Bowtie

Bowtie and bowtie2 are ultra-fast and memory-efficient tools for aligning sequencing reads to long reference sequences. Bowtie indexes the genome with a Burrows-Wheeler index and Bowtie 2 indexes the genome with a related FM Index. Both bowtie and bowtie2 can output alignments in SAM format, enabling their incorporation with a large number of tools.     

For more detailed information refer to manuals:

• bowtie

• bowtie2

Which bowtie to Use

Both bowtie and bowtie2 are currently available on the coeus cluster:

> module avail 2>&1 | grep bowtie

Biosciences/bowtie/1.2.2/gcc

Biosciences/bowtie/2.3.4/gcc

Detailed information on the differences between bowtie and bowtie2 can be found in the bowtie2 manual. 

Bowtie2 is optimized for longer reads than bowtie, if your reads are over 50bp in length you will likely want to use bowtie2. In addition bowtie does not support gapped alignment whereas bowtie2 does. It is important to note that bowtie and bowtie2 scripts will not be cross compatible as the flags are different

Example SLURM sbatch Scripts

Bowtie alignment

This job is based on  e-coli examples from the bowtie manual                                                                                                                                                                                          First create script sub_bowtie_ex1.sh:

#!/bin/bash

#SBATCH --job-name bowtie_example

#SBATCH --partition medium

#SBATCH --output=bowtie_align_ex_%j

module purge

module load Biosciences/bowtie/1.2.2/gcc

srun bowtie -a -v 2 e_coli --suppress 1,5,6,7 -c ATGCATCATGCGCCAT

Bowtie flags used: 

• The -a flag specifies all valid alignments per read or pair will be reported

• The -v flag specifies maximum number of mismatches per alignment

• The --suppress flag specifies which columns of output (from the default output mode) will be omitted here that is the read name, read sequence, read quality, and name of reference sequence where mate's alignment occurs fields 

• The -c flag specifies the reference sequences are given on the command line rather than by file

Then submit this job to the SLURM scheduler using command sbatch:

> sbatch sub_bowtie_ex1.sh

 Bowtie2 indexing

     This job is based on Lamda phage examples from the bowtie2 manual 

#!/bin/bash

#SBATCH --job-name bowtie2_ex1

#SBATCH --partition medium

#SBATCH --output=bowtie2_index_ex_t%j

module purge

module load Biosciences/bowtie/2.3.4/gcc

srun bowtie2-build ~/example/reference/lambda_virus.fa lambda_virus

Bowtie2 alignment

This job is based on Lamda phage examples from the bowtie2 manual 

#!/bin/bash

#SBATCH --job-name bowtie2_ex2

#SBATCH --partition medium

#SBATCH --output=bowtie2_align_ex_%j

module purge

module load Biosciences/bowtie/2.3.4/gcc

srun bowtie2 -x lambda_virus -U ~/example/reads/reads_1.fq -S eg1.sam

Bowtie2 flags used: 

• The -x flag specifies the basename of the index for the reference genome

• The -U flag specifies files containing unpaired reads to be aligned

• The -S flag specifies a file to write SAM alignments to

Bowtie2 paired end read alignment

This job is based on Lamda phage examples from the bowtie2 manual 

#!/bin/bash

#SBATCH --job-name bowtie2_ex3

#SBATCH --partition medium

#SBATCH --output=bowtie2_per_ex_%j

module purge

module load Biosciences/bowtie/2.3.4/gcc

srun bowtie2 -x lambda_virus -1 ~/example/reads/reads_1.fq -2 ~/example/reads/reads_2.fq -S eg2.sam

Bowtie2 flags used: 

• The -1 flag specifies files containing mate 1s 

• The -2 flag specifies files containing mate 2s