-
Notifications
You must be signed in to change notification settings - Fork 30
/
Copy pathwtcorr.c
1539 lines (1487 loc) · 53.3 KB
/
wtcorr.c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
/*
*
* Copyright (c) 2011, Jue Ruan <ruanjue@gmail.com>
*
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
#include "list.h"
#include "string.h"
#include "dna.h"
#include "bitvec.h"
#include "hashset.h"
#include "counting_bloom_filter.h"
#include "heap.h"
#include "kswx.h"
#include "file_reader.h"
#include "thread.h"
#define DBG_MAX_KSIZE 31
#define DBG_MAX_BF_KSIZE 128
#define DBG_MAX_KLINK 0xFF
static int verbose = 0;
static char *debug_opt1 = NULL;
static char *debug_opt2 = NULL;
typedef struct {
uint64_t val;
uint8_t links[4+4];
} kmer_t;
#define dbg_kmer_smear(K) ((K) ^ ((K) >> 4) ^ ((K) >> 7) ^ ((K) >> 12))
#define dbg_kmer_hc(E) u64hashcode((E).val)
#define dbg_kmer_he(E1, E2) (E1).val == (E2).val
define_hashset(kmerhash, kmer_t, dbg_kmer_hc, dbg_kmer_he);
typedef struct {
kmerhash **hashs;
CBF **bfs; // max count is 3
uint32_t ksize, bf_ksize, nh; // nh = n_hash
uint64_t kmask;
uint8_t min_link_cov;
} DBG;
size_t _dbg_n_hash_count(void *obj, int idx){ if(idx == 0) return ((DBG*)obj)->nh; else if(idx == 1) return ((DBG*)obj)->nh; else return 1; }
static const struct obj_desc_t DBG_obj_desc = (obj_desc_t){sizeof(DBG), 2, {3, 3}, {offsetof(DBG, hashs), offsetof(DBG, bfs)}, {&kmerhash_obj_desc, &cbf_obj_desc}, _dbg_n_hash_count};
DBG* init_dbg(uint32_t ksize, uint32_t bf_ksize, uint32_t ncpu, uint32_t min_link_cov){
DBG *g;
uint32_t i;
if(ncpu < 1) ncpu = 1;
g = calloc(1, sizeof(DBG));
g->nh = ncpu;
g->hashs = malloc(sizeof(kmerhash*) * ncpu);
for(i=0;i<ncpu;i++) g->hashs[i] = init_kmerhash(1023);
g->bfs = NULL;
if((ksize & 0x01) == 0){
fprintf(stderr, " -- Automaticly change kmer_size from %u to %u in %s -- %s:%d --\n", ksize, ksize + 1, __FUNCTION__, __FILE__, __LINE__);
ksize ++;
}
if(ksize > DBG_MAX_KSIZE){
fprintf(stderr, " -- Automaticly change kmer_size from %u to %u in %s -- %s:%d --\n", ksize, DBG_MAX_KSIZE, __FUNCTION__, __FILE__, __LINE__);
ksize = DBG_MAX_KSIZE;
}
g->ksize = ksize;
if(bf_ksize){
if((bf_ksize & 0x01) == 0){
fprintf(stderr, " -- Automaticly change bf_kmer_size from %u to %u in %s -- %s:%d --\n", bf_ksize, bf_ksize + 1, __FUNCTION__, __FILE__, __LINE__);
bf_ksize ++;
}
if(bf_ksize > DBG_MAX_BF_KSIZE){
fprintf(stderr, " -- Automaticly change bf_kmer_size from %u to %u in %s -- %s:%d --\n", bf_ksize, DBG_MAX_BF_KSIZE - 1, __FUNCTION__, __FILE__, __LINE__);
bf_ksize = DBG_MAX_BF_KSIZE - 1;
}
if(bf_ksize < ksize){
fprintf(stderr, " -- bf_kmer_size (%u) is no larger than kmer_size (%u), skip to build bloom-filter in %s -- %s:%d --\n", bf_ksize, ksize, __FUNCTION__, __FILE__, __LINE__);
bf_ksize = 0;
}
}
g->bf_ksize = bf_ksize;
g->kmask = 0xFFFFFFFFFFFFFFFFLLU >> ((32 - ksize) << 1);
g->min_link_cov = min_link_cov;
return g;
}
void free_dbg(DBG *g){
uint32_t i;
for(i=0;i<g->nh;i++) free_kmerhash(g->hashs[i]);
free(g->hashs);
if(g->bfs){
for(i=0;i<g->nh;i++) free_cbf(g->bfs[i]);
free(g->bfs);
}
free(g);
}
// Thread: reading sequence
thread_beg_def(mseq);
BaseBank *rdseqs;
u32list *rdlens;
u8list *seqs;
uint64_t max;
FileReader *fr;
Sequence *seq;
thread_end_def(mseq);
thread_beg_func(mseq);
unsigned long long nseq, tseq, toff;
uint32_t nlen;
int i;
tseq = toff = 0;
mseq->seq = NULL;
thread_beg_loop(mseq);
clear_u8list(mseq->seqs);
nseq = 0;
if(mseq->fr){
while(fread_seq_adv((Sequence**)&mseq->seq, mseq->fr, SEQ_FLAG_NO_NAME | SEQ_FLAG_NO_QUAL)){
nseq ++;
push_u8list(mseq->seqs, 4);
for(i=0;i<mseq->seq->seq.size;i++) push_u8list(mseq->seqs, base_bit_table[(int)mseq->seq->seq.string[i]]);
if(mseq->rdseqs){
seq2basebank(mseq->rdseqs, mseq->seq->seq.string, mseq->seq->seq.size);
push_u32list(mseq->rdlens, mseq->seq->seq.size);
}
if(mseq->seqs->size >= mseq->max) break;
}
} else if(mseq->rdseqs){
while(tseq + nseq < mseq->rdlens->size){
push_u8list(mseq->seqs, 4);
nlen = mseq->rdlens->buffer[tseq + (nseq ++)];
encap_u8list(mseq->seqs, nlen);
bitseq_basebank(mseq->rdseqs, toff, nlen, mseq->seqs->buffer + mseq->seqs->size);
mseq->seqs->size += nlen;
toff += nlen;
if(mseq->seqs->size >= mseq->max) break;
}
}
if(mseq->seqs->size) push_u8list(mseq->seqs, 4);
tseq += nseq;
if(verbose){ fprintf(stderr, "[%s] + %llu = %llu reads.\n", date(), nseq, tseq); fflush(stderr); }
thread_end_loop(mseq);
if(mseq->seq) free_sequence(mseq->seq);
thread_end_func(mseq);
// Thread: hashing kmers
thread_beg_def(midx);
DBG *g;
u8list *seqs;
thread_end_def(midx);
thread_beg_func(midx);
kmerhash *hash;
kmer_t *k, KMER;
uint64_t i, j, kmer;
uint32_t tidx;
int exists;
uint8_t cur, lst, nxt;
tidx = midx->t_idx;
hash = midx->g->hashs[tidx];
memset(&KMER, 0, sizeof(kmer_t));
thread_beg_loop(midx);
kmer = 0;
for(i=j=0;i<midx->seqs->size;i++){
cur = midx->seqs->buffer[i];
if(cur == 4){ j = 0; kmer = 0; continue; }
kmer = ((kmer << 2) | cur) & midx->g->kmask;
j ++;
if(j < midx->g->ksize) continue;
lst = j == midx->g->ksize? 8 : 3 - midx->seqs->buffer[i - midx->g->ksize];
nxt = midx->seqs->buffer[i + 1];
if(nxt == 4) nxt = 8;
KMER.val = dna_rev_seq(kmer, midx->g->ksize);
if(KMER.val < kmer){
nxt = nxt == 8? 8 : 4 + nxt;
} else {
KMER.val = kmer;
lst = lst == 8? 8 : 4 + lst;
}
if((dbg_kmer_smear(KMER.val) % midx->g->nh) != tidx) continue;
k = prepare_kmerhash(hash, KMER, &exists);
if(!exists) *k = KMER;
if(lst != 8 && k->links[lst] < DBG_MAX_KLINK) k->links[lst] ++;
if(nxt != 8 && k->links[nxt] < DBG_MAX_KLINK) k->links[nxt] ++;
}
thread_end_loop(midx);
thread_end_func(midx);
// Thread: bf-hashing kmers
thread_beg_def(mbf);
DBG *g;
u8list *seqs;
thread_end_def(mbf);
thread_beg_func(mbf);
CBF *bf;
uint64_t i, j, *k, kmer[DBG_MAX_BF_KSIZE / 32], krev[DBG_MAX_BF_KSIZE / 32];
uint32_t tidx, bfn;
uint8_t cur;
tidx = mbf->t_idx;
bf = mbf->g->bfs[tidx];
bfn = (mbf->g->bf_ksize + 31) / 32 * 8;
thread_beg_loop(mbf);
memset(kmer, 0, bfn);
for(i=j=0;i<mbf->seqs->size;i++){
cur = mbf->seqs->buffer[i];
if(cur == 4){ j = 0; memset(kmer, 0, bfn); continue; }
dna_shl_seqs(kmer, mbf->g->bf_ksize, cur);
j ++;
if(j < mbf->g->ksize) continue;
memcpy(krev, kmer, bfn);
dna_rev_seqs(krev, mbf->g->bf_ksize);
if(dna_cmp_seqs(krev, kmer, mbf->g->bf_ksize) == -1){
k = krev;
} else {
k = kmer;
}
if((dbg_kmer_smear(k[0]) % mbf->g->nh) != tidx) continue;
put_cbf(bf, k, bfn);
}
thread_end_loop(mbf);
thread_end_func(mbf);
void build_dbg(DBG *g, uint64_t seq_batch_size, FileReader *fr){
BaseBank *rdseqs;
u32list *rdlens;
u8list *seqs[2];
kmer_t KMER;
size_t size, totk;
uint32_t i, seq_idx;
thread_preprocess(mseq);
thread_preprocess(midx);
thread_preprocess(mbf);
rdseqs = init_basebank();
rdlens = init_u32list(1024);
memset(&KMER, 0, sizeof(kmer_t));
seqs[0] = init_u8list(1024);
seqs[1] = init_u8list(1024);
seq_idx = 0;
thread_beg_init(mseq, 1);
mseq->rdseqs = g->bf_ksize? rdseqs : NULL;
mseq->rdlens = g->bf_ksize? rdlens : NULL;
mseq->seqs = seqs[seq_idx];
mseq->max = seq_batch_size;
mseq->fr = fr;
thread_end_init(mseq);
thread_beg_operate(mseq, 0);
thread_wake(mseq);
thread_beg_init(midx, g->nh);
midx->g = g;
midx->seqs = seqs[seq_idx];
thread_end_init(midx);
while(1){
thread_wait_all(midx);
thread_wait(mseq);
if(mseq->seqs->size == 0) break;
thread_beg_iter(midx);
midx->seqs = seqs[seq_idx];
thread_wake(midx);
thread_end_iter(midx);
seq_idx = !seq_idx;
mseq->seqs = seqs[seq_idx];
thread_wake(mseq);
}
thread_beg_close(midx);
thread_end_close(midx);
thread_beg_close(mseq);
thread_end_close(mseq);
if(g->bf_ksize == 0) return;
size = mem_size_obj(g->hashs, 3, &kmerhash_obj_desc, 0, g->nh);
for(i=totk=0;i<g->nh;i++) totk += g->hashs[i]->count;
if(verbose){
fprintf(stdout, "[%s] Total size of kmer hash is %llu bytes\n", date(), (unsigned long long)size); fflush(stdout);
fprintf(stdout, "[%s] Total number of kmers is %llu\n", date(), (unsigned long long)totk); fflush(stdout);
}
// bits = 1.44 * log2(1/FP), here we set FP = 0.001, bits ~ 14
// The total size of CBF is 14 * kmer_count, here we assume the number of small-kmers as number of big-kmers
size = 14 * (totk / g->nh);
size = _rj_hashset_find_prime(size);
if(verbose){
fprintf(stdout, "[%s] Total memory of counting bloom filter is %llu bytes\n", date(), (unsigned long long)size * g->nh / 4); fflush(stdout);
}
g->bfs = malloc(sizeof(CBF*) * g->nh);
for(i=0;i<g->nh;i++) g->bfs[i] = init_cbf(size, 2, 3);
thread_beg_init(mseq, 1);
mseq->rdseqs = g->bf_ksize? rdseqs : NULL;
mseq->rdlens = g->bf_ksize? rdlens : NULL;
mseq->seqs = seqs[seq_idx];
mseq->max = seq_batch_size;
mseq->fr = NULL;
thread_end_init(mseq);
thread_beg_operate(mseq, 0);
thread_wake(mseq);
thread_beg_init(mbf, g->nh);
mbf->g = g;
mbf->seqs = seqs[seq_idx];
thread_end_init(mbf);
while(1){
thread_wait_all(mbf);
thread_wait(mseq);
if(mseq->seqs->size == 0) break;
thread_beg_iter(mbf);
mbf->seqs = seqs[seq_idx];
thread_wake(mbf);
thread_end_iter(mbf);
seq_idx = !seq_idx;
mseq->seqs = seqs[seq_idx];
thread_wake(mseq);
}
thread_beg_close(mbf);
thread_end_close(mbf);
thread_beg_close(mseq);
thread_end_close(mseq);
free_u8list(seqs[0]);
free_u8list(seqs[1]);
free_basebank(rdseqs);
free_u32list(rdlens);
}
void chk_sequences_dbg(DBG *g, char *tag, char *seq, uint32_t seqlen, uint8_t min_link_cov, uint8_t min_bf_cov, FILE *out){
kmer_t *k, KMER;
uint64_t kmer, krev;
uint64_t *bk, bkmer[DBG_MAX_BF_KSIZE / 32], bkrev[DBG_MAX_BF_KSIZE / 32];
uint32_t i, j, bfn, hidx;
uint8_t cur, lst;
kmer = 0;
bfn = (g->bf_ksize + 31) / 32 * 8;
memset(&KMER, 0, sizeof(kmer_t));
if(bfn) memset(bkmer, 0, bfn);
for(i=j=0;i<=seqlen;i++){
if((cur = base_bit_table[(int)seq[i]]) == 4){
if(j > g->ksize) fprintf(out, "%s\t%d\t%d\t%d\n", tag, i - j, i - 1, j);
j = 0; kmer = 0; if(bfn) memset(bkmer, 0, bfn);
continue;
}
kmer = ((kmer << 2) | cur) & g->kmask;
dna_shl_seqs(bkmer, g->bf_ksize, cur);
j ++;
if(j < g->ksize) continue;
lst = (j == g->ksize)? 8 : 3 - base_bit_table[(int)seq[i - g->ksize]];
krev = dna_rev_seq(kmer, g->ksize);
if(krev > kmer){ krev = kmer; lst = lst == 8? 8 : 4 + lst; }
hidx = dbg_kmer_smear(krev) % g->nh;
KMER.val = krev;
if(((k = get_kmerhash(g->hashs[hidx], KMER)) == NULL || (lst != 8 && k->links[lst] < min_link_cov))){
if(j > g->ksize) fprintf(out, "%s\t%d\t%d\t%d\n", tag, i - j + 1, i - 1, j - 2);
j = 0; kmer = 0; if(bfn) memset(bkmer, 0, bfn);
continue;
}
if(bfn && j >= g->bf_ksize){
//uint32_t l;
//for(l=0;l<g->bf_ksize;l++) fputc(bit_base_table[bits2bit(bkmer, l)], stdout);
memcpy(bkrev, bkmer, bfn);
dna_rev_seqs(bkrev, g->bf_ksize);
if(dna_cmp_seqs(bkrev, bkmer, g->bf_ksize) == -1) bk = bkrev;
else bk = bkmer;
hidx = dbg_kmer_smear(bk[0]) % g->nh;
//fprintf(stdout, " -- %d in %s -- %s:%d --\n", get_cbf(g->bfs[hidx], bk, bfn), __FUNCTION__, __FILE__, __LINE__); fflush(stdout);
if(get_cbf(g->bfs[hidx], bk, bfn) < min_bf_cov){
if(j > g->ksize) fprintf(out, "%s\t%d\t%d\t%d\n", tag, i - j + 1, i - 1, j - 2);
j = 0; kmer = 0; memset(bkmer, 0, bfn);
}
}
}
}
typedef struct { uint32_t kidx:31, dir:1; } klnk_t;
typedef struct {
kmer_t *k;
klnk_t links[8]; // kidx = 0 : NULL
} kmer_cache_t;
define_list(kcachev, kmer_cache_t);
// need to call set_userdata before use kcachehash
#define E(idx) ((kcachev*)set->userdata)->buffer[idx]
#define kmer_cache_hc(idx) 64hashcode(E(idx).k->val)
#define kmer_cache_he(idx1, idx2) (E(idx1).k->val == E(idx2).k->val)
define_hashset(kcachehash, uint32_t, kmer_cache_hc, kmer_cache_he);
#define DBG_MAX_BTIDX 0x7FFFFFFFU
#define DBG_MAX_SEED 0x00000FFFU
typedef struct {
uint32_t qidx:12;
int score:20;
uint32_t link:1; // qdp is stacked togethter in qdpv, link == 1 means existing next linked qdp.
uint32_t bt_idx:31;
} q_dp_t;
define_list(qdpv, q_dp_t);
typedef struct {
union {
uint32_t kidx:28, path:2, qdir:1, dir:1;
uint32_t identifier;
};
union {
int score:21, inc:10;
uint32_t cached:1;
uint32_t scoreinfo;
};
uint32_t bt_idx;
} dbg_dp_t;
define_list(dbgdpv, dbg_dp_t);
int dbg_dp_heap_cmp(uint32_t idx1, uint32_t idx2, void *ref){
dbgdpv *dps = (dbgdpv*)ref;
dbg_dp_t *dp1 = ref_dbgdpv(dps, idx1);
dbg_dp_t *dp2 = ref_dbgdpv(dps, idx2);
cmp_2nums_proc(dp2->score, dp1->score);
return 0;
}
// set->userdata is bulit-in field of hashset
#define E(idx) ((gdpv*)set->userdata)->buffer[idx]
#define dbg_dp_hc(idx) (u64hashcode(E(idx).identifier))
#define dbg_dp_he(idx1, idx2) ((E(idx1)).identifier == (E(idx2)).identifier)
define_hashset(dphash, uint32_t, dbg_dp_hc, dbg_dp_he);
#undef E
//Speedup strategies:
//TODO: 1, kmer_cache can cache traveled kmers and their linked kmers
//TODO: 2, dphash can be divided by (qpos, path, qdir, dir), to reduce search time, aln->qtop - aln->W determinates the life-time of qpos's dphash
//TODO: 3, maintianing qtop, qmaxs(length eq qlen, max score for each pos) for all qmers. First start the 1th qmer, when its qtop come to the next qmer position, start the next qmer, and so on
//TODO: the dbg_dp_t are shared between all qmers, the max score dp will live. How to find multiple best local alignments? analysis the qmaxs! We need
//TODO: a complicated qmaxs, should be {int score; uint32_t dp_idx;}.
typedef struct {
uint32_t idx;
int score;
int x, y;
} dbg_hsp_t;
define_list(dbghspv, dbg_hsp_t);
typedef struct {
DBG *g;
uint8_t min_link_cov, link_cov_cutoff, kmer_cov_cutoff; // link_cov is for edges of small-kmer, kmer_cov is for big-kmer's coverage
float min_mat_ratio; // in any move, if total mat is less than this ratio of qry run, discard it
int lowcov_bonus; // penalty for path with cov < min_link_cov
float cov_bonus; // add coverage * cov_bonus to alignment score, usually set to FLT_MIN
int Z, W, M, X, I, D, E, T; // M > 0, Z is somehow like score bandwidth, W is like qpos bandwidth. T * (X or QMIS) is end-clip penalty
int QMAT, QMIS, QDEL, QEXT; // global mat/mis/del/gap_ext PhredQV for 5q alignment
int aln_mode; // 0: global; 1: local
int all_kmer; // whether to try all qmers
size_t max_nodes;
uint8_t *qry;
int qlen;
u8list **qvs;
int has_5q; // for quality values guided alignment
dbgdpv *qmers; // solid k-mer in qry
dphash *qhash; // hash of qmers
dbghspv *hsps; // local alignment for each qmer
kcachev *kcs; // cache searched kmer and links
kcachehash *kch;
dbgdpv *nodes;
u64list *bfks; // recording bf-kmers for each node
uint32_t bfk_n; // how many uint64_t for each record
Heap *heap; // sort living (to be extended) nodes by score
dphash *hash; // hash of nodes, avoid to search traved kmer-pos
int qbeg;
int qmax;
int local_max; // max score, like local alignment
uint32_t local_idx; // node idx of local_max
//int qtop; // The longest move on query
//int smin; // smin is the low bound of dp_t->score
//int last_cached_pos;
dbgdpv *paths; // best paths of extension for each solid k-mer, the first element will be used for querying dphash
dphash *phash; // hash of best paths
int max_score;
uint32_t best_idx;
} dbg_aligner;
dbg_aligner* init_dbgaln(DBG *g, int Z, int W, int T, int M, int X, int I, int D, int E, int QMAT, int QMIS, int QDEL, int QEXT){
//uint32_t i;
dbg_aligner *aln;
aln = malloc(sizeof(dbg_aligner));
aln->g = g;
aln->min_mat_ratio = 0.5;
aln->min_link_cov = g->min_link_cov;
aln->link_cov_cutoff = 0;
aln->kmer_cov_cutoff = 0;
aln->lowcov_bonus = -2;
aln->cov_bonus = FLT_MIN;
aln->Z = Z;
aln->W = W;
aln->T = T;
aln->M = M;
aln->X = X;
aln->I = I;
aln->D = D;
aln->E = E;
aln->QMAT = QMAT;
aln->QMIS = QMIS;
aln->QDEL = QDEL;
aln->QEXT = QEXT;
aln->aln_mode = 1;
aln->all_kmer = 0;
aln->max_nodes = 0xFFFFFFFFFFFFFFFFLLU;
aln->qry = NULL;
aln->qlen = 0;
aln->qvs = NULL;
aln->has_5q = 0;
aln->qmers = init_dbgdpv(1024);
aln->qhash = init_dphash(1023);
set_userdata_dphash(aln->qhash, aln->qmers);
aln->kcs = init_kcachev(1024);
aln->kch = init_kcachehash(1023);
aln->hsps = init_dbghspv(1024);
aln->nodes = init_dbgdpv(1024);
aln->bfks = init_u64list(1024);
aln->bfk_n = g->bf_ksize? (g->bf_ksize + 31) / 32 : 0;
aln->heap = init_heap(dbg_dp_heap_cmp, aln->nodes);
aln->hash = init_dphash(1023);
set_userdata_dphash(aln->hash, aln->nodes);
aln->qmax = - 0x7FFFFFFF;
aln->qbeg = 0;
aln->local_max = 0;
aln->local_idx = 1;
aln->paths = init_dbgdpv(1024);
aln->phash = init2_dphash(1023, 0.33);
set_userdata_dphash(aln->phash, aln->paths);
aln->max_score = -0x0FFFFFFF;
aln->best_idx = 0;
return aln;
}
void free_dbgaln(dbg_aligner *aln){
free_kcachev(aln->kcs);
free_kcachehash(aln->kch);
free_dbgdpv(aln->qmers);
free_dphash(aln->qhash);
free_dbghspv(aln->hsps);
free_dbgdpv(aln->nodes);
free_u64list(aln->bfks);
free_heap(aln->heap);
free_dphash(aln->hash);
free_dbgdpv(aln->paths);
free_dphash(aln->phash);
free(aln);
}
inline uint8_t last_base_dp_kmer(dbg_dp_t *d, uint32_t ksize){
return d->dir? ((~(d->k->val >> ((ksize - 1) << 1))) & 0x03) : (d->k->val & 0x03);
}
void trace_aln_paths(dbg_aligner *aln, uint32_t idx, uint32_t limit, FILE *out){
dbg_dp_t *d;
uint32_t cnt;
cnt = 0;
while(idx){
cnt ++;
if(cnt == limit) break;
d = ref_dbgdpv(aln->paths, idx);
fprintf(out, "%d[%d,%d,%d,%p];\n", idx, d->bt_idx, d->qpos, d->path, d->k);
idx = d->bt_idx;
}
fflush(out);
}
int count_covered_qmers(dbg_aligner *aln, uint32_t sidx){
dbg_dp_t *d;
int cnt;
cnt = 0;
while(sidx){
d = ref_dbgdpv(aln->nodes, sidx);
sidx = d->bt_idx;
if(d->path != 0) continue;
aln->qmers->buffer[aln->qmers->size] = *d;
d = ref_dbgdpv(aln->qmers, aln->qmers->size);
if(d->qdir){
d->qdir = 0;
d->dir = d->dir ^ 1;
d->qpos = aln->qlen - (d->qpos - aln->g->ksize);
}
if(exists_dphash(aln->qhash, aln->qmers->size)) cnt ++;
}
return cnt;
}
typedef struct {
int score;
int qb, qe;
int alns[4];
} dbg_hit_t;
dbg_hit_t stat_aln_paths(dbg_aligner *aln, uint32_t idx){
dbg_hit_t R;
dbg_dp_t *d;
memset(&R, 0, sizeof(dbg_hit_t));
while(idx){
d = ref_dbgdpv(aln->nodes, idx);
R.alns[d->path] ++;
R.score += d->inc;
idx = d->bt_idx;
}
return R;
}
static const int dbg_path2cigar[] = {0, 2, 1, 0};
dbg_hit_t call_correct_seq(dbg_aligner *aln, uint32_t qmer_idx, u8list *seq, u32list *cigars){
dbg_hit_t R;
dbg_dp_t *d;
uint64_t kmer;
uint32_t *ptr, idx, i;
uint8_t v;
if(seq) clear_u8list(seq);
if(cigars) clear_u32list(cigars);
d = ref_dbgdpv(aln->qmers, qmer_idx);
set_dbgdpv(aln->paths, 0, *d);
aln->paths->buffer[0].qdir = 1; // first pass of the reverse strand
aln->paths->buffer[0].dir = d->dir ^ 1;
aln->paths->buffer[0].qpos = aln->qlen - (d->qpos - aln->g->ksize);
ptr = get_dphash(aln->phash, 0);
if(ptr == NULL){
fprintf(stderr, " -- Unexpected error in %s -- %s:%d --\n", __FUNCTION__, __FILE__, __LINE__); exit(1);
}
idx = *ptr;
memset(&R, 0, sizeof(dbg_hit_t));
while(idx){
d = ref_dbgdpv(aln->paths, idx);
if(seq && d->path != 1){
// v = dir? ((~(d->k->val >> ((aln->g->ksize - 1) << 1))) & 0x03) : (d->k->val & 0x03);
// We want (~v) & 0x03, thre complementary
v = d->dir? (d->k->val >> ((aln->g->ksize - 1) << 1)) : ((~d->k->val) & 0x03);
if(d->path) v += 4;
push_u8list(seq, v);
}
if(cigars) kswx_push_cigar(cigars, dbg_path2cigar[d->path], 1);
idx = d->bt_idx;
R.alns[d->path] ++;
}
R.qb = aln->qlen - d->qpos;
if(seq) reverse_u8list(seq);
if(cigars) reverse_u32list(cigars);
d = ref_dbgdpv(aln->qmers, qmer_idx);
set_dbgdpv(aln->paths, 0, *d);
aln->paths->buffer[0].qdir = 0; // the forward strand
aln->paths->buffer[0].dir = d->dir ^ 0;
aln->paths->buffer[0].qpos = d->qpos;
ptr = get_dphash(aln->phash, 0);
if(ptr == NULL){
fprintf(stderr, " -- Unexpected error in %s -- %s:%d --\n", __FUNCTION__, __FILE__, __LINE__); exit(1);
}
idx = *ptr;
if(seq){
d = ref_dbgdpv(aln->paths, idx);
kmer = d->dir? dna_rev_seq(d->k->val, aln->g->ksize) : d->k->val;
//fputc('\n', stdout);
for(i=1;i+1<aln->g->ksize;i++){
v = (kmer >> ((aln->g->ksize - 1 - i) << 1)) & 0x03;
push_u8list(seq, v);
//fputc(bit_base_table[v], stdout);
}
//fputc('\n', stdout);
}
if(cigars) kswx_push_cigar(cigars, 0, aln->g->ksize - 2);
R.alns[0] += aln->g->ksize - 2;
while(idx){
d = ref_dbgdpv(aln->paths, idx);
if(seq && d->path != 1){
v = last_base_dp_kmer(d, aln->g->ksize);
if(d->path) v += 4;
push_u8list(seq, v);
}
if(cigars) kswx_push_cigar(cigars, dbg_path2cigar[d->path], 1);
idx = d->bt_idx;
R.alns[d->path] ++;
}
R.qe = d->qpos;
return R;
}
inline void init_bf_kmer_dbgaln(dbg_aligner *aln, dbg_dp_t *d){
uint64_t *k, kmer;
uint32_t i;
encap_u64list(aln->bfks, aln->bfk_n);
kmer = d->dir? dna_rev_seq(d->k->val, aln->g->ksize) : d->k->val;
k = aln->bfks->buffer + (d - aln->nodes->buffer) * aln->bfk_n;
memset(k, 0, aln->bfk_n * sizeof(uint64_t));
for(i=0;i<aln->g->ksize;i++) dna_shl_seqs(k, aln->g->bf_ksize, (kmer >> ((aln->g->ksize - 1 - i) << 1)) & 0x03U);
}
inline int prepare_bf_kmer_dbgaln(dbg_aligner *aln, uint32_t dpidx, uint8_t link){
uint64_t *k1, *k2, *k3;
uint32_t hidx;
encap_u64list(aln->bfks, 2 * aln->bfk_n);
k1 = aln->bfks->buffer + ((size_t)dpidx) * aln->bfk_n;
k2 = aln->bfks->buffer + aln->nodes->size * aln->bfk_n;
k3 = aln->bfks->buffer + (aln->nodes->size + 1) * aln->bfk_n;
memcpy(k2, k1, aln->bfk_n * sizeof(uint64_t));
dna_shl_seqs(k2, aln->g->bf_ksize, link);
memcpy(k3, k2, aln->bfk_n * sizeof(uint64_t));
dna_rev_seqs(k3, aln->g->bf_ksize);
if(dna_cmp_seqs(k2, k3, aln->g->bf_ksize) < 0) k3 = k2;
hidx = dbg_kmer_smear(k3[0]) % aln->g->nh;
return get_cbf(aln->g->bfs[hidx], k3, aln->bfk_n * sizeof(uint64_t));
}
inline void cpy_bf_kmer_dbgaln(dbg_aligner *aln, uint32_t dpidx){
uint64_t *k1, *k2;
encap_u64list(aln->bfks, aln->bfk_n);
k1 = aln->bfks->buffer + ((size_t)dpidx) * aln->bfk_n;
k2 = aln->bfks->buffer + (aln->nodes->size) * aln->bfk_n;
memcpy(k2, k1, aln->bfk_n * sizeof(uint64_t));
}
inline void print_dp_kmers(dbg_aligner *aln, uint32_t dpidx){
dbg_dp_t *d;
uint64_t kmer, *k;
uint32_t i;
d = ref_dbgdpv(aln->nodes, dpidx);
kmer = d->dir? dna_rev_seq(d->k->val, aln->g->ksize) : d->k->val;
for(i=0;i<aln->g->ksize;i++) fputc(bit_base_table[(kmer >> ((aln->g->ksize - 1 - i) << 1)) & 0x03], stdout);
fputc('\t', stdout);
k = aln->bfks->buffer + dpidx * aln->bfk_n;
for(i=0;i<aln->g->bf_ksize;i++) fputc(bit_base_table[bits2bit(k, i)], stdout);
}
static const int hashing_dp_path[2][4] = {{1, 1, 1, 1}, {1, 0, 0, 0}};
#define DBG_DP_CORE_FINISH 0
#define DBG_DP_CORE_EXTEND 1
#define DBG_DP_CORE_DISCARD 2
#define DBG_DP_CORE_DEADEND 3
#define DBG_DP_CORE_FOUND 4
#define REMOVE_HEAP(aln, idx) array_heap_remove((aln)->heap->ptrs->buffer, (aln)->heap->ptrs->size, (aln)->heap->ptrs->cap, uint32_t, idx, num_cmp((aln)->nodes->buffer[b].score, (aln)->nodes->buffer[a].score))
#define PUSH_HEAP(aln, idx) array_heap_push((aln)->heap->ptrs->buffer, (aln)->heap->ptrs->size, (aln)->heap->ptrs->cap, uint32_t, idx, num_cmp((aln)->nodes->buffer[b].score, (aln)->nodes->buffer[a].score))
// TODO: Supports we know some or all of variants in population or especially in multiple copies of repetitive sequences,
// TODO: need to lower the score penalty for known variants. Finding bubbles on DBG? will help reduce false alignment step
// TODO: No, if the right path exists on DBG, we only need to care about sequencing errors
inline int dbg_aln_core(dbg_aligner *aln){
dbg_dp_t *dp, *dp2, *dp1;
kmer_cache_t *kc;
kmer_t *k, KMER;
uint64_t kmer, knew, *bk, BK[DBG_MAX_BF_KSIZE / 32], BR[DBG_MAX_BF_KSIZE / 32];
uint32_t idx, i, d, q, hidx, cnt, *ptr, didx, pidx, path, _qpos;
uint8_t lst_base;
int score, exists, kc_exists, found, qpos;
if((idx = peer_heap(aln->heap)) == 0xFFFFFFFFU) return DBG_DP_CORE_FINISH;
encap_dbgdpv(aln->nodes, 9);
dp = ref_dbgdpv(aln->nodes, idx);
if(verbose > 3){
printf("DP: %u<-%u\tpos=%d\tpath=%d\tscore=%d\theap=%d\tsearch=%d\tqmax=%d", idx, dp->bt_idx, dp->qpos, dp->path, (int)dp->score, (int)aln->heap->ptrs->size, (int)aln->nodes->size, aln->qmaxs->buffer[dp->qpos]);
if(verbose > 4){
if(verbose > 5){
dbg_hit_t R = stat_aln_paths(aln, idx);
printf("\t[%d,%d,%d,%d,%d]", R.score, R.alns[0], R.alns[3], R.alns[1], R.alns[2]);
}
printf("\t");
print_dp_kmers(aln, idx);
}
printf("\n");
}
if(dp->score > aln->local_max){
aln->local_max = dp->score;
aln->local_idx = idx;
}
if(dp->qpos >= aln->qlen){
return DBG_DP_CORE_FOUND;
}
{
if(dp->score < aln->smin){
REMOVE_HEAP(aln, 0); return DBG_DP_CORE_DISCARD;
}
//if(dp->mat < (dp->qpos - aln->qbeg) * aln->min_mat_ratio){
//REMOVE_HEAP(aln, 0); return DBG_DP_CORE_DISCARD;
//}
if(dp->qpos + aln->W < aln->qtop){ // W bandwidth
REMOVE_HEAP(aln, 0); return DBG_DP_CORE_DISCARD;
} else if(dp->qpos > aln->qtop){
aln->qtop = dp->qpos;
// Tidy heap to improve the speed of heap_push
for(i=aln->heap->ptrs->size-1;i>0;i--){
if(aln->nodes->buffer[aln->heap->ptrs->buffer[i]].qpos + aln->W < aln->qtop){
REMOVE_HEAP(aln, i);
}
}
}
if(dp->score < aln->qmaxs->buffer[dp->qpos]){ // Z bandwidth
REMOVE_HEAP(aln, 0); return DBG_DP_CORE_DISCARD;
} else if(dp->score + aln->Z * (aln->has_5q? aln->QMIS : aln->X) > aln->qmaxs->buffer[dp->qpos]){
aln->qmaxs->buffer[dp->qpos] = dp->score + aln->Z * (aln->has_5q? aln->QMIS : aln->X);
}
}
if(1){
ptr = prepare_dphash(aln->hash, idx, &exists);
if(exists){ // Found cached search
dp2 = ref_dbgdpv(aln->nodes, *ptr);
if(dp2->cached){
return process_cached_dps_dbgaln(aln, idx, *ptr);
} else if(dp->score > dp2->score){
*ptr = idx;
} else{
REMOVE_HEAP(aln, 0); return DBG_DP_CORE_DISCARD;
}
} else {
*ptr = idx;
}
}
// extending
REMOVE_HEAP(aln, 0);
cnt = 0;
for(i=0;i<4;i++){
if(dp->k->links[dp->dir * 4 + i] < aln->link_cov_cutoff) continue;
cnt ++;
}
if(cnt == 0) return DBG_DP_CORE_DEADEND;
dp->fw_idx = aln->nodes->size;
kmer = dp->dir? dna_rev_seq(dp->k->val, aln->g->ksize) : dp->k->val;
lst_base = kmer & 0x03;
if(dp->qdir){
_qpos = aln->qlen - 1 - dp->qpos;
q = (~aln->qry[_qpos]) & 0x03;
} else {
_qpos = dp->qpos;
q = aln->qry[_qpos];
}
memset(&KMER, 0, sizeof(kmer_t));
cnt = 0;
dc = prepare_dbgcachehash(aln->gcache, (dbg_cache_t){dp->k, (kmer_t*[8]){0, 0, 0, 0, 0, 0, 0, 0}}, &dc_exists);
if(!dc_exists) memset(dc->links, 0, sizeof(kmer_t*) * 8);
for(i=0;i<4;i++){
if(dp->k->links[dp->dir * 4 + i] < aln->link_cov_cutoff) continue;
if(aln->bfk_n && (prepare_bf_kmer_dbgaln(aln, idx, i) < aln->kmer_cov_cutoff && dp->qpos - aln->qbeg >= (int)aln->g->bf_ksize)) continue; // Check whether this move exists in big-kmer
// match or mismatch
if(aln->has_5q){
if(i == q) score = dp->score + aln->QMAT;
else if(dp->qdir){
score = dp->score + ((aln->qvs[5]->buffer[_qpos] < 4 && i == ((~aln->qvs[5]->buffer[_qpos]) & 0x03U))? - aln->qvs[1]->buffer[_qpos] : aln->QMIS);
} else {
score = dp->score + ((i == aln->qvs[5]->buffer[_qpos])? - aln->qvs[1]->buffer[_qpos] : aln->QMIS);
}
if(dp->path == 2){
if(i == q && i == lst_base){
if(aln->qvs[4]->buffer[_qpos] < - dp->inc){
score = score - (dp->inc) + (- aln->qvs[4]->buffer[_qpos]);
}
}
}
} else score = dp->score + (i == q? aln->M : aln->X) + dp->k->links[dp->dir * 4 + i] * aln->cov_bonus;
if(dc_exists) k = dc->links[dp->dir * 4 + i];
if(k == NULL){
knew = ((kmer << 2) | i) & aln->g->kmask;
KMER.val = dna_rev_seq(knew, aln->g->ksize);
if(KMER.val < knew){ d = 1; } else { KMER.val = knew; d = 0; }
hidx = (dbg_kmer_smear(KMER.val) % aln->g->nh);
k = get_kmerhash(aln->g->hashs[hidx], KMER);
dc->links[dp->dir * 4 + i] = k;
}
if(k == NULL){
fprintf(stderr, " -- Unexpected error in %s -- %s:%d --\n", __FUNCTION__, __FILE__, __LINE__); exit(1);
}
dp2 = next_ref_dbgdpv(aln->nodes);
dp2->aux1 = 0;
aln->bfks->size += aln->bfk_n;
dp2->score = score;// + (i == q? FLT_MIN : 0); // favor on match
if(dp->k->links[dp->dir * 4 + i] < aln->min_link_cov) dp2->score += aln->lowcov_bonus;
dp2->inc = dp2->score - dp->score;
//dp2->mat = dp->mat + (i == q? 1 : 0);// + (dp->k->links[dp->dir * 4 + i] / 256.0f); // favor on high coverage
dp2->qpos = dp->qpos + 1;
dp2->k = k;
dp2->bt_idx = idx;
dp2->path = (i == q? 0 : 3);
dp2->qdir = dp->qdir;
dp2->dir = d;
PUSH_HEAP(aln, aln->nodes->size - 1);
dp->link |= 0x1 << (i * 2 + 0);
cnt ++;
// deletion
if(dp->path != 1){ // affine gap alignment
if(aln->bfk_n) cpy_bf_kmer_dbgaln(aln, aln->nodes->size - 1);
aln->bfks->size += aln->bfk_n;
dp2 = next_ref_dbgdpv(aln->nodes);
if(aln->has_5q){ // complicated code for deletions may be more than 1bp, caution of the deletionTag matching
if(dp->qdir){
score = dp->score + ((aln->qvs[6]->buffer[_qpos] < 4 && i == ((~aln->qvs[6]->buffer[_qpos]) & 0x03))? -aln->qvs[3]->buffer[_qpos] : (dp->path == 2? aln->QEXT : aln->QDEL));
} else {
score = dp->score + (i == aln->qvs[6]->buffer[_qpos + 1]? -aln->qvs[3]->buffer[_qpos + 1] : (dp->path == 2? aln->QEXT : aln->QDEL));
}
} else score = dp->score + aln->E + (dp->path == 2? 0 : aln->D);
dp2->aux1 = 0;
dp2->score = score;
if(dp->k->links[dp->dir * 4 + i] < aln->min_link_cov) dp2->score += aln->lowcov_bonus;
dp2->inc = dp2->score - dp->score;
dp2->qpos = dp->qpos;
dp2->k = k;
dp2->bt_idx = idx;
dp2->path = 2;
dp2->qdir = dp->qdir;
dp2->dir = d;
PUSH_HEAP(aln, aln->nodes->size - 1);
dp->link |= 0x1 << (i * 2 + 1);
}
}
if(cnt == 0) return DBG_DP_CORE_DEADEND;
// inserion
if(dp->path != 2){ // affine gap alignment
if(aln->has_5q){
score = dp->score + (- aln->qvs[2]->buffer[_qpos]);
} else score = dp->score + aln->E + (dp->path == 1? 0 : aln->I);
if(aln->bfk_n) cpy_bf_kmer_dbgaln(aln, idx);
aln->bfks->size += aln->bfk_n;
dp2 = next_ref_dbgdpv(aln->nodes);
dp2->aux1 = 0;
dp2->score = score;// + FLT_MIN; // just a bit favor on insertion
dp2->inc = dp2->score - dp->score;
dp2->qpos = dp->qpos + 1;
dp2->k = dp->k;
dp2->bt_idx = idx;
dp2->path = 1;
dp2->qdir = dp->qdir;
dp2->dir = dp->dir;
PUSH_HEAP(aln, aln->nodes->size - 1);
dp->link |= 0x1 << 8;
}
return DBG_DP_CORE_EXTEND;
}
void dbg_aln(dbg_aligner *aln, uint8_t *qry, uint32_t qlen, u8list *qvs[7]){
dbg_dp_t *d, *d1, *d2, *d3;
kmer_t *k, KMER;
uint64_t kmer;
uint32_t ii, i, j, idx, sidx, hidx, *ptr;
int score, mat, dir, smax, ret, goon, found;
uint8_t lst;
clear_dbgcachehash(aln->gcache);
clear_dbgdpv(aln->qmers);
clear_dphash(aln->qhash);
clear_dbghspv(aln->hsps);
clear_dbgdpv(aln->paths);
next_ref_dbgdpv(aln->paths);
clear_dphash(aln->phash);
clear_b32list(aln->qmaxs);
for(i=0;i<=qlen;i++) push_b32list(aln->qmaxs, -0x0FFFFFFF);
aln->best_idx = 0xFFFFFFFFU;
aln->max_score = - 0x0FFFFFFF;
aln->qry = qry;
aln->qlen = qlen;
aln->qvs = qvs;
aln->has_5q = (qvs != NULL);
memset(&KMER, 0, sizeof(kmer_t));
kmer = 0;
for(i=0;i<qlen;i++){
kmer = ((kmer << 2) | qry[i]) & aln->g->kmask;
if(i + 1 < aln->g->ksize) continue;
KMER.val = dna_rev_seq(kmer, aln->g->ksize);
lst = i >= aln->g->ksize? 3 - (qry[i - aln->g->ksize] & 0x03) : 8;
if(KMER.val < kmer){ dir = 1; }
else { dir = 0; lst = lst == 8? 8 : 4 + lst; KMER.val = kmer; }
hidx = dbg_kmer_smear(KMER.val) % aln->g->nh;
k = get_kmerhash(aln->g->hashs[hidx], KMER);
if(k == NULL) continue;
//if(!aln->all_kmer && lst < 8){
//if(k->links[lst] > aln->link_cov_cutoff) continue;
//}
d = next_ref_dbgdpv(aln->qmers);
memset(d, 0, sizeof(dbg_dp_t));
d->qpos = i + 1;
d->k = k;
d->dir = dir;
push_dbghspv(aln->hsps, (dbg_hsp_t){aln->qmers->size - 1, -0x0FFFFFFF, i + 1, i + 1});
put_dphash(aln->qhash, aln->qmers->size - 1);
}
encap_dbgdpv(aln->qmers, 1);
if(verbose){
fprintf(stdout, "Found %d seed-kmers\n", (int)aln->qmers->size); fflush(stdout);
}
for(dir=0;dir<2;dir++){
for(ii=0;ii<aln->qmers->size;ii++){
i = dir? ii : aln->qmers->size - 1 - ii;
if(debug_opt1 && (int)i != atoi(debug_opt1)) continue;
d = ref_dbgdpv(aln->qmers, i);
if(d->cached) continue;
score = 0;
mat = 0;
goon = 1;
clear_dbgdpv(aln->nodes);
clear_u64list(aln->bfks);
next_ref_dbgdpv(aln->nodes);
encap_and_inc_u64list(aln->bfks, aln->bfk_n);
clear_heap(aln->heap);
clear_dphash(aln->hash);
d1 = next_ref_dbgdpv(aln->nodes);
d1->aux1 = 0;
d1->aux2 = 0;
d1->score = aln->g->ksize * (aln->has_5q? aln->QMAT : aln->M);