if (freeMem) {
safe_free(icm_mesg->consensus);
safe_free(icm_mesg->quality);
}
}
}
fclose(infp);
fclose(outfp);
return 0;
}
