xxay/hifiasm
1
0
Fork 0
mirror of https://github.com/chhylp123/hifiasm.git synced 2026-09-24 18:08:12 +08:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #445 from chhylp123/hifiasm_dev_debug

Browse Source 
Hifiasm dev debug
This commit is contained in:
chhylp123
2023-04-14 16:44:10 -04:00
committed by GitHub
parent b49b4a6cc7 0efcc14499
commit fbfcf72eb3
6 changed files with 1026 additions and 118 deletions
Download Patch File Download Diff File Expand all files Collapse all files
CommandLines.h
+1 -1
View File
@@ -5,7 +5,7 @@
#include <pthread.h>
#include <stdint.h>
#define HA_VERSION "0.19.4-r575"
#define HA_VERSION "0.19.4-r584"
#define VERBOSE 0
Overlaps.cpp
+1003 -104
View File
File diff suppressed because it is too large Load Diff
Overlaps.h
+2 -2
View File
@@ -1066,7 +1066,7 @@ ma_ug_t* copy_untig_graph(ma_ug_t *src);
ma_ug_t* output_trio_unitig_graph(asg_t *sg, ma_sub_t* coverage_cut, char* output_file_name,
uint8_t flag, ma_hit_t_alloc* sources, ma_hit_t_alloc* reverse_sources,
long long tipsLen, float tip_drop_ratio, long long stops_threshold, R_to_U* ruIndex,
float chimeric_rate, float drop_ratio, int max_hang, int min_ovlp, int is_bench, bub_label_t* b_mask_t, char *f_prefix, uint8_t *kpt_buf, kvec_asg_arc_t_warp *r_edges);
float chimeric_rate, float drop_ratio, int max_hang, int min_ovlp, int gap_fuzz, int is_bench, bub_label_t* b_mask_t, char *f_prefix, uint8_t *kpt_buf, kvec_asg_arc_t_warp *r_edges);
asg_t* copy_read_graph(asg_t *src);
ma_ug_t *ma_ug_gen(asg_t *g);
void ma_ug_destroy(ma_ug_t *ug);
@@ -1135,7 +1135,7 @@ void adjust_utg_by_trio(ma_ug_t **ug, asg_t* read_g, uint8_t flag, float drop_ra
ma_hit_t_alloc* sources, ma_hit_t_alloc* reverse_sources, ma_sub_t* coverage_cut,
long long tipsLen, float tip_drop_ratio, long long stops_threshold,
R_to_U* ruIndex, float chimeric_rate, float drop_ratio, int max_hang, int min_ovlp,
kvec_asg_arc_t_warp* new_rtg_edges, bub_label_t* b_mask_t);
int gap_fuzz, kvec_asg_arc_t_warp* new_rtg_edges, bub_label_t* b_mask_t);
uint32_t cmp_untig_graph(ma_ug_t *src, ma_ug_t *dest);
void reduce_hamming_error(asg_t *sg, ma_hit_t_alloc* sources, ma_sub_t *coverage_cut,
int max_hang, int min_ovlp, long long gap_fuzz);
README.md
+1 -1
View File
@@ -51,7 +51,7 @@ Its latest release could support the telomere-to-telomere assembly by utilizing
## <a name="why"></a>Why Hifiasm?
* Hifiasm delivers high-quality assemblies. It tends to generate longer contigs
* Hifiasm delivers high-quality telomere-to-telomere assemblies. It tends to generate longer contigs
and resolve more segmental duplications than other assemblers.
* Given Hi-C reads or short reads from the parents, hifiasm can produce overall the best
gfa_ut.cpp
+18 -9
View File
@@ -7365,7 +7365,7 @@ void rebuid_idx(ul_resolve_t *uidx)
init_ul_str_idx_t(uidx);
}
void shrink_1b(ma_ug_t *ug, uc_block_t *z, uint32_t is_forward)
void shrink_1b(ma_ug_t *ug, uc_block_t *z, uc_block_t *lim, uint32_t is_forward)
{
if(z->ts != 0 || z->te != ug->g->seq[z->hid].len) return;
uc_block_t bc = *z;
@@ -7377,12 +7377,14 @@ void shrink_1b(ma_ug_t *ug, uc_block_t *z, uint32_t is_forward)
} else {
z->te -= 1; z->qs += off;
}
if((lim) && (!((z->qs <= lim->qs) && (z->qe <= lim->qe)))) *z = bc;
} else {
if((!z->rev)) {
z->te -= 1; z->qe -= off;
} else {
z->ts += 1; z->qe -= off;
}
if((lim) && (!((z->qs >= lim->qs) && (z->qe >= lim->qe)))) *z = bc;
}
if((ugl_cover_check(bc.ts, bc.te, &(ug->u.a[bc.hid]))) &&
(!ugl_cover_check(z->ts, z->te, &(ug->u.a[z->hid])))) {
@@ -7436,7 +7438,7 @@ void renew_ul_vec_t(ul_vec_t *x, ma_ug_t *ug)
void shrink_ul0(all_ul_t *uls, ul_str_t *str, uint64_t id, integer_t *buf, ma_ug_t *ug)
{
uint32_t k, c_k, p_k, cv, pv, bl, i; uc_block_t *xi; buf->u.n = 0; nid_t *np = NULL;
uint32_t k, c_k, p_k, cv, pv, bl, i; uc_block_t *xi, *yi; buf->u.n = 0; nid_t *np = NULL;
asg_arc_t *av; uint32_t nv, s, e, m, d, mm, is_conn; uint64_t *z; ul_vec_t *x;
if(str->cn < 2) return;
for (k = 0, bl = 0, c_k = p_k = pv = (uint32_t)-1; k < str->cn; k++) {
@@ -7464,6 +7466,11 @@ void shrink_ul0(all_ul_t *uls, ul_str_t *str, uint64_t id, integer_t *buf, ma_ug
is_conn = 1;
}
}
if(is_conn) {
is_conn = 0; assert(k);
yi = &(uls->a[id].bb.a[str->a[k-1]>>32]);
if((xi->qs >= yi->qs) && (xi->qe >= yi->qe)) is_conn = 1;
}
if(is_conn) {
bl++;
} else {
@@ -7514,10 +7521,9 @@ void shrink_ul0(all_ul_t *uls, ul_str_t *str, uint64_t id, integer_t *buf, ma_ug
x->bb.n = m;
assert(x->bb.n > 1);
shrink_1b(ug, &(x->bb.a[0]), 1);
shrink_1b(ug, &(x->bb.a[x->bb.n-1]), 0);
shrink_1b(ug, &(x->bb.a[0]), ((x->bb.n>=2)?&(x->bb.a[1]):(NULL)), 1);
shrink_1b(ug, &(x->bb.a[x->bb.n-1]), ((x->bb.n>=2)?&(x->bb.a[x->bb.n-2]):(NULL)), 0);
d = x->bb.a[0].qs;
for (k = 0; k < x->bb.n; k++) {
x->bb.a[k].qs -= d; x->bb.a[k].qe -= d;
@@ -7546,8 +7552,8 @@ void shrink_ul0(all_ul_t *uls, ul_str_t *str, uint64_t id, integer_t *buf, ma_ug
}
x->bb.n = m;
assert(x->bb.n > 1);
shrink_1b(ug, &(x->bb.a[0]), 1);
shrink_1b(ug, &(x->bb.a[x->bb.n-1]), 0);
shrink_1b(ug, &(x->bb.a[0]), ((x->bb.n>=2)?&(x->bb.a[1]):(NULL)), 1);
shrink_1b(ug, &(x->bb.a[x->bb.n-1]), ((x->bb.n>=2)?&(x->bb.a[x->bb.n-2]):(NULL)), 0);
d = x->bb.a[0].qs;
for (k = 0; k < x->bb.n; k++) {
x->bb.a[k].qs -= d; x->bb.a[k].qe -= d;
@@ -15686,6 +15692,9 @@ void u2g_hybrid_clean(ul_resolve_t *uidx, ulg_opt_t *ulopt, usg_t *ng, asg64_v *
// prt_usg_t(uidx, ng, sb);
// usg_arc_cut_length(ng, b, ub, mm_tip>>1, drop, ulopt->is_trio, 1, NULL);
usg_bub_clean(ng, &bb, b, ub, mm_tip>>1, drop, 1, bs, f);
// fprintf(stderr, "-1bub-[M::%s::] i::%ld, drop::%f\n", __func__, i, drop);
// sprintf(sb, "ng_ss::%ld_i::%ld_drop::%f_b::bub", ss, i, drop);
// prt_usg_t(uidx, ng, sb);
usg_arc_cut_srt_length(ng, b, ub, mm_tip>>1, drop, ulopt->is_trio, 1, NULL, bs);
// fprintf(stderr, "-1-[M::%s::] i::%ld, drop::%f\n", __func__, i, drop);
// sprintf(sb, "ng_ss::%ld_i::%ld_drop::%f_b", ss, i, drop);
@@ -16580,7 +16589,7 @@ ma_ug_t* output_trio_unitig_graph_ul(ug_opt_t *uopt, ul_resolve_t *uidx, char* o
adjust_utg_by_trio(&ug, uidx->sg, flag, TRIO_THRES, uopt->sources, uopt->reverse_sources,
uopt->coverage_cut, uopt->tipsLen, uopt->tip_drop_ratio, uopt->stops_threshold, uopt->ruIndex,
uopt->chimeric_rate, uopt->drop_ratio, uopt->max_hang, uopt->min_ovlp, &ne, uopt->b_mask_t);
uopt->chimeric_rate, uopt->drop_ratio, uopt->max_hang, uopt->min_ovlp, uopt->gap_fuzz, &ne, uopt->b_mask_t);
// if(asm_opt.b_low_cov > 0) {
// break_ug_contig(&ug, uidx->sg, &R_INF, uopt->coverage_cut, uopt->sources, uopt->ruIndex, &ne,
horder.cpp
+1 -1
View File
@@ -781,7 +781,7 @@ ma_ug_t* get_trio_unitig_graph(asg_t *sg, uint8_t flag, ug_opt_t *opt)
adjust_utg_by_trio(&ug, sg, flag, TRIO_THRES, opt->sources, opt->reverse_sources,
opt->coverage_cut, opt->tipsLen, opt->tip_drop_ratio, opt->stops_threshold,
opt->ruIndex, opt->chimeric_rate, opt->drop_ratio, opt->max_hang, opt->min_ovlp,
&new_rtg_edges, opt->b_mask_t);
opt->gap_fuzz, &new_rtg_edges, opt->b_mask_t);
kv_destroy(new_rtg_edges.a);
return ug;