bntseq_t *bwa_open_nt(const char *prefix)
{
bntseq_t *ntbns;
char *str;
str = (char*)calloc(strlen(prefix) + 10, 1);
strcat(strcpy(str, prefix), ".nt");
ntbns = bns_restore(str);
free(str);
return ntbns;
}
int main(int argc, char *argv[])
{
bntseq_t *bns;
bns = bns_restore(argv[1]);
uint8_t *pac;
pac = calloc(bns->l_pac/2+2, 1);
fread(pac, 1, bns->l_pac/2+2, bns->fp_pac);
int i;
for(i = 0; i < bns->l_pac; ++i){
putchar( "ACGT#"[_get_pac(pac, i)]);
}
bns_destroy(bns);
}
extern "C" void bwa_seed2genome_init(const char *prefix, gap_opt_t *opt)
{
if (!opt)
opt=mygap_init_opt() ;
opt->mode=BWA_MODE_BAM_SE ;
bwt_opt = opt ;
{ // load BWT
char *str = (char*)calloc(strlen(prefix) + 10, 1);
strcpy(str, prefix); strcat(str, ".bwt"); bwt_bwt[0] = bwt_restore_bwt(str);
strcpy(str, prefix); strcat(str, ".sa"); bwt_restore_sa(str, bwt_bwt[0]);
strcpy(str, prefix); strcat(str, ".rbwt"); bwt_bwt[1] = bwt_restore_bwt(str);
strcpy(str, prefix); strcat(str, ".rsa"); bwt_restore_sa(str, bwt_bwt[1]);
free(str);
bwt_bns = bns_restore(prefix);
}
}
int bwa_bwtsw2(int argc, char *argv[])
{
bsw2opt_t *opt;
bwt_t *target[2];
char buf[1024];
bntseq_t *bns;
int c;
opt = bsw2_init_opt();
srand48(11);
optind = 1;
while ((c = getopt(argc, argv, "q:r:a:b:t:T:w:d:z:m:y:s:c:N:Hf:")) >= 0) {
switch (c) {
case 'q': opt->q = atoi(optarg); break;
case 'r': opt->r = atoi(optarg); break;
case 'a': opt->a = atoi(optarg); break;
case 'b': opt->b = atoi(optarg); break;
case 'w': opt->bw = atoi(optarg); break;
case 'T': opt->t = atoi(optarg); break;
case 't': opt->n_threads = atoi(optarg); break;
case 'z': opt->z = atoi(optarg); break;
case 'y': opt->yita = atof(optarg); break;
case 's': opt->is = atoi(optarg); break;
case 'm': opt->mask_level = atof(optarg); break;
case 'c': opt->coef = atof(optarg); break;
case 'N': opt->t_seeds = atoi(optarg); break;
case 'H': opt->hard_clip = 1; break;
case 'f': xreopen(optarg, "w", stdout); break;
}
}
opt->qr = opt->q + opt->r;
if (optind + 2 > argc) {
fprintf(stderr, "\n");
fprintf(stderr, "Usage: bwa bwasw [options] <target.prefix> <query.fa>\n\n");
fprintf(stderr, "Options: -a INT score for a match [%d]\n", opt->a);
fprintf(stderr, " -b INT mismatch penalty [%d]\n", opt->b);
fprintf(stderr, " -q INT gap open penalty [%d]\n", opt->q);
fprintf(stderr, " -r INT gap extension penalty [%d]\n", opt->r);
// fprintf(stderr, " -y FLOAT error recurrence coef. (4..16) [%.1f]\n", opt->yita);
fprintf(stderr, "\n");
fprintf(stderr, " -t INT number of threads [%d]\n", opt->n_threads);
fprintf(stderr, " -s INT size of a chunk of reads [%d]\n", opt->chunk_size);
fprintf(stderr, "\n");
fprintf(stderr, " -w INT band width [%d]\n", opt->bw);
fprintf(stderr, " -m FLOAT mask level [%.2f]\n", opt->mask_level);
fprintf(stderr, "\n");
fprintf(stderr, " -T INT score threshold divided by a [%d]\n", opt->t);
fprintf(stderr, " -s INT maximum seeding interval size [%d]\n", opt->is);
fprintf(stderr, " -z INT Z-best [%d]\n", opt->z);
fprintf(stderr, " -N INT # seeds to trigger reverse alignment [%d]\n", opt->t_seeds);
fprintf(stderr, " -c FLOAT coefficient of length-threshold adjustment [%.1f]\n", opt->coef);
fprintf(stderr, " -H in SAM output, use hard clipping rather than soft\n");
fprintf(stderr, " -f FILE file to output results to instead of stdout\n\n");
fprintf(stderr, "Note: For long Illumina, 454 and Sanger reads, assembly contigs, fosmids and\n");
fprintf(stderr, " BACs, the default setting usually works well. For the current PacBio\n");
fprintf(stderr, " reads (end of 2010), '-b5 -q2 -r1 -z10' is recommended. One may also\n");
fprintf(stderr, " increase '-z' for better sensitivity.\n");
fprintf(stderr, "\n");
if (0) {
double c, theta, eps, delta;
c = opt->a / log(opt->yita);
theta = exp(-opt->b / c) / opt->yita;
eps = exp(-opt->q / c);
delta = exp(-opt->r / c);
fprintf(stderr, "mismatch: %lf, gap_open: %lf, gap_ext: %lf\n\n",
theta, eps, delta);
}
return 1;
}
// adjust opt for opt->a
opt->t *= opt->a;
opt->coef *= opt->a;
strcpy(buf, argv[optind]); target[0] = bwt_restore_bwt(strcat(buf, ".bwt"));
strcpy(buf, argv[optind]); bwt_restore_sa(strcat(buf, ".sa"), target[0]);
strcpy(buf, argv[optind]); target[1] = bwt_restore_bwt(strcat(buf, ".rbwt"));
strcpy(buf, argv[optind]); bwt_restore_sa(strcat(buf, ".rsa"), target[1]);
bns = bns_restore(argv[optind]);
bsw2_aln(opt, bns, target, argv[optind+1]);
bns_destroy(bns);
bwt_destroy(target[0]); bwt_destroy(target[1]);
free(opt);
fflush(stdout);
xreopen("/dev/tty","w",stdout);
return 0;
}
int bwa_bwtsw2(int argc, char *argv[])
{
extern char *bwa_infer_prefix(const char *hint);
bsw2opt_t *opt;
bwt_t *target;
char buf[1024], *prefix;
bntseq_t *bns;
int c;
opt = bsw2_init_opt();
srand48(11);
while ((c = getopt(argc, argv, "q:r:a:b:t:T:w:d:z:m:s:c:N:Hf:MI:SG:C")) >= 0) {
switch (c) {
case 'q': opt->q = atoi(optarg); break;
case 'r': opt->r = atoi(optarg); break;
case 'a': opt->a = atoi(optarg); break;
case 'b': opt->b = atoi(optarg); break;
case 'w': opt->bw = atoi(optarg); break;
case 'T': opt->t = atoi(optarg); break;
case 't': opt->n_threads = atoi(optarg); break;
case 'z': opt->z = atoi(optarg); break;
case 's': opt->is = atoi(optarg); break;
case 'm': opt->mask_level = atof(optarg); break;
case 'c': opt->coef = atof(optarg); break;
case 'N': opt->t_seeds = atoi(optarg); break;
case 'M': opt->multi_2nd = 1; break;
case 'H': opt->hard_clip = 1; break;
case 'f': xreopen(optarg, "w", stdout); break;
case 'I': opt->max_ins = atoi(optarg); break;
case 'S': opt->skip_sw = 1; break;
case 'C': opt->cpy_cmt = 1; break;
case 'G': opt->max_chain_gap = atoi(optarg); break;
}
}
opt->qr = opt->q + opt->r;
if (optind + 2 > argc) {
fprintf(stderr, "\n");
fprintf(stderr, "Usage: bwa bwasw [options] <target.prefix> <query.fa> [query2.fa]\n\n");
fprintf(stderr, "Options: -a INT score for a match [%d]\n", opt->a);
fprintf(stderr, " -b INT mismatch penalty [%d]\n", opt->b);
fprintf(stderr, " -q INT gap open penalty [%d]\n", opt->q);
fprintf(stderr, " -r INT gap extension penalty [%d]\n", opt->r);
fprintf(stderr, " -w INT band width [%d]\n", opt->bw);
fprintf(stderr, " -m FLOAT mask level [%.2f]\n", opt->mask_level);
fprintf(stderr, "\n");
fprintf(stderr, " -t INT number of threads [%d]\n", opt->n_threads);
fprintf(stderr, " -f FILE file to output results to instead of stdout\n");
fprintf(stderr, " -H in SAM output, use hard clipping instead of soft clipping\n");
fprintf(stderr, " -C copy FASTA/Q comment to SAM output\n");
fprintf(stderr, " -M mark multi-part alignments as secondary\n");
fprintf(stderr, " -S skip Smith-Waterman read pairing\n");
fprintf(stderr, " -I INT ignore pairs with insert >=INT for inferring the size distr [%d]\n", opt->max_ins);
fprintf(stderr, "\n");
fprintf(stderr, " -T INT score threshold divided by a [%d]\n", opt->t);
fprintf(stderr, " -c FLOAT coefficient of length-threshold adjustment [%.1f]\n", opt->coef);
fprintf(stderr, " -z INT Z-best [%d]\n", opt->z);
fprintf(stderr, " -s INT maximum seeding interval size [%d]\n", opt->is);
fprintf(stderr, " -N INT # seeds to trigger rev aln; 2*INT is also the chaining threshold [%d]\n", opt->t_seeds);
fprintf(stderr, " -G INT maximum gap size during chaining [%d]\n", opt->max_chain_gap);
fprintf(stderr, "\n");
fprintf(stderr, "Note: For long Illumina, 454 and Sanger reads, assembly contigs, fosmids and\n");
fprintf(stderr, " BACs, the default setting usually works well. For the current PacBio\n");
fprintf(stderr, " reads (end of 2010), '-b5 -q2 -r1 -z10' is recommended. One may also\n");
fprintf(stderr, " increase '-z' for better sensitivity.\n");
fprintf(stderr, "\n");
return 1;
}
// adjust opt for opt->a
opt->t *= opt->a;
opt->coef *= opt->a;
if ((prefix = bwa_infer_prefix(argv[optind])) == 0) {
fprintf(stderr, "[%s] fail to locate the index\n", __func__);
return 0;
}
strcpy(buf, prefix); target = bwt_restore_bwt(strcat(buf, ".bwt"));
strcpy(buf, prefix); bwt_restore_sa(strcat(buf, ".sa"), target);
bns = bns_restore(prefix);
bsw2_aln(opt, bns, target, argv[optind+1], optind+2 < argc? argv[optind+2] : 0);
bns_destroy(bns);
bwt_destroy(target);
free(opt); free(prefix);
return 0;
}
int bwa_bwtsw2(int argc, char *argv[])
{
bsw2opt_t *opt;
bwt_t *target[2];
char buf[1024];
bntseq_t *bns;
int c;
opt = bsw2_init_opt();
srand48(11);
while ((c = getopt(argc, argv, "q:r:a:b:t:T:w:d:z:m:y:s:c:N:H")) >= 0) {
switch (c) {
case 'q': opt->q = atoi(optarg); break;
case 'r': opt->r = atoi(optarg); break;
case 'a': opt->a = atoi(optarg); break;
case 'b': opt->b = atoi(optarg); break;
case 'w': opt->bw = atoi(optarg); break;
case 'T': opt->t = atoi(optarg); break;
case 't': opt->n_threads = atoi(optarg); break;
case 'z': opt->z = atoi(optarg); break;
case 'y': opt->yita = atof(optarg); break;
case 's': opt->is = atoi(optarg); break;
case 'm': opt->mask_level = atof(optarg); break;
case 'c': opt->coef = atof(optarg); break;
case 'N': opt->t_seeds = atoi(optarg); break;
case 'H': opt->hard_clip = 1; break;
}
}
opt->qr = opt->q + opt->r;
if (optind + 2 > argc) {
fprintf(stderr, "\n");
fprintf(stderr, "Usage: bwa dbwtsw [options] <target.prefix> <query.fa>\n\n");
fprintf(stderr, "Options: -a INT score for a match [%d]\n", opt->a);
fprintf(stderr, " -b INT mismatch penalty [%d]\n", opt->b);
fprintf(stderr, " -q INT gap open penalty [%d]\n", opt->q);
fprintf(stderr, " -r INT gap extension penalty [%d]\n", opt->r);
// fprintf(stderr, " -y FLOAT error recurrence coef. (4..16) [%.1f]\n", opt->yita);
fprintf(stderr, "\n");
fprintf(stderr, " -t INT nmber of threads [%d]\n", opt->n_threads);
fprintf(stderr, " -s INT size of a chunk of reads [%d]\n", opt->chunk_size);
fprintf(stderr, "\n");
fprintf(stderr, " -w INT band width [%d]\n", opt->bw);
fprintf(stderr, " -m FLOAT mask level [%.2f]\n", opt->mask_level);
fprintf(stderr, "\n");
fprintf(stderr, " -T INT score threshold divided by a [%d]\n", opt->t);
fprintf(stderr, " -s INT maximum seeding interval size [%d]\n", opt->is);
fprintf(stderr, " -z INT Z-best [%d]\n", opt->z);
fprintf(stderr, " -N INT # seeds to trigger reverse alignment [%d]\n", opt->t_seeds);
fprintf(stderr, " -c FLOAT coefficient of length-threshold adjustment [%.1f]\n", opt->coef);
fprintf(stderr, " -H in SAM output, use hard clipping rather than soft\n");
fprintf(stderr, "\n");
{
double c, theta, eps, delta;
c = opt->a / log(opt->yita);
theta = exp(-opt->b / c) / opt->yita;
eps = exp(-opt->q / c);
delta = exp(-opt->r / c);
fprintf(stderr, "mismatch: %lf, gap_open: %lf, gap_ext: %lf\n\n",
theta, eps, delta);
}
return 1;
}
// adjust opt for opt->a
opt->t *= opt->a;
opt->coef *= opt->a;
strcpy(buf, argv[optind]); target[0] = bwt_restore_bwt(strcat(buf, ".bwt"));
strcpy(buf, argv[optind]); bwt_restore_sa(strcat(buf, ".sa"), target[0]);
strcpy(buf, argv[optind]); target[1] = bwt_restore_bwt(strcat(buf, ".rbwt"));
strcpy(buf, argv[optind]); bwt_restore_sa(strcat(buf, ".rsa"), target[1]);
bns = bns_restore(argv[optind]);
bsw2_aln(opt, bns, target, argv[optind+1]);
bns_destroy(bns);
bwt_destroy(target[0]); bwt_destroy(target[1]);
free(opt);
return 0;
}