1bwa(1) Bioinformatics tools bwa(1)
2
6 bwa - Burrows-Wheeler Alignment Tool
7
9 bwa index ref.fa
10 bwa mem ref.fa reads.fq > aln-se.sam
12 bwa mem ref.fa read1.fq read2.fq > aln-pe.sam
14 bwa aln ref.fa short_read.fq > aln_sa.sai
16 bwa samse ref.fa aln_sa.sai short_read.fq > aln-se.sam
18 bwa sampe ref.fa aln_sa1.sai aln_sa2.sai read1.fq read2.fq > aln-pe.sam
20 bwa bwasw ref.fa long_read.fq > aln.sam
22
25 BWA is a software package for mapping low-divergent sequences against a
26 large reference genome, such as the human genome. It consists of three
27 algorithms: BWA-backtrack, BWA-SW and BWA-MEM. The first algorithm is
28 designed for Illumina sequence reads up to 100bp, while the rest two
29 for longer sequences ranged from 70bp to 1Mbp. BWA-MEM and BWA-SW share
30 similar features such as long-read support and split alignment, but
31 BWA-MEM, which is the latest, is generally recommended for high-quality
32 queries as it is faster and more accurate. BWA-MEM also has better
33 performance than BWA-backtrack for 70-100bp Illumina reads.
34 For all the algorithms, BWA first needs to construct the FM-index for
36 the reference genome (the index command). Alignment algorithms are
37 invoked with different sub-commands: aln/samse/sampe for BWA-backtrack,
38 bwasw for BWA-SW and mem for the BWA-MEM algorithm.
39
42 index bwa index [-p prefix] [-a algoType] db.fa
43 Index database sequences in the FASTA format.
45 OPTIONS:
47 -p STR Prefix of the output database [same as db filename]
49 -a STR Algorithm for constructing BWT index. BWA implements
51 three algorithms for BWT construction: is, bwtsw and
52 rb2. The first algorithm is a little faster for small
53 database but requires large RAM and does not work for
54 databases with total length longer than 2GB. The sec‐
55 ond algorithm is adapted from the BWT-SW source code.
56 It in theory works with database with trillions of
57 bases. When this option is not specified, the appro‐
58 priate algorithm will be chosen automatically.
59 mem bwa mem [-aCHjMpP] [-t nThreads] [-k minSeedLen] [-w bandWidth]
62 [-d zDropoff] [-r seedSplitRatio] [-c maxOcc] [-D chainShadow]
63 [-m maxMateSW] [-W minSeedMatch] [-A matchScore] [-B mmPenalty]
64 [-O gapOpenPen] [-E gapExtPen] [-L clipPen] [-U unpairPen] [-R
65 RGline] [-H HDlines] [-v verboseLevel] db.prefix reads.fq
66 [mates.fq]
67 Align 70bp-1Mbp query sequences with the BWA-MEM algorithm.
69 Briefly, the algorithm works by seeding alignments with maximal
70 exact matches (MEMs) and then extending seeds with the affine-
71 gap Smith-Waterman algorithm (SW).
72 If mates.fq file is absent and option -p is not set, this com‐
74 mand regards input reads are single-end. If mates.fq is present,
75 this command assumes the i-th read in reads.fq and the i-th read
76 in mates.fq constitute a read pair. If -p is used, the command
77 assumes the 2i-th and the (2i+1)-th read in reads.fq constitute
78 a read pair (such input file is said to be interleaved). In this
79 case, mates.fq is ignored. In the paired-end mode, the mem com‐
80 mand will infer the read orientation and the insert size distri‐
81 bution from a batch of reads.
82 The BWA-MEM algorithm performs local alignment. It may produce
84 multiple primary alignments for different part of a query
85 sequence. This is a crucial feature for long sequences. However,
86 some tools such as Picard's markDuplicates does not work with
87 split alignments. One may consider to use option -M to flag
88 shorter split hits as secondary.
89 ALGORITHM OPTIONS:
92 -t INT Number of threads [1]
94 -k INT Minimum seed length. Matches shorter than INT will be
96 missed. The alignment speed is usually insensitive to
97 this value unless it significantly deviates from 20.
98 [19]
99 -w INT Band width. Essentially, gaps longer than INT will not
101 be found. Note that the maximum gap length is also
102 affected by the scoring matrix and the hit length, not
103 solely determined by this option. [100]
104 -d INT Off-diagonal X-dropoff (Z-dropoff). Stop extension
106 when the difference between the best and the current
107 extension score is above |i-j|*A+INT, where i and j
108 are the current positions of the query and reference,
109 respectively, and A is the matching score. Z-dropoff
110 is similar to BLAST's X-dropoff except that it doesn't
111 penalize gaps in one of the sequences in the align‐
112 ment. Z-dropoff not only avoids unnecessary extension,
113 but also reduces poor alignments inside a long good
114 alignment. [100]
115 -r FLOAT Trigger re-seeding for a MEM longer than min‐
117 SeedLen*FLOAT. This is a key heuristic parameter for
118 tuning the performance. Larger value yields fewer
119 seeds, which leads to faster alignment speed but lower
120 accuracy. [1.5]
121 -c INT Discard a MEM if it has more than INT occurence in the
123 genome. This is an insensitive parameter. [500]
124 -D FLOAT Drop chains shorter than FLOAT fraction of the longest
126 overlapping chain [0.5]
127 -m INT Perform at most INT rounds of mate-SW [50]
129 -W INT Drop a chain if the number of bases in seeds is
131 smaller than INT. This option is primarily used for
132 longer contigs/reads. When positive, it also affects
133 seed filtering. [0]
134 -P In the paired-end mode, perform SW to rescue missing
136 hits only but do not try to find hits that fit a
137 proper pair.
138 SCORING OPTIONS:
141 -A INT Matching score. [1]
143 -B INT Mismatch penalty. The sequence error rate is approxi‐
145 mately: {.75 * exp[-log(4) * B/A]}. [4]
146 -O INT[,INT]
148 Gap open penalty. If two numbers are specified, the
149 first is the penalty of openning a deletion and the
150 second for openning an insertion. [6]
151 -E INT[,INT]
153 Gap extension penalty. If two numbers are specified,
154 the first is the penalty of extending a deletion and
155 second for extending an insertion. A gap of length k
156 costs O + k*E (i.e. -O is for opening a zero-length
157 gap). [1]
158 -L INT[,INT]
160 Clipping penalty. When performing SW extension, BWA-
161 MEM keeps track of the best score reaching the end of
162 query. If this score is larger than the best SW score
163 minus the clipping penalty, clipping will not be
164 applied. Note that in this case, the SAM AS tag
165 reports the best SW score; clipping penalty is not
166 deduced. If two numbers are provided, the first is for
167 5'-end clipping and second for 3'-end clipping. [5]
168 -U INT Penalty for an unpaired read pair. BWA-MEM scores an
170 unpaired read pair as scoreRead1+scoreRead2-INT and
171 scores a paired as scoreRead1+scoreRead2-insert‐
172 Penalty. It compares these two scores to determine
173 whether we should force pairing. A larger value leads
174 to more aggressive read pair. [17]
175 INPUT/OUTPUT OPTIONS:
178 -p Smart pairing. If two adjacent reads have the same
180 name, they are considered to form a read pair. This
181 way, paired-end and single-end reads can be mixed in a
182 single FASTA/Q stream.
183 -R STR Complete read group header line. '\t' can be used in
185 STR and will be converted to a TAB in the output SAM.
186 The read group ID will be attached to every read in
187 the output. An example is '@RG\tID:foo\tSM:bar'.
188 [null]
189 -H ARG If ARG starts with @, it is interpreted as a string
191 and gets inserted into the output SAM header; other‐
192 wise, ARG is interpreted as a file with all lines
193 starting with @ in the file inserted into the SAM
194 header. [null]
195 -T INT Don't output alignment with score lower than INT.
197 This option affects output and occasionally SAM flag
198 2. [30]
199 -j Treat ALT contigs as part of the primary assembly
201 (i.e. ignore the db.prefix.alt file).
202 -h INT[,INT2]
204 If a query has not more than INT hits with score
205 higher than 80% of the best hit, output them all in
206 the XA tag. If INT2 is specified, BWA-MEM outputs up
207 to INT2 hits if the list contains a hit to an ALT con‐
208 tig. [5,200]
209 -a Output all found alignments for single-end or unpaired
211 paired-end reads. These alignments will be flagged as
212 secondary alignments.
213 -C Append append FASTA/Q comment to SAM output. This
215 option can be used to transfer read meta information
216 (e.g. barcode) to the SAM output. Note that the
217 FASTA/Q comment (the string after a space in the
218 header line) must conform the SAM spec (e.g.
219 BC:Z:CGTAC). Malformated comments lead to incorrect
220 SAM output.
221 -Y Use soft clipping CIGAR operation for supplementary
223 alignments. By default, BWA-MEM uses soft clipping for
224 the primary alignment and hard clipping for supplemen‐
225 tary alignments.
226 -M Mark shorter split hits as secondary (for Picard com‐
228 patibility).
229 -v INT Control the verbose level of the output. This option
231 has not been fully supported throughout BWA. Ideally,
232 a value 0 for disabling all the output to stderr; 1
233 for outputting errors only; 2 for warnings and errors;
234 3 for all normal messages; 4 or higher for debugging.
235 When this option takes value 4, the output is not SAM.
236 [3]
237 -I FLOAT[,FLOAT[,INT[,INT]]]
239 Specify the mean, standard deviation (10% of the mean
240 if absent), max (4 sigma from the mean if absent) and
241 min (4 sigma if absent) of the insert size distribu‐
242 tion. Only applicable to the FR orientation. By
243 default, BWA-MEM infers these numbers and the pair
244 orientations given enough reads. [inferred]
245 aln bwa aln [-n maxDiff] [-o maxGapO] [-e maxGapE] [-d nDelTail] [-i
249 nIndelEnd] [-k maxSeedDiff] [-l seedLen] [-t nThrds] [-cRN] [-M
250 misMsc] [-O gapOsc] [-E gapEsc] [-q trimQual] <in.db.fasta>
251 <in.query.fq> > <out.sai>
252 Find the SA coordinates of the input reads. Maximum maxSeedDiff
254 differences are allowed in the first seedLen subsequence and
255 maximum maxDiff differences are allowed in the whole sequence.
256 OPTIONS:
258 -n NUM Maximum edit distance if the value is INT, or the
260 fraction of missing alignments given 2% uniform base
261 error rate if FLOAT. In the latter case, the maximum
262 edit distance is automatically chosen for different
263 read lengths. [0.04]
264 -o INT Maximum number of gap opens [1]
266 -e INT Maximum number of gap extensions, -1 for k-difference
268 mode (disallowing long gaps) [-1]
269 -d INT Disallow a long deletion within INT bp towards the
271 3'-end [16]
272 -i INT Disallow an indel within INT bp towards the ends [5]
274 -l INT Take the first INT subsequence as seed. If INT is
276 larger than the query sequence, seeding will be dis‐
277 abled. For long reads, this option is typically ranged
278 from 25 to 35 for `-k 2'. [inf]
279 -k INT Maximum edit distance in the seed [2]
281 -t INT Number of threads (multi-threading mode) [1]
283 -M INT Mismatch penalty. BWA will not search for suboptimal
285 hits with a score lower than (bestScore-misMsc). [3]
286 -O INT Gap open penalty [11]
288 -E INT Gap extension penalty [4]
290 -R INT Proceed with suboptimal alignments if there are no
292 more than INT equally best hits. This option only
293 affects paired-end mapping. Increasing this threshold
294 helps to improve the pairing accuracy at the cost of
295 speed, especially for short reads (~32bp).
296 -c Reverse query but not complement it, which is required
298 for alignment in the color space. (Disabled since
299 0.6.x)
300 -N Disable iterative search. All hits with no more than
302 maxDiff differences will be found. This mode is much
303 slower than the default.
304 -q INT Parameter for read trimming. BWA trims a read down to
306 argmax_x{\sum_{i=x+1}^l(INT-q_i)} if q_l<INT where l
307 is the original read length. [0]
308 -I The input is in the Illumina 1.3+ read format (quality
310 equals ASCII-64).
311 -B INT Length of barcode starting from the 5'-end. When INT
313 is positive, the barcode of each read will be trimmed
314 before mapping and will be written at the BC SAM tag.
315 For paired-end reads, the barcode from both ends are
316 concatenated. [0]
317 -b Specify the input read sequence file is the BAM for‐
319 mat. For paired-end data, two ends in a pair must be
320 grouped together and options -1 or -2 are usually
321 applied to specify which end should be mapped. Typical
322 command lines for mapping pair-end data in the BAM
323 format are:
324 bwa aln ref.fa -b1 reads.bam > 1.sai
326 bwa aln ref.fa -b2 reads.bam > 2.sai
327 bwa sampe ref.fa 1.sai 2.sai reads.bam reads.bam >
328 aln.sam
329 -0 When -b is specified, only use single-end reads in
331 mapping.
332 -1 When -b is specified, only use the first read in a
334 read pair in mapping (skip single-end reads and the
335 second reads).
336 -2 When -b is specified, only use the second read in a
338 read pair in mapping.
339 samse bwa samse [-n maxOcc] <in.db.fasta> <in.sai> <in.fq> > <out.sam>
342 Generate alignments in the SAM format given single-end reads.
344 Repetitive hits will be randomly chosen.
345 OPTIONS:
347 -n INT Maximum number of alignments to output in the XA tag
349 for reads paired properly. If a read has more than INT
350 hits, the XA tag will not be written. [3]
351 -r STR Specify the read group in a format like
353 `@RG\tID:foo\tSM:bar'. [null]
354 sampe bwa sampe [-a maxInsSize] [-o maxOcc] [-n maxHitPaired] [-N max‐
357 HitDis] [-P] <in.db.fasta> <in1.sai> <in2.sai> <in1.fq> <in2.fq>
358 > <out.sam>
359 Generate alignments in the SAM format given paired-end reads.
361 Repetitive read pairs will be placed randomly.
362 OPTIONS:
364 -a INT Maximum insert size for a read pair to be considered
366 being mapped properly. Since 0.4.5, this option is only
367 used when there are not enough good alignment to infer
368 the distribution of insert sizes. [500]
369 -o INT Maximum occurrences of a read for pairing. A read with
371 more occurrneces will be treated as a single-end read.
372 Reducing this parameter helps faster pairing. [100000]
373 -P Load the entire FM-index into memory to reduce disk
375 operations (base-space reads only). With this option, at
376 least 1.25N bytes of memory are required, where N is the
377 length of the genome.
378 -n INT Maximum number of alignments to output in the XA tag for
380 reads paired properly. If a read has more than INT hits,
381 the XA tag will not be written. [3]
382 -N INT Maximum number of alignments to output in the XA tag for
384 disconcordant read pairs (excluding singletons). If a
385 read has more than INT hits, the XA tag will not be
386 written. [10]
387 -r STR Specify the read group in a format like
389 `@RG\tID:foo\tSM:bar'. [null]
390 bwasw bwa bwasw [-a matchScore] [-b mmPen] [-q gapOpenPen] [-r
393 gapExtPen] [-t nThreads] [-w bandWidth] [-T thres] [-s hspIntv]
394 [-z zBest] [-N nHspRev] [-c thresCoef] <in.db.fasta> <in.fq>
395 [mate.fq]
396 Align query sequences in the in.fq file. When mate.fq is
398 present, perform paired-end alignment. The paired-end mode only
399 works for reads Illumina short-insert libraries. In the paired-
400 end mode, BWA-SW may still output split alignments but they are
401 all marked as not properly paired; the mate positions will not
402 be written if the mate has multiple local hits.
403 OPTIONS:
405 -a INT Score of a match [1]
407 -b INT Mismatch penalty [3]
409 -q INT Gap open penalty [5]
411 -r INT Gap extension penalty. The penalty for a contiguous
413 gap of size k is q+k*r. [2]
414 -t INT Number of threads in the multi-threading mode [1]
416 -w INT Band width in the banded alignment [33]
418 -T INT Minimum score threshold divided by a [37]
420 -c FLOAT Coefficient for threshold adjustment according to
422 query length. Given an l-long query, the threshold for
423 a hit to be retained is a*max{T,c*log(l)}. [5.5]
424 -z INT Z-best heuristics. Higher -z increases accuracy at the
426 cost of speed. [1]
427 -s INT Maximum SA interval size for initiating a seed. Higher
429 -s increases accuracy at the cost of speed. [3]
430 -N INT Minimum number of seeds supporting the resultant
432 alignment to skip reverse alignment. [5]
433
436 The output of the `aln' command is binary and designed for BWA use
437 only. BWA outputs the final alignment in the SAM (Sequence Align‐
438 ment/Map) format. Each line consists of:
439 ┌────┬───────┬──────────────────────────────────────────────────────────┐
442 │Col │ Field │ Description │
443 ├────┼───────┼──────────────────────────────────────────────────────────┤
444 │ 1 │ QNAME │ Query (pair) NAME │
445 │ 2 │ FLAG │ bitwise FLAG │
446 │ 3 │ RNAME │ Reference sequence NAME │
447 │ 4 │ POS │ 1-based leftmost POSition/coordinate of clipped sequence │
448 │ 5 │ MAPQ │ MAPping Quality (Phred-scaled) │
449 │ 6 │ CIAGR │ extended CIGAR string │
450 │ 7 │ MRNM │ Mate Reference sequence NaMe (`=' if same as RNAME) │
451 │ 8 │ MPOS │ 1-based Mate POSistion │
452 │ 9 │ ISIZE │ Inferred insert SIZE │
453 │10 │ SEQ │ query SEQuence on the same strand as the reference │
454 │11 │ QUAL │ query QUALity (ASCII-33 gives the Phred base quality) │
455 │12 │ OPT │ variable OPTional fields in the format TAG:VTYPE:VALUE │
456 └────┴───────┴──────────────────────────────────────────────────────────┘
457 Each bit in the FLAG field is defined as:
459 ┌────┬────────┬───────────────────────────────────────┐
462 │Chr │ Flag │ Description │
463 ├────┼────────┼───────────────────────────────────────┤
464 │ p │ 0x0001 │ the read is paired in sequencing │
465 │ P │ 0x0002 │ the read is mapped in a proper pair │
466 │ u │ 0x0004 │ the query sequence itself is unmapped │
467 │ U │ 0x0008 │ the mate is unmapped │
468 │ r │ 0x0010 │ strand of the query (1 for reverse) │
469 │ R │ 0x0020 │ strand of the mate │
470 │ 1 │ 0x0040 │ the read is the first read in a pair │
471 │ 2 │ 0x0080 │ the read is the second read in a pair │
472 │ s │ 0x0100 │ the alignment is not primary │
473 │ f │ 0x0200 │ QC failure │
474 │ d │ 0x0400 │ optical or PCR duplicate │
475 │ S │ 0x0800 │ supplementary alignment │
476 └────┴────────┴───────────────────────────────────────┘
477 The Please check <http://samtools.sourceforge.net> for the format spec‐
479 ification and the tools for post-processing the alignment.
480 BWA generates the following optional fields. Tags starting with `X' are
482 specific to BWA.
483 ┌────┬──────────────────────────────────────────────────┐
486 │Tag │ Meaning │
487 ├────┼──────────────────────────────────────────────────┤
488 │NM │ Edit distance │
489 │MD │ Mismatching positions/bases │
490 │AS │ Alignment score │
491 │BC │ Barcode sequence │
492 │SA │ Supplementary alignments │
493 ├────┼──────────────────────────────────────────────────┤
494 │X0 │ Number of best hits │
495 │X1 │ Number of suboptimal hits found by BWA │
496 │XN │ Number of ambiguous bases in the referenece │
497 │XM │ Number of mismatches in the alignment │
498 │XO │ Number of gap opens │
499 │XG │ Number of gap extentions │
500 │XT │ Type: Unique/Repeat/N/Mate-sw │
501 │XA │ Alternative hits; format: /(chr,pos,CIGAR,NM;)*/ │
502 ├────┼──────────────────────────────────────────────────┤
503 │XS │ Suboptimal alignment score │
504 │XF │ Support from forward/reverse alignment │
505 │XE │ Number of supporting seeds │
506 └────┴──────────────────────────────────────────────────┘
507 Note that XO and XG are generated by BWT search while the CIGAR string
509 by Smith-Waterman alignment. These two tags may be inconsistent with
510 the CIGAR string. This is not a bug.
511
514 Alignment Accuracy
515 When seeding is disabled, BWA guarantees to find an alignment contain‐
516 ing maximum maxDiff differences including maxGapO gap opens which do
517 not occur within nIndelEnd bp towards either end of the query. Longer
518 gaps may be found if maxGapE is positive, but it is not guaranteed to
519 find all hits. When seeding is enabled, BWA further requires that the
520 first seedLen subsequence contains no more than maxSeedDiff differ‐
521 ences.
522 When gapped alignment is disabled, BWA is expected to generate the same
524 alignment as Eland version 1, the Illumina alignment program. However,
525 as BWA change `N' in the database sequence to random nucleotides, hits
526 to these random sequences will also be counted. As a consequence, BWA
527 may mark a unique hit as a repeat, if the random sequences happen to be
528 identical to the sequences which should be unqiue in the database.
529 By default, if the best hit is not highly repetitive (controlled by
531 -R), BWA also finds all hits contains one more mismatch; otherwise, BWA
532 finds all equally best hits only. Base quality is NOT considered in
533 evaluating hits. In the paired-end mode, BWA pairs all hits it found.
534 It further performs Smith-Waterman alignment for unmapped reads to res‐
535 cue reads with a high erro rate, and for high-quality anomalous pairs
536 to fix potential alignment errors.
537 Estimating Insert Size Distribution
540 BWA estimates the insert size distribution per 256*1024 read pairs. It
541 first collects pairs of reads with both ends mapped with a single-end
542 quality 20 or higher and then calculates median (Q2), lower and higher
543 quartile (Q1 and Q3). It estimates the mean and the variance of the
544 insert size distribution from pairs whose insert sizes are within
545 interval [Q1-2(Q3-Q1), Q3+2(Q3-Q1)]. The maximum distance x for a pair
546 considered to be properly paired (SAM flag 0x2) is calculated by solv‐
547 ing equation Phi((x-mu)/sigma)=x/L*p0, where mu is the mean, sigma is
548 the standard error of the insert size distribution, L is the length of
549 the genome, p0 is prior of anomalous pair and Phi() is the standard
550 cumulative distribution function. For mapping Illumina short-insert
551 reads to the human genome, x is about 6-7 sigma away from the mean.
552 Quartiles, mean, variance and x will be printed to the standard error
553 output.
554 Memory Requirement
557 With bwtsw algorithm, 5GB memory is required for indexing the complete
558 human genome sequences. For short reads, the aln command uses ~3.2GB
559 memory and the sampe command uses ~5.4GB.
560 Speed
563 Indexing the human genome sequences takes 3 hours with bwtsw algorithm.
564 Indexing smaller genomes with IS algorithms is faster, but requires
565 more memory.
566 The speed of alignment is largely determined by the error rate of the
568 query sequences (r). Firstly, BWA runs much faster for near perfect
569 hits than for hits with many differences, and it stops searching for a
570 hit with l+2 differences if a l-difference hit is found. This means BWA
571 will be very slow if r is high because in this case BWA has to visit
572 hits with many differences and looking for these hits is expensive.
573 Secondly, the alignment algorithm behind makes the speed sensitive to
574 [k log(N)/m], where k is the maximum allowed differences, N the size of
575 database and m the length of a query. In practice, we choose k w.r.t. r
576 and therefore r is the leading factor. I would not recommend to use BWA
577 on data with r>0.02.
578 Pairing is slower for shorter reads. This is mainly because shorter
580 reads have more spurious hits and converting SA coordinates to chromo‐
581 somal coordinates are very costly.
582
585 Since version 0.6, BWA has been able to work with a reference genome
586 longer than 4GB. This feature makes it possible to integrate the for‐
587 ward and reverse complemented genome in one FM-index, which speeds up
588 both BWA-short and BWA-SW. As a tradeoff, BWA uses more memory because
589 it has to keep all positions and ranks in 64-bit integers, twice larger
590 than 32-bit integers used in the previous versions.
591 The latest BWA-SW also works for paired-end reads longer than 100bp. In
593 comparison to BWA-short, BWA-SW tends to be more accurate for highly
594 unique reads and more robust to relative long INDELs and structural
595 variants. Nonetheless, BWA-short usually has higher power to distin‐
596 guish the optimal hit from many suboptimal hits. The choice of the map‐
597 ping algorithm may depend on the application.
598
601 BWA website <http://bio-bwa.sourceforge.net>, Samtools website
602 <http://samtools.sourceforge.net>
603
606 Heng Li at the Sanger Institute wrote the key source codes and inte‐
607 grated the following codes for BWT construction: bwtsw
608 <http://i.cs.hku.hk/~ckwong3/bwtsw/>, implemented by Chi-Kwong Wong at
609 the University of Hong Kong and IS
610 <http://yuta.256.googlepages.com/sais> originally proposed by Nong Ge
611 <http://www.cs.sysu.edu.cn/nong/> at the Sun Yat-Sen University and
612 implemented by Yuta Mori.
613
616 The full BWA package is distributed under GPLv3 as it uses source codes
617 from BWT-SW which is covered by GPL. Sorting, hash table, BWT and IS
618 libraries are distributed under the MIT license.
619 If you use the BWA-backtrack algorithm, please cite the following
621 paper:
622 Li H. and Durbin R. (2009) Fast and accurate short read alignment with
624 Burrows-Wheeler transform. Bioinformatics, 25, 1754-1760. [PMID:
625 19451168]
626 If you use the BWA-SW algorithm, please cite:
628 Li H. and Durbin R. (2010) Fast and accurate long-read alignment with
630 Burrows-Wheeler transform. Bioinformatics, 26, 589-595. [PMID:
631 20080505]
632 If you use BWA-MEM or the fastmap component of BWA, please cite:
634 Li H. (2013) Aligning sequence reads, clone sequences and assembly con‐
636 tigs with BWA-MEM. arXiv:1303.3997v1 [q-bio.GN].
637 It is likely that the BWA-MEM manuscript will not appear in a peer-
639 reviewed journal.
640
643 BWA is largely influenced by BWT-SW. It uses source codes from BWT-SW
644 and mimics its binary file formats; BWA-SW resembles BWT-SW in several
645 ways. The initial idea about BWT-based alignment also came from the
646 group who developed BWT-SW. At the same time, BWA is different enough
647 from BWT-SW. The short-read alignment algorithm bears no similarity to
648 Smith-Waterman algorithm any more. While BWA-SW learns from BWT-SW, it
649 introduces heuristics that can hardly be applied to the original algo‐
650 rithm. In all, BWA does not guarantee to find all local hits as what
651 BWT-SW is designed to do, but it is much faster than BWT-SW on both
652 short and long query sequences.
653 I started to write the first piece of codes on 24 May 2008 and got the
655 initial stable version on 02 June 2008. During this period, I was
656 acquainted that Professor Tak-Wah Lam, the first author of BWT-SW
657 paper, was collaborating with Beijing Genomics Institute on SOAP2, the
658 successor to SOAP (Short Oligonucleotide Analysis Package). SOAP2 has
659 come out in November 2008. According to the SourceForge download page,
660 the third BWT-based short read aligner, bowtie, was first released in
661 August 2008. At the time of writing this manual, at least three more
662 BWT-based short-read aligners are being implemented.
663 The BWA-SW algorithm is a new component of BWA. It was conceived in No‐
665 vember 2008 and implemented ten months later.
666 The BWA-MEM algorithm is based on an algorithm finding super-maximal
668 exact matches (SMEMs), which was first published with the fermi assem‐
669 bler paper in 2012. I first implemented the basic SMEM algorithm in the
670 fastmap command for an experiment and then extended the basic algorithm
671 and added the extension part in Feburary 2013 to make BWA-MEM a fully
672 featured mapper.
673bwa-0.7.15-r1140 31 May 2016 bwa(1)