BLESS、BLESS2はショートリードのエラー補正ツール。BLESSはやや遅いという問題点があった。BLESS2は並列化に対応することで高速化された。またメモリ効率、および精度の要素でもBLESSより改善しているとされる。特に複数のノードが使える条件でランタイムは最速になっている（BLESS2論文）。
インストール
cent OSに導入した。
brewで0.24が導入できる（0.2がBless2のPreliminary release）
> bless
user$ bless
----------------------------------------------------------------------
BLESS: Bloom-filter based error correction tool
----------------------------------------------------------------------
VERSION: 0.24
DATE : Jan. 29, 2015
----------------------------------------------------------------------
1. USAGE
bless <OPTIONS>
2. OPTIONS
1) REQUIRED
-read <file>: Unpaired input fastq file. One of "-read" or
"-read1/-read2" should be used.
-read1 <file>: Forward fastq file of paired-end reads. It should
be used with "-read2".
-read2 <file>: Reverse fastq file of paired-end reads. It should
be used with "-read1".
-kmerlength <number>: Length of k-mers.
-prefix <string>: Output file prefix.
2) OPTIONAL
-count <integer>: Minimum occurrences for solid k-mers.
This number is automatically determined when it is not
given.
-extend <integer>: Read extension amount. Default: 5.
-fpr <float>: Target false positive probability for the Bloom
filter. Default: 0.001.
-smpthread <integer>: Number of threads used in a SMP node.
Default: number of cores in each SMP node.
-load <prefix>: Skip the solid k-mer finding step and load
pre-built Bloom filter data. When BLESS is executed with
"-prefix <prefix>" option <prefix>.bf.data and
<prefix>.bf.size are generated. If "-load <prefix>" option
is used, BLESS does not try to fiile solid k-mers in inputs
and load the existing Bloom filter data files.
-max_mem <4-1024>: Set maximum memory usage for KMC. Default: 4.
-notrim: Turn trimming off.
-seed <integer>: Set a seed for random number generation.
Default: 0.
3. EXAMPLES
1) PAIRED INPUT READS
bless -read1 in1.fastq -read2 in2.fastq -prefix \
directory/prefix -kmerlength 31
2) UNPAIRED INPUT READ
bless -read in.fastq -prefix directory/prefix -kmerlength 31
4. CONTACT
Yun Heo (yunheo1@illinois.edu)
実行方法
シングルエンドのエラー補正。
bless -read input.fq -prefix outut -kmerlength 31
- -read <file> Unpaired input fastq file. One of "-read" or "-read1/-read2" should be used.
- -prefix <string> Output file prefix.
- -kmerlength <number> Length of k-mers.
ペアードエンドのエラー補正。
bless -read1 pair1.fq -read2 pair2.fq -prefix outut
- -read1 <file> Forward fastq file of paired-end reads. It should be used with "-read2".
- -read2 <file> Reverse fastq file of paired-end reads. It should be used with "-read1".
引用
BLESS: bloom filter-based error correction solution for high-throughput sequencing reads
Heo Y, Wu XL, Chen D, Ma J, Hwu WM
Bioinformatics. 2014 May 15;30(10):1354-62
