0
|
1 #ifndef __AC_KSW_H
|
|
2 #define __AC_KSW_H
|
|
3
|
|
4 #include <stdint.h>
|
|
5
|
|
6 #define KSW_XBYTE 0x10000
|
|
7 #define KSW_XSTOP 0x20000
|
|
8 #define KSW_XSUBO 0x40000
|
|
9 #define KSW_XSTART 0x80000
|
|
10
|
|
11 struct _kswq_t;
|
|
12 typedef struct _kswq_t kswq_t;
|
|
13
|
|
14 typedef struct {
|
|
15 int score; // best score
|
|
16 int te, qe; // target end and query end
|
|
17 int score2, te2; // second best score and ending position on the target
|
|
18 int tb, qb; // target start and query start
|
|
19 } kswr_t;
|
|
20
|
|
21 #ifdef __cplusplus
|
|
22 extern "C" {
|
|
23 #endif
|
|
24
|
|
25 /**
|
|
26 * Aligning two sequences
|
|
27 *
|
|
28 * @param qlen length of the query sequence (typically <tlen)
|
|
29 * @param query query sequence with 0 <= query[i] < m
|
|
30 * @param tlen length of the target sequence
|
|
31 * @param target target sequence
|
|
32 * @param m number of residue types
|
|
33 * @param mat m*m scoring matrix in one-dimension array
|
|
34 * @param gapo gap open penalty; a gap of length l cost "-(gapo+l*gape)"
|
|
35 * @param gape gap extension penalty
|
|
36 * @param xtra extra information (see below)
|
|
37 * @param qry query profile (see below)
|
|
38 *
|
|
39 * @return alignment information in a struct; unset values to -1
|
|
40 *
|
|
41 * When xtra==0, ksw_align() uses a signed two-byte integer to store a
|
|
42 * score and only finds the best score and the end positions. The 2nd best
|
|
43 * score or the start positions are not attempted. The default behavior can
|
|
44 * be tuned by setting KSW_X* flags:
|
|
45 *
|
|
46 * KSW_XBYTE: use an unsigned byte to store a score. If overflow occurs,
|
|
47 * kswr_t::score will be set to 255
|
|
48 *
|
|
49 * KSW_XSUBO: track the 2nd best score and the ending position on the
|
|
50 * target if the 2nd best is higher than (xtra&0xffff)
|
|
51 *
|
|
52 * KSW_XSTOP: stop if the maximum score is above (xtra&0xffff)
|
|
53 *
|
|
54 * KSW_XSTART: find the start positions
|
|
55 *
|
|
56 * When *qry==NULL, ksw_align() will compute and allocate the query profile
|
|
57 * and when the function returns, *qry will point to the profile, which can
|
|
58 * be deallocated simply by free(). If one query is aligned against multiple
|
|
59 * target sequences, *qry should be set to NULL during the first call and
|
|
60 * freed after the last call. Note that qry can equal 0. In this case, the
|
|
61 * query profile will be deallocated in ksw_align().
|
|
62 */
|
|
63 kswr_t ksw_align(int qlen, uint8_t *query, int tlen, uint8_t *target, int m, const int8_t *mat, int gapo, int gape, int xtra, kswq_t **qry);
|
|
64 kswr_t ksw_align2(int qlen, uint8_t *query, int tlen, uint8_t *target, int m, const int8_t *mat, int o_del, int e_del, int o_ins, int e_ins, int xtra, kswq_t **qry);
|
|
65
|
|
66 /**
|
|
67 * Banded global alignment
|
|
68 *
|
|
69 * @param qlen query length
|
|
70 * @param query query sequence with 0 <= query[i] < m
|
|
71 * @param tlen target length
|
|
72 * @param target target sequence with 0 <= target[i] < m
|
|
73 * @param m number of residue types
|
|
74 * @param mat m*m scoring mattrix in one-dimension array
|
|
75 * @param gapo gap open penalty; a gap of length l cost "-(gapo+l*gape)"
|
|
76 * @param gape gap extension penalty
|
|
77 * @param w band width
|
|
78 * @param n_cigar (out) number of CIGAR elements
|
|
79 * @param cigar (out) BAM-encoded CIGAR; caller need to deallocate with free()
|
|
80 *
|
|
81 * @return score of the alignment
|
|
82 */
|
|
83 int ksw_global(int qlen, const uint8_t *query, int tlen, const uint8_t *target, int m, const int8_t *mat, int gapo, int gape, int w, int *n_cigar, uint32_t **cigar);
|
|
84 int ksw_global2(int qlen, const uint8_t *query, int tlen, const uint8_t *target, int m, const int8_t *mat, int o_del, int e_del, int o_ins, int e_ins, int w, int *n_cigar, uint32_t **cigar);
|
|
85
|
|
86 /**
|
|
87 * Extend alignment
|
|
88 *
|
|
89 * The routine aligns $query and $target, assuming their upstream sequences,
|
|
90 * which are not provided, have been aligned with score $h0. In return,
|
|
91 * region [0,*qle) on the query and [0,*tle) on the target sequences are
|
|
92 * aligned together. If *gscore>=0, *gscore keeps the best score such that
|
|
93 * the entire query sequence is aligned; *gtle keeps the position on the
|
|
94 * target where *gscore is achieved. Returning *gscore and *gtle helps the
|
|
95 * caller to decide whether an end-to-end hit or a partial hit is preferred.
|
|
96 *
|
|
97 * The first 9 parameters are identical to those in ksw_global()
|
|
98 *
|
|
99 * @param h0 alignment score of upstream sequences
|
|
100 * @param _qle (out) length of the query in the alignment
|
|
101 * @param _tle (out) length of the target in the alignment
|
|
102 * @param _gtle (out) length of the target if query is fully aligned
|
|
103 * @param _gscore (out) score of the best end-to-end alignment; negative if not found
|
|
104 *
|
|
105 * @return best semi-local alignment score
|
|
106 */
|
|
107 int ksw_extend(int qlen, const uint8_t *query, int tlen, const uint8_t *target, int m, const int8_t *mat, int gapo, int gape, int w, int end_bonus, int zdrop, int h0, int *qle, int *tle, int *gtle, int *gscore, int *max_off);
|
|
108 int ksw_extend2(int qlen, const uint8_t *query, int tlen, const uint8_t *target, int m, const int8_t *mat, int o_del, int e_del, int o_ins, int e_ins, int w, int end_bonus, int zdrop, int h0, int *qle, int *tle, int *gtle, int *gscore, int *max_off);
|
|
109
|
|
110 #ifdef __cplusplus
|
|
111 }
|
|
112 #endif
|
|
113
|
|
114 #endif
|