forked from DaehwanKimLab/hisat2
-
Notifications
You must be signed in to change notification settings - Fork 6
/
aligner_sw_common.h
305 lines (277 loc) · 8.53 KB
/
aligner_sw_common.h
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
/*
* Copyright 2011, Ben Langmead <[email protected]>
*
* This file is part of Bowtie 2.
*
* Bowtie 2 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.
*
* Bowtie 2 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 Bowtie 2. If not, see <http://www.gnu.org/licenses/>.
*/
#ifndef ALIGNER_SW_COMMON_H_
#define ALIGNER_SW_COMMON_H_
#include "aligner_result.h"
/**
* Encapsulates the result of a dynamic programming alignment, including
* colorspace alignments. In our case, the result is a combination of:
*
* 1. All the nucleotide edits
* 2. All the "edits" where an ambiguous reference char is resolved to
* an unambiguous char.
* 3. All the color edits (if applicable)
* 4. All the color miscalls (if applicable). This is a subset of 3.
* 5. The score of the best alginment
* 6. The score of the second-best alignment
*
* Having scores for the best and second-best alignments gives us an
* idea of where gaps may make reassembly beneficial.
*/
struct SwResult {
SwResult() :
alres(),
sws(0),
swcups(0),
swrows(0),
swskiprows(0),
swskip(0),
swsucc(0),
swfail(0),
swbts(0)
{ }
/**
* Clear all contents.
*/
void reset() {
sws = swcups = swrows = swskiprows = swskip = swsucc =
swfail = swbts = 0;
alres.reset();
}
/**
* Reverse all edit lists.
*/
void reverse() {
alres.reverseEdits();
}
/**
* Return true iff no result has been installed.
*/
bool empty() const {
return alres.empty();
}
#ifndef NDEBUG
/**
* Check that result is internally consistent.
*/
bool repOk() const {
assert(alres.repOk());
return true;
}
/**
* Check that result is internally consistent w/r/t read.
*/
bool repOk(const Read& rd) const {
assert(alres.repOk(rd));
return true;
}
#endif
AlnRes alres;
uint64_t sws; // # DP problems solved
uint64_t swcups; // # DP cell updates
uint64_t swrows; // # DP row updates
uint64_t swskiprows; // # skipped DP row updates (b/c no valid alignments can go thru row)
uint64_t swskip; // # DP problems skipped by sse filter
uint64_t swsucc; // # DP problems resulting in alignment
uint64_t swfail; // # DP problems not resulting in alignment
uint64_t swbts; // # DP backtrace steps
int nup; // upstream decoded nucleotide; for colorspace reads
int ndn; // downstream decoded nucleotide; for colorspace reads
};
/**
* Encapsulates counters that measure how much work has been done by
* the dynamic programming driver and aligner.
*/
struct SwMetrics {
SwMetrics() : mutex_m() {
reset();
}
void reset() {
sws = swcups = swrows = swskiprows = swskip = swsucc = swfail = swbts =
sws10 = sws5 = sws3 =
rshit = ungapsucc = ungapfail = ungapnodec = 0;
exatts = exranges = exrows = exsucc = exooms = 0;
mm1atts = mm1ranges = mm1rows = mm1succ = mm1ooms = 0;
sdatts = sdranges = sdrows = sdsucc = sdooms = 0;
}
void init(
uint64_t sws_,
uint64_t sws10_,
uint64_t sws5_,
uint64_t sws3_,
uint64_t swcups_,
uint64_t swrows_,
uint64_t swskiprows_,
uint64_t swskip_,
uint64_t swsucc_,
uint64_t swfail_,
uint64_t swbts_,
uint64_t rshit_,
uint64_t ungapsucc_,
uint64_t ungapfail_,
uint64_t ungapnodec_,
uint64_t exatts_,
uint64_t exranges_,
uint64_t exrows_,
uint64_t exsucc_,
uint64_t exooms_,
uint64_t mm1atts_,
uint64_t mm1ranges_,
uint64_t mm1rows_,
uint64_t mm1succ_,
uint64_t mm1ooms_,
uint64_t sdatts_,
uint64_t sdranges_,
uint64_t sdrows_,
uint64_t sdsucc_,
uint64_t sdooms_)
{
sws = sws_;
sws10 = sws10_;
sws5 = sws5_;
sws3 = sws3_;
swcups = swcups_;
swrows = swrows_;
swskiprows = swskiprows_;
swskip = swskip_;
swsucc = swsucc_;
swfail = swfail_;
swbts = swbts_;
ungapsucc = ungapsucc_;
ungapfail = ungapfail_;
ungapnodec = ungapnodec_;
// Exact end-to-end attempts
exatts = exatts_;
exranges = exranges_;
exrows = exrows_;
exsucc = exsucc_;
exooms = exooms_;
// 1-mismatch end-to-end attempts
mm1atts = mm1atts_;
mm1ranges = mm1ranges_;
mm1rows = mm1rows_;
mm1succ = mm1succ_;
mm1ooms = mm1ooms_;
// Seed attempts
sdatts = sdatts_;
sdranges = sdranges_;
sdrows = sdrows_;
sdsucc = sdsucc_;
sdooms = sdooms_;
}
/**
* Merge (add) the counters in the given SwResult object into this
* SwMetrics object.
*/
void update(const SwResult& r) {
sws += r.sws;
swcups += r.swcups;
swrows += r.swrows;
swskiprows += r.swskiprows;
swskip += r.swskip;
swsucc += r.swsucc;
swfail += r.swfail;
swbts += r.swbts;
}
/**
* Merge (add) the counters in the given SwMetrics object into this
* object. This is the only safe way to update a SwMetrics shared
* by multiple threads.
*/
void merge(const SwMetrics& r, bool getLock = false) {
ThreadSafe ts(&mutex_m, getLock);
sws += r.sws;
sws10 += r.sws10;
sws5 += r.sws5;
sws3 += r.sws3;
swcups += r.swcups;
swrows += r.swrows;
swskiprows += r.swskiprows;
swskip += r.swskip;
swsucc += r.swsucc;
swfail += r.swfail;
swbts += r.swbts;
rshit += r.rshit;
ungapsucc += r.ungapsucc;
ungapfail += r.ungapfail;
ungapnodec += r.ungapnodec;
exatts += r.exatts;
exranges += r.exranges;
exrows += r.exrows;
exsucc += r.exsucc;
exooms += r.exooms;
mm1atts += r.mm1atts;
mm1ranges += r.mm1ranges;
mm1rows += r.mm1rows;
mm1succ += r.mm1succ;
mm1ooms += r.mm1ooms;
sdatts += r.sdatts;
sdranges += r.sdranges;
sdrows += r.sdrows;
sdsucc += r.sdsucc;
sdooms += r.sdooms;
}
void tallyGappedDp(size_t readGaps, size_t refGaps) {
size_t mx = max(readGaps, refGaps);
if(mx < 10) sws10++;
if(mx < 5) sws5++;
if(mx < 3) sws3++;
}
uint64_t sws; // # DP problems solved
uint64_t sws10; // # DP problems solved where max gaps < 10
uint64_t sws5; // # DP problems solved where max gaps < 5
uint64_t sws3; // # DP problems solved where max gaps < 3
uint64_t swcups; // # DP cell updates
uint64_t swrows; // # DP row updates
uint64_t swskiprows; // # skipped DP rows (b/c no valid alns go thru row)
uint64_t swskip; // # DP problems skipped by sse filter
uint64_t swsucc; // # DP problems resulting in alignment
uint64_t swfail; // # DP problems not resulting in alignment
uint64_t swbts; // # DP backtrace steps
uint64_t rshit; // # DP problems avoided b/c seed hit was redundant
uint64_t ungapsucc; // # DP problems avoided b/c seed hit was redundant
uint64_t ungapfail; // # DP problems avoided b/c seed hit was redundant
uint64_t ungapnodec; // # DP problems avoided b/c seed hit was redundant
uint64_t exatts; // total # attempts at exact-hit end-to-end aln
uint64_t exranges; // total # ranges returned by exact-hit queries
uint64_t exrows; // total # rows returned by exact-hit queries
uint64_t exsucc; // exact-hit yielded non-empty result
uint64_t exooms; // exact-hit offset memory exhausted
uint64_t mm1atts; // total # attempts at 1mm end-to-end aln
uint64_t mm1ranges; // total # ranges returned by 1mm-hit queries
uint64_t mm1rows; // total # rows returned by 1mm-hit queries
uint64_t mm1succ; // 1mm-hit yielded non-empty result
uint64_t mm1ooms; // 1mm-hit offset memory exhausted
uint64_t sdatts; // total # attempts to find seed alignments
uint64_t sdranges; // total # seed-alignment ranges found
uint64_t sdrows; // total # seed-alignment rows found
uint64_t sdsucc; // # times seed alignment yielded >= 1 hit
uint64_t sdooms; // # times an OOM occurred during seed alignment
MUTEX_T mutex_m;
};
// The various ways that one might backtrack from a later cell (either oall,
// rdgap or rfgap) to an earlier cell
enum {
SW_BT_OALL_DIAG, // from oall cell to oall cell
SW_BT_OALL_REF_OPEN, // from oall cell to oall cell
SW_BT_OALL_READ_OPEN, // from oall cell to oall cell
SW_BT_RDGAP_EXTEND, // from rdgap cell to rdgap cell
SW_BT_RFGAP_EXTEND // from rfgap cell to rfgap cell
};
#endif /*def ALIGNER_SW_COMMON_H_*/