zzh
/
minimap2
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

r477: renamed multi_seg to frag_mode

This commit is contained in:
Heng Li 2017-10-05 15:48:17 -04:00
parent ae2adf04d4
commit 9c5767f9ed
5 changed files with 25 additions and 24 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
bseq.c
View File

@ -61,7 +61,7 @@ static inline void kseq2bseq(kseq_t *ks, mm_bseq1_t *s, int with_qual)
s->l_seq = ks->seq.l; s->l_seq = ks->seq.l;
} }
mm_bseq1_t *mm_bseq_read2(mm_bseq_file_t *fp, int chunk_size, int with_qual, int check_name, int *n_) mm_bseq1_t *mm_bseq_read2(mm_bseq_file_t *fp, int chunk_size, int with_qual, int frag_mode, int *n_)
{ {
int64_t size = 0; int64_t size = 0;
kvec_t(mm_bseq1_t) a = {0,0,0}; kvec_t(mm_bseq1_t) a = {0,0,0};
@ -81,7 +81,7 @@ mm_bseq1_t *mm_bseq_read2(mm_bseq_file_t *fp, int chunk_size, int with_qual, int
kseq2bseq(ks, s, with_qual); kseq2bseq(ks, s, with_qual);
size += s->l_seq; size += s->l_seq;
if (size >= chunk_size) { if (size >= chunk_size) {
if (check_name && a.a[a.n-1].l_seq < CHECK_PAIR_THRES) { if (frag_mode && a.a[a.n-1].l_seq < CHECK_PAIR_THRES) {
while (kseq_read(ks) >= 0) { while (kseq_read(ks) >= 0) {
kseq2bseq(ks, &fp->s, with_qual); kseq2bseq(ks, &fp->s, with_qual);
if (mm_qname_same(fp->s.name, a.a[a.n-1].name)) { if (mm_qname_same(fp->s.name, a.a[a.n-1].name)) {
@ -102,7 +102,7 @@ mm_bseq1_t *mm_bseq_read(mm_bseq_file_t *fp, int chunk_size, int with_qual, int
return mm_bseq_read2(fp, chunk_size, with_qual, 0, n_); return mm_bseq_read2(fp, chunk_size, with_qual, 0, n_);
} }
mm_bseq1_t *mm_bseq_read_multi(int n_fp, mm_bseq_file_t **fp, int chunk_size, int with_qual, int *n_) mm_bseq1_t *mm_bseq_read_frag(int n_fp, mm_bseq_file_t **fp, int chunk_size, int with_qual, int *n_)
{ {
int i; int i;
int64_t size = 0; int64_t size = 0;

4
bseq.h
View File

@ -18,9 +18,9 @@ typedef struct {
mm_bseq_file_t *mm_bseq_open(const char *fn); mm_bseq_file_t *mm_bseq_open(const char *fn);
void mm_bseq_close(mm_bseq_file_t *fp); void mm_bseq_close(mm_bseq_file_t *fp);
mm_bseq1_t *mm_bseq_read2(mm_bseq_file_t *fp, int chunk_size, int with_qual, int check_name, int *n_); mm_bseq1_t *mm_bseq_read2(mm_bseq_file_t *fp, int chunk_size, int with_qual, int frag_mode, int *n_);
mm_bseq1_t *mm_bseq_read(mm_bseq_file_t *fp, int chunk_size, int with_qual, int *n_); mm_bseq1_t *mm_bseq_read(mm_bseq_file_t *fp, int chunk_size, int with_qual, int *n_);
mm_bseq1_t *mm_bseq_read_multi(int n_fp, mm_bseq_file_t **fp, int chunk_size, int with_qual, int *n_); mm_bseq1_t *mm_bseq_read_frag(int n_fp, mm_bseq_file_t **fp, int chunk_size, int with_qual, int *n_);
int mm_bseq_eof(mm_bseq_file_t *fp); int mm_bseq_eof(mm_bseq_file_t *fp);
extern unsigned char seq_nt4_table[256]; extern unsigned char seq_nt4_table[256];

15
main.c
View File

@ -6,7 +6,7 @@
#include "mmpriv.h" #include "mmpriv.h"
#include "getopt.h" #include "getopt.h"
#define MM_VERSION "2.2-r476-dirty" #define MM_VERSION "2.2-r477-dirty"
#ifdef __linux__ #ifdef __linux__
#include <sys/resource.h> #include <sys/resource.h>
@ -131,8 +131,8 @@ int main(int argc, char *argv[])
else if (c == 0 && long_idx ==13) opt.flag |= MM_F_SR; // --sr else if (c == 0 && long_idx ==13) opt.flag |= MM_F_SR; // --sr
else if (c == 0 && long_idx == 14) { // --frag else if (c == 0 && long_idx == 14) { // --frag
if (optarg == 0 || strcmp(optarg, "yes") == 0 || strcmp(optarg, "y") == 0) if (optarg == 0 || strcmp(optarg, "yes") == 0 || strcmp(optarg, "y") == 0)
opt.flag |= MM_F_MULTI_SEG; opt.flag |= MM_F_FRAG_MODE;
else opt.flag &= ~MM_F_MULTI_SEG; else opt.flag &= ~MM_F_FRAG_MODE;
} else if (c == 0 && long_idx == 15) { // --print-2nd } else if (c == 0 && long_idx == 15) { // --print-2nd
if (optarg == 0 || strcmp(optarg, "yes") == 0 || strcmp(optarg, "y") == 0) if (optarg == 0 || strcmp(optarg, "yes") == 0 || strcmp(optarg, "y") == 0)
opt.flag &= ~MM_F_NO_PRINT_2ND; opt.flag &= ~MM_F_NO_PRINT_2ND;
@ -180,7 +180,7 @@ int main(int argc, char *argv[])
} }
} }
if (max_gap_ref > 0) { if (max_gap_ref > 0) {
if (opt.flag & MM_F_MULTI_SEG) if (opt.flag & MM_F_FRAG_MODE)
opt.max_gap_ref = max_gap_ref; opt.max_gap_ref = max_gap_ref;
if (opt.flag & MM_F_SPLICE) if (opt.flag & MM_F_SPLICE)
opt.max_gap_ref = opt.bw = max_gap_ref; opt.max_gap_ref = opt.bw = max_gap_ref;
@ -267,11 +267,12 @@ int main(int argc, char *argv[])
__func__, realtime() - mm_realtime0, cputime() / (realtime() - mm_realtime0), mi->n_seq); __func__, realtime() - mm_realtime0, cputime() / (realtime() - mm_realtime0), mi->n_seq);
if (argc != optind + 1) mm_mapopt_update(&opt, mi); if (argc != optind + 1) mm_mapopt_update(&opt, mi);
if (mm_verbose >= 3) mm_idx_stat(mi); if (mm_verbose >= 3) mm_idx_stat(mi);
if (opt.flag & MM_F_MULTI_SEG) if (!(opt.flag & MM_F_FRAG_MODE)) {
mm_map_file_multi_seg(mi, argc - (optind + 1), (const char**)&argv[optind + 1], &opt, n_threads);
else
for (i = optind + 1; i < argc; ++i) for (i = optind + 1; i < argc; ++i)
mm_map_file(mi, argv[i], &opt, n_threads); mm_map_file(mi, argv[i], &opt, n_threads);
} else {
mm_map_file_frag(mi, argc - (optind + 1), (const char**)&argv[optind + 1], &opt, n_threads);
}
mm_idx_destroy(mi); mm_idx_destroy(mi);
} }
mm_idx_reader_close(idx_rdr); mm_idx_reader_close(idx_rdr);

20
map.c
View File

@ -80,7 +80,7 @@ int mm_set_opt(const char *preset, mm_idxopt_t *io, mm_mapopt_t *mo)
mo->min_dp_max = 200; mo->min_dp_max = 200;
} else if (strcmp(preset, "short") == 0 || strcmp(preset, "sr") == 0) { } else if (strcmp(preset, "short") == 0 || strcmp(preset, "sr") == 0) {
io->is_hpc = 0, io->k = 21, io->w = 11; io->is_hpc = 0, io->k = 21, io->w = 11;
mo->flag |= MM_F_SR | MM_F_MULTI_SEG | MM_F_NO_PRINT_2ND; mo->flag |= MM_F_SR | MM_F_FRAG_MODE | MM_F_NO_PRINT_2ND;
mo->pe_ori = 0<<1|1; // FR mo->pe_ori = 0<<1|1; // FR
mo->a = 2, mo->b = 8, mo->q = 12, mo->e = 2, mo->q2 = 32, mo->e2 = 1; mo->a = 2, mo->b = 8, mo->q = 12, mo->e = 2, mo->q2 = 32, mo->e2 = 1;
mo->max_gap = 200; mo->max_gap = 200;
@ -261,7 +261,7 @@ static mm_reg1_t *align_regs(const mm_mapopt_t *opt, const mm_idx_t *mi, void *k
return regs; return regs;
} }
void mm_map_multi(const mm_idx_t *mi, int n_segs, const int *qlens, const char **seqs, int *n_regs, mm_reg1_t **regs, mm_tbuf_t *b, const mm_mapopt_t *opt, const char *qname) void mm_map_frag(const mm_idx_t *mi, int n_segs, const int *qlens, const char **seqs, int *n_regs, mm_reg1_t **regs, mm_tbuf_t *b, const mm_mapopt_t *opt, const char *qname)
{ {
int i, j, max_gap_ref, rep_len, qlen_sum, n_regs0, rechain = 0; int i, j, max_gap_ref, rep_len, qlen_sum, n_regs0, rechain = 0;
uint32_t hash; uint32_t hash;
@ -350,7 +350,7 @@ void mm_map_multi(const mm_idx_t *mi, int n_segs, const int *qlens, const char *
mm_reg1_t *mm_map(const mm_idx_t *mi, int qlen, const char *seq, int *n_regs, mm_tbuf_t *b, const mm_mapopt_t *opt, const char *qname) mm_reg1_t *mm_map(const mm_idx_t *mi, int qlen, const char *seq, int *n_regs, mm_tbuf_t *b, const mm_mapopt_t *opt, const char *qname)
{ {
mm_reg1_t *regs; mm_reg1_t *regs;
mm_map_multi(mi, 1, &qlen, &seq, n_regs, &regs, b, opt, qname); mm_map_frag(mi, 1, &qlen, &seq, n_regs, &regs, b, opt, qname);
return regs; return regs;
} }
@ -391,7 +391,7 @@ static void worker_for(void *_data, long i, int tid) // kt_for() callback
qlens[j] = s->seq[off + j].l_seq; qlens[j] = s->seq[off + j].l_seq;
qseqs[j] = s->seq[off + j].seq; qseqs[j] = s->seq[off + j].seq;
} }
mm_map_multi(s->p->mi, s->n_seg[i], qlens, qseqs, &s->n_reg[off], &s->reg[off], b, s->p->opt, s->seq[off].name); mm_map_frag(s->p->mi, s->n_seg[i], qlens, qseqs, &s->n_reg[off], &s->reg[off], b, s->p->opt, s->seq[off].name);
for (j = 0; j < s->n_seg[i]; ++j) // flip the query strand and coordinate to the original read strand for (j = 0; j < s->n_seg[i]; ++j) // flip the query strand and coordinate to the original read strand
if (s->n_seg[i] == 2 && ((j == 0 && (pe_ori>>1&1)) || (j == 1 && (pe_ori&1)))) { if (s->n_seg[i] == 2 && ((j == 0 && (pe_ori>>1&1)) || (j == 1 && (pe_ori&1)))) {
int k, t; int k, t;
@ -414,11 +414,11 @@ static void *worker_pipeline(void *shared, int step, void *in)
pipeline_t *p = (pipeline_t*)shared; pipeline_t *p = (pipeline_t*)shared;
if (step == 0) { // step 0: read sequences if (step == 0) { // step 0: read sequences
int with_qual = (!!(p->opt->flag & MM_F_OUT_SAM) && !(p->opt->flag & MM_F_NO_QUAL)); int with_qual = (!!(p->opt->flag & MM_F_OUT_SAM) && !(p->opt->flag & MM_F_NO_QUAL));
int multi_seg = (p->n_fp > 1 || !!(p->opt->flag & MM_F_MULTI_SEG)); int frag_mode = (p->n_fp > 1 || !!(p->opt->flag & MM_F_FRAG_MODE));
step_t *s; step_t *s;
s = (step_t*)calloc(1, sizeof(step_t)); s = (step_t*)calloc(1, sizeof(step_t));
if (p->n_fp > 1) s->seq = mm_bseq_read_multi(p->n_fp, p->fp, p->mini_batch_size, with_qual, &s->n_seq); if (p->n_fp > 1) s->seq = mm_bseq_read_frag(p->n_fp, p->fp, p->mini_batch_size, with_qual, &s->n_seq);
else s->seq = mm_bseq_read2(p->fp[0], p->mini_batch_size, with_qual, multi_seg, &s->n_seq); else s->seq = mm_bseq_read2(p->fp[0], p->mini_batch_size, with_qual, frag_mode, &s->n_seq);
if (s->seq) { if (s->seq) {
s->p = p; s->p = p;
for (i = 0; i < s->n_seq; ++i) for (i = 0; i < s->n_seq; ++i)
@ -431,7 +431,7 @@ static void *worker_pipeline(void *shared, int step, void *in)
s->n_seg = s->seg_off + s->n_seq; s->n_seg = s->seg_off + s->n_seq;
s->reg = (mm_reg1_t**)calloc(s->n_seq, sizeof(mm_reg1_t*)); s->reg = (mm_reg1_t**)calloc(s->n_seq, sizeof(mm_reg1_t*));
for (i = 1, j = 0; i <= s->n_seq; ++i) for (i = 1, j = 0; i <= s->n_seq; ++i)
if (i == s->n_seq || !multi_seg || !mm_qname_same(s->seq[i-1].name, s->seq[i].name)) { if (i == s->n_seq || !frag_mode || !mm_qname_same(s->seq[i-1].name, s->seq[i].name)) {
s->n_seg[s->n_frag] = i - j; s->n_seg[s->n_frag] = i - j;
s->seg_off[s->n_frag++] = j; s->seg_off[s->n_frag++] = j;
j = i; j = i;
@ -484,7 +484,7 @@ static void *worker_pipeline(void *shared, int step, void *in)
return 0; return 0;
} }
int mm_map_file_multi_seg(const mm_idx_t *idx, int n_segs, const char **fn, const mm_mapopt_t *opt, int n_threads) int mm_map_file_frag(const mm_idx_t *idx, int n_segs, const char **fn, const mm_mapopt_t *opt, int n_threads)
{ {
int i, j; int i, j;
pipeline_t pl; pipeline_t pl;
@ -516,5 +516,5 @@ int mm_map_file_multi_seg(const mm_idx_t *idx, int n_segs, const char **fn, cons
int mm_map_file(const mm_idx_t *idx, const char *fn, const mm_mapopt_t *opt, int n_threads) int mm_map_file(const mm_idx_t *idx, const char *fn, const mm_mapopt_t *opt, int n_threads)
{ {
return mm_map_file_multi_seg(idx, 1, &fn, opt, n_threads); return mm_map_file_frag(idx, 1, &fn, opt, n_threads);
} }

4
minimap.h
View File

@ -18,7 +18,7 @@
#define MM_F_NO_LJOIN 0x400 #define MM_F_NO_LJOIN 0x400
#define MM_F_OUT_CS_LONG 0x800 #define MM_F_OUT_CS_LONG 0x800
#define MM_F_SR 0x1000 #define MM_F_SR 0x1000
#define MM_F_MULTI_SEG 0x2000 #define MM_F_FRAG_MODE 0x2000
#define MM_F_NO_PRINT_2ND 0x4000 #define MM_F_NO_PRINT_2ND 0x4000
#define MM_IDX_MAGIC "MMI\2" #define MM_IDX_MAGIC "MMI\2"
@ -258,7 +258,7 @@ mm_reg1_t *mm_map(const mm_idx_t *mi, int l_seq, const char *seq, int *n_regs, m
*/ */
int mm_map_file(const mm_idx_t *idx, const char *fn, const mm_mapopt_t *opt, int n_threads); int mm_map_file(const mm_idx_t *idx, const char *fn, const mm_mapopt_t *opt, int n_threads);
int mm_map_file_multi_seg(const mm_idx_t *idx, int n_segs, const char **fn, const mm_mapopt_t *opt, int n_threads); int mm_map_file_frag(const mm_idx_t *idx, int n_segs, const char **fn, const mm_mapopt_t *opt, int n_threads);
// deprecated APIs for backward compatibility // deprecated APIs for backward compatibility
void mm_mapopt_init(mm_mapopt_t *opt); void mm_mapopt_init(mm_mapopt_t *opt);