int main(int argc, char *argv[])
{
fml_opt_t opt;
int c, n_seqs, n_utg, gfa_out = 0;
bseq1_t *seqs;
fml_utg_t *utg;
fml_opt_init(&opt);
while ((c = getopt(argc, argv, "gOAe:l:r:t:c:d:v:")) >= 0) {
if (c == 'e') opt.ec_k = atoi(optarg);
else if (c == 'l') opt.min_asm_ovlp = atoi(optarg);
else if (c == 'r') opt.mag_opt.min_dratio1 = atof(optarg);
else if (c == 'A') opt.mag_opt.flag |= MAG_F_AGGRESSIVE;
else if (c == 'O') opt.mag_opt.flag &= ~MAG_F_POPOPEN;
else if (c == 'd') opt.mag_opt.max_bdiff = atoi(optarg);
else if (c == 't') opt.n_threads = atoi(optarg);
else if (c == 'g') gfa_out = 1;
else if (c == 'v') fm_verbose = atoi(optarg);
else if (c == 'c') {
char *p;
opt.min_cnt = strtol(optarg, &p, 10);
if (*p == ',') opt.max_cnt = strtol(p + 1, &p, 10);
}
}
if (argc == optind) {
fprintf(stderr, "Usage: fml-asm [options] <in.fq>\n");
fprintf(stderr, "Options:\n");
fprintf(stderr, " -e INT k-mer length for error correction (0 for auto; -1 to disable) [%d]\n", opt.ec_k);
fprintf(stderr, " -c INT1[,INT2] range of k-mer & read count thresholds for ec and graph cleaning [%d,%d]\n", opt.min_cnt, opt.max_cnt);
fprintf(stderr, " -l INT min overlap length during initial assembly [%d]\n", opt.min_asm_ovlp);
fprintf(stderr, " -r FLOAT drop an overlap if its length is below maxOvlpLen*FLOAT [%g]\n", opt.mag_opt.min_dratio1);
fprintf(stderr, " -t INT number of threads (don't use multi-threading for small data sets) [%d]\n", opt.n_threads);
fprintf(stderr, " -d INT retain a bubble if one side is longer than the other side by >INT-bp [%d]\n", opt.mag_opt.max_bdiff);
fprintf(stderr, " -A discard heterozygotes (apply this to assemble bacterial genomes; override -O)\n");
fprintf(stderr, " -O don't apply aggressive tip trimming\n");
fprintf(stderr, " -g output the assembly graph in the GFA format\n");
return 1;
}
seqs = bseq_read(argv[optind], &n_seqs);
utg = fml_assemble(&opt, n_seqs, seqs, &n_utg);
if (!gfa_out) fml_utg_print(n_utg, utg);
else fml_utg_print_gfa(n_utg, utg);
fml_utg_destroy(n_utg, utg);
return 0;
}
int main(int argc, char *argv[])
{
fml_opt_t opt;
int c, n_seqs, n_utg;
bseq1_t *seqs;
fml_utg_t *utg;
fml_opt_init(&opt);
while ((c = getopt(argc, argv, "Ae:l:r:t:c:")) >= 0) {
if (c == 'e') opt.ec_k = atoi(optarg);
else if (c == 'l') opt.min_asm_ovlp = atoi(optarg);
else if (c == 'r') opt.mag_opt.min_dratio1 = atof(optarg);
else if (c == 'A') opt.mag_opt.flag |= MAG_F_AGGRESSIVE;
else if (c == 't') opt.n_threads = atoi(optarg);
else if (c == 'c') {
char *p;
opt.min_cnt = strtol(optarg, &p, 10);
if (*p == ',') opt.max_cnt = strtol(p + 1, &p, 10);
}
}
if (argc == optind) {
fprintf(stderr, "Usage: fml-asm [options] <in.fq>\n");
fprintf(stderr, "Options:\n");
fprintf(stderr, " -e INT k-mer length for error correction (0 for auto; -1 to disable) [%d]\n", opt.ec_k);
fprintf(stderr, " -c INT1[,INT2] range of k-mer & read count thresholds for ec and graph cleaning [%d,%d]\n", opt.min_cnt, opt.max_cnt);
fprintf(stderr, " -l INT min overlap length during initial assembly [%d]\n", opt.min_asm_ovlp);
fprintf(stderr, " -r FLOAT drop an overlap if its length is below maxOvlpLen*FLOAT [%g]\n", opt.mag_opt.min_dratio1);
fprintf(stderr, " -t INT number of threads (don't use multi-threading for small data sets) [%d]\n", opt.n_threads);
fprintf(stderr, " -A discard heterozygotes (apply this to assemble bacterial genomes)\n");
return 1;
}
seqs = bseq_read(argv[optind], &n_seqs);
utg = fml_assemble(&opt, n_seqs, seqs, &n_utg);
fml_utg_print(n_utg, utg);
fml_utg_destroy(n_utg, utg);
return 0;
}
int main(int argc, char **argv)
{
fml_opt_t options;
bseq1_t *seqs = NULL;
PONE_READ reads = NULL;
size_t readCount = 0;
ERR_VALUE ret = ERR_INTERNAL_ERROR;
fml_opt_init(&options);
options.n_threads = omp_get_num_procs();
options.ec_k = 31;
utils_allocator_init(options.n_threads);
fprintf(stderr, "Loading reads from %s...\n", argv[1]);
ret = input_get_reads(argv[1], "sam", &reads, &readCount);
if (ret == ERR_SUCCESS) {
fprintf(stderr, "Converting to fermi-lite format...\n");
ret = utils_calloc(readCount, sizeof(bseq1_t), &seqs);
if (ret == ERR_SUCCESS) {
for (size_t i = 0; i < readCount; ++i) {
memset(seqs + i, 0, sizeof(seqs[i]));
seqs[i].l_seq = reads[i].ReadSequenceLen;
read_quality_encode(reads + i);
seqs[i].seq = _copy_string(reads[i].ReadSequence, reads[i].ReadSequenceLen);
if (reads[i].Quality != NULL)
seqs[i].qual = _copy_string(reads[i].Quality, reads[i].QualityLen);
read_quality_decode(reads + i);
}
fml_opt_adjust(&options, readCount, seqs);
fprintf(stderr, "Correcting...\n");
fml_correct(&options, readCount, seqs);
fprintf(stderr, "Fitting unique k-mers...\n");
fml_fltuniq(&options, readCount, seqs);
fprintf(stderr, "Converting back to our format...\n");
for (size_t i = 0; i < readCount; ++i) {
if (reads[i].ReadSequenceLen != seqs[i].l_seq) {
utils_copy_string("*", &reads[i].CIGAR);
reads[i].CIGARLen = 1;
}
reads[i].ReadSequenceLen = seqs[i].l_seq;
reads[i].QualityLen = seqs[i].l_seq;
ret = utils_copy_string(seqs[i].seq, &reads[i].ReadSequence);
if (ret == ERR_SUCCESS)
ret = utils_copy_string(seqs[i].qual, &reads[i].Quality);
for (size_t j = 0; j < reads[i].ReadSequenceLen; ++j)
reads[i].ReadSequence[j] = toupper(reads[i].ReadSequence[j]);
if (reads[i].ReadSequenceLen > 0 && reads[i].QualityLen > 0)
read_write_sam(stdout, reads + i);
read_quality_decode(reads + i);
}
fprintf(stderr, "Freeing fermi-lite resources...\n");
utils_free(seqs);
}
fprintf(stderr, "Freeing our reads...\n");
read_set_destroy(reads, readCount);
}
return 0;
}
