int main(int argc, char **argv)
{
ReadSet *allSequences = NULL;
SplayTable *splayTable;
int hashLength, hashLengthStep, hashLengthMax, h;
char *directory, *filename, *seqFilename, *buf;
boolean double_strand = true;
boolean multiple_kmers = false;
DIR *dir;
setProgramName("velveth");
if (argc < 4) {
printf("velveth - simple hashing program\n");
printf("Version %i.%i.%2.2i\n", VERSION_NUMBER,
RELEASE_NUMBER, UPDATE_NUMBER);
printf("\nCopyright 2007, 2008 Daniel Zerbino ([email protected])\n");
printf("This is free software; see the source for copying conditions. There is NO\n");
printf("warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.\n\n");
printf("Compilation settings:\n");
printf("CATEGORIES = %i\n", CATEGORIES);
printf("MAXKMERLENGTH = %i\n", MAXKMERLENGTH);
printf("\n");
printUsage();
return 0;
}
if ( strstr(argv[2],"," ) )
{
sscanf(argv[2],"%d,%d,%d",&hashLength,&hashLengthMax,&hashLengthStep);
multiple_kmers = true;
}
else
{
hashLength = atoi(argv[2]);
hashLengthMax = hashLength + 1;
hashLengthStep = 2;
}
if (hashLengthMax > MAXKMERLENGTH) {
velvetLog
("Velvet can't handle k-mers as long as %i! We'll stick to %i if you don't mind.\n",
hashLength, MAXKMERLENGTH);
hashLength = MAXKMERLENGTH;
} else if (hashLength <= 0) {
velvetLog("Invalid hash length: %s\n", argv[2]);
printUsage();
return 0;
} else if ( hashLength > hashLengthMax ) {
velvetLog("hashLengthMin <= hashLengthMax is required %s", argv[2]);
printUsage();
return 0;
}
if (hashLength % 2 == 0) {
velvetLog
("Velvet can't work with even length k-mers, such as %i. We'll use %i instead, if you don't mind.\n",
hashLength, hashLength - 1);
hashLength--;
}
if (hashLengthStep % 2 == 1) {
velvetLog
("Velvet can't work with an odd length k-mer step, such as %i. We'll use %i instead, if you don't mind.\n",
hashLengthStep, hashLengthStep - 1);
hashLengthStep--;
}
for (h = hashLength; h < hashLengthMax; h += hashLengthStep) {
resetWordFilter(h);
buf = mallocOrExit(strlen(argv[1]) + 100, char);
if ( multiple_kmers ) {
sprintf(buf,"%s_%d",argv[1],h);
directory = mallocOrExit(strlen(buf) + 100, char);
strcpy(directory,buf);
} else
directory = argv[1];
filename = mallocOrExit(strlen(directory) + 100, char);
seqFilename = mallocOrExit(strlen(directory) + 100, char);
dir = opendir(directory);
if (dir == NULL)
mkdir(directory, 0777);
else {
sprintf(buf, "%s/PreGraph", directory);
remove(buf);
sprintf(buf, "%s/Graph", directory);
remove(buf);
sprintf(buf, "%s/Graph2", directory);
remove(buf);
sprintf(buf, "%s/Graph3", directory);
remove(buf);
sprintf(buf, "%s/Graph4", directory);
remove(buf);
sprintf(buf, "%s/Log", directory);
remove(buf);
}
logInstructions(argc, argv, directory);
strcpy(seqFilename, directory);
strcat(seqFilename, "/Sequences");
if ( h == hashLength ) {
parseDataAndReadFiles(seqFilename, argc - 2, &(argv[2]), &double_strand);
} else {
sprintf(buf,"ln -s ../%s_%d/Sequences %s",argv[1],hashLength,seqFilename);
system(buf);
}
splayTable = newSplayTable(h, double_strand);
if (!allSequences)
allSequences = importReadSet(seqFilename);
velvetLog("%li sequences in total.\n", (long) allSequences->readCount);
strcpy(filename, directory);
strcat(filename, "/Roadmaps");
inputSequenceArrayIntoSplayTableAndArchive(allSequences,
splayTable, filename, seqFilename);
destroySplayTable(splayTable);
if (dir)
closedir(dir);
if (directory != argv[1])
free(directory);
free(filename);
free(seqFilename);
free(buf);
}
// Imports roadmap from the appropriate file format
// Memory allocated within the function
RoadMapArray *importRoadMapArray(char *filename)
{
FILE *file;
const int maxline = 100;
char *line = mallocOrExit(maxline, char);
RoadMap *array;
RoadMap *rdmap = NULL;
IDnum rdmapIndex = 0;
IDnum seqID;
Coordinate position, start, finish;
Annotation *nextAnnotation;
RoadMapArray *result = mallocOrExit(1, RoadMapArray);
IDnum sequenceCount;
IDnum annotationCount = 0;
short short_var;
long long_var;
long long longlong_var, longlong_var2, longlong_var3;
printf("Reading roadmap file %s\n", filename);
file = fopen(filename, "r");
if (!fgets(line, maxline, file))
exitErrorf(EXIT_FAILURE, true, "%s incomplete.", filename);
sscanf(line, "%ld\t%i\t%hi\n", &long_var, &(result->WORDLENGTH), &short_var);
sequenceCount = (IDnum) long_var;
resetWordFilter(result->WORDLENGTH);
result->length = sequenceCount;
array = mallocOrExit(sequenceCount, RoadMap);
result->array = array;
result->double_strand = (boolean) short_var;
while (fgets(line, maxline, file) != NULL)
if (line[0] != 'R')
annotationCount++;
result->annotations = callocOrExit(annotationCount, Annotation);
nextAnnotation = result->annotations;
fclose(file);
file = fopen(filename, "r");
if (!fgets(line, maxline, file))
exitErrorf(EXIT_FAILURE, true, "%s incomplete.", filename);
while (fgets(line, maxline, file) != NULL) {
if (line[0] == 'R') {
rdmap = getRoadMapInArray(result, rdmapIndex++);
rdmap->annotationCount = 0;
} else {
sscanf(line, "%ld\t%lld\t%lld\t%lld\n", &long_var,
&longlong_var, &longlong_var2, &longlong_var3);
seqID = (IDnum) long_var;
position = (Coordinate) longlong_var;
start = (Coordinate) longlong_var2;
finish = (Coordinate) longlong_var3;
nextAnnotation->sequenceID = seqID;
nextAnnotation->position = position;
nextAnnotation->start.coord = start;
nextAnnotation->finish.coord = finish;
if (seqID > 0)
nextAnnotation->length = finish - start;
else
nextAnnotation->length = start - finish;
rdmap->annotationCount++;
nextAnnotation++;
}
}
printf("%d roadmaps reads\n", rdmapIndex);
fclose(file);
free(line);
return result;
}
int main(int argc, char **argv)
{
ReadSet *allSequences = NULL;
SplayTable *splayTable;
int hashLength, hashLengthStep, hashLengthMax, h;
char *directory, *filename, *seqFilename, *baseSeqName, *buf;
char * token;
boolean double_strand = true;
boolean noHash = false;
boolean multiple_kmers = false;
char buffer[100];
DIR *dir;
setProgramName("velveth");
if (argc < 4) {
printf("velveth - simple hashing program\n");
printf("Version %i.%i.%2.2i\n", VERSION_NUMBER,
RELEASE_NUMBER, UPDATE_NUMBER);
printf("\nCopyright 2007, 2008 Daniel Zerbino ([email protected])\n");
printf("This is free software; see the source for copying conditions. There is NO\n");
printf("warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.\n\n");
printf("Compilation settings:\n");
printf("CATEGORIES = %i\n", CATEGORIES);
printf("MAXKMERLENGTH = %i\n", MAXKMERLENGTH);
#ifdef _OPENMP
puts("OPENMP");
#endif
#ifdef LONGSEQUENCES
puts("LONGSEQUENCES");
#endif
#ifdef BIGASSEMBLY
puts("BIGASSEMBLY");
#endif
#ifdef COLOR
puts("COLOR");
#endif
#ifdef DEBUG
puts("DEBUG");
#endif
printf("\n");
printUsage();
return 0;
}
strcpy(buffer, argv[2]);
token = strtok(buffer, ",");
hashLength = atoi(token);
token = strtok(NULL, ",");
if (token == NULL) {
multiple_kmers = false;
hashLengthMax = hashLength + 1;
} else {
multiple_kmers = true;
hashLengthMax = atoi(token);
}
token = strtok(NULL, ",");
if (token == NULL) {
hashLengthStep = 2;
} else {
hashLengthStep = atoi(token);
}
if (hashLength > MAXKMERLENGTH) {
velvetLog
("Velvet can't handle k-mers as long as %i! We'll stick to %i if you don't mind.\n",
hashLength, MAXKMERLENGTH);
hashLength = MAXKMERLENGTH;
}
if (hashLength <= 0) {
velvetLog("Invalid hash length: %s\n", argv[2]);
printUsage();
return 0;
}
if (hashLength % 2 == 0) {
velvetLog
("Velvet can't work with even length k-mers, such as %i. We'll use %i instead, if you don't mind.\n",
hashLength, hashLength - 1);
hashLength--;
}
if (multiple_kmers) {
if (hashLengthMax > MAXKMERLENGTH + 1) {
velvetLog
("Velvet can't handle k-mers as long as %i! We'll stick to %i if you don't mind.\n",
hashLengthMax, MAXKMERLENGTH + 1);
hashLengthMax = MAXKMERLENGTH + 1;
}
if (hashLengthMax <= hashLength) {
velvetLog("hashLengthMin < hashLengthMax is required %s", argv[2]);
printUsage();
return 0;
}
if (hashLengthStep <= 0) {
velvetLog("Non-positive hash length! Setting it to 2\n");
hashLengthStep = 2;
}
if (hashLengthStep % 2 == 1) {
velvetLog
("Velvet can't work with an odd length k-mer step, such as %i. We'll use %i instead, if you don't mind.\n",
hashLengthStep, hashLengthStep + 1);
hashLengthStep++;
}
}
// check if binary sequences should be used
int argIndex;
for (argIndex = 3; argIndex < argc; argIndex++)
if (strcmp(argv[argIndex], "-create_binary") == 0 || strcmp(argv[argIndex], "-reuse_binary") == 0)
setCreateBinary(true);
for (h = hashLength; h < hashLengthMax; h += hashLengthStep) {
resetWordFilter(h);
buf = mallocOrExit(2 * strlen(argv[1]) + 500, char);
if ( multiple_kmers ) {
sprintf(buf,"%s_%d",argv[1],h);
directory = mallocOrExit(strlen(buf) + 100, char);
strcpy(directory,buf);
} else
directory = argv[1];
filename = mallocOrExit(strlen(directory) + 100, char);
seqFilename = mallocOrExit(strlen(directory) + 100, char);
baseSeqName = mallocOrExit(100, char);
dir = opendir(directory);
if (dir == NULL)
mkdir(directory, 0777);
else {
sprintf(buf, "%s/PreGraph", directory);
remove(buf);
sprintf(buf, "%s/Graph", directory);
remove(buf);
sprintf(buf, "%s/Graph2", directory);
remove(buf);
sprintf(buf, "%s/Graph3", directory);
remove(buf);
sprintf(buf, "%s/Graph4", directory);
remove(buf);
}
logInstructions(argc, argv, directory);
strcpy(seqFilename, directory);
if (isCreateBinary()) {
// use the CNY unified seq writer
strcpy(baseSeqName, "/CnyUnifiedSeq");
// remove other style sequences file
sprintf(buf, "%s/Sequences", directory);
remove(buf);
} else {
strcpy(baseSeqName, "/Sequences");
// remove other style sequences file
sprintf(buf, "%s/CnyUnifiedSeq", directory);
remove(buf);
sprintf(buf, "%s/CnyUnifiedSeq.names", directory);
remove(buf);
}
strcat(seqFilename, baseSeqName);
if ( h == hashLength ) {
parseDataAndReadFiles(seqFilename, argc - 2, &(argv[2]), &double_strand, &noHash);
} else {
sprintf(buf,"rm -f %s",seqFilename);
if (system(buf)) {
velvetLog("Command failed!\n");
velvetLog("%s\n", buf);
#ifdef DEBUG
abort();
#endif
exit(1);
}
if (argv[1][0] == '/')
sprintf(buf,"ln -s %s_%d%s %s",argv[1],hashLength,baseSeqName,seqFilename);
else
sprintf(buf,"ln -s `pwd`/%s_%d%s %s",argv[1],hashLength,baseSeqName,seqFilename);
if (system(buf)) {
velvetLog("Command failed!\n");
velvetLog("%s\n", buf);
#ifdef DEBUG
abort();
#endif
exit(1);
}
}
if (noHash)
continue;
splayTable = newSplayTable(h, double_strand);
if (isCreateBinary()) {
allSequences = importCnyReadSet(seqFilename);
} else {
allSequences = importReadSet(seqFilename);
}
velvetLog("%li sequences in total.\n", (long) allSequences->readCount);
strcpy(filename, directory);
strcat(filename, "/Roadmaps");
inputSequenceArrayIntoSplayTableAndArchive(allSequences,
splayTable, filename, seqFilename);
destroySplayTable(splayTable);
if (dir)
closedir(dir);
if (directory != argv[1])
free(directory);
free(filename);
free(seqFilename);
free(baseSeqName);
free(buf);
if (allSequences) {
destroyReadSet(allSequences);
}
}