Bowtie2 - Sequence Alignment

After trimming adapters and low-quality sequences with Cutadapt, the next step is to align the cleaned reads to a reference genome using Bowtie2.

Alignment is often the time-limiting step.

Bowtie can be used to align short reads.

End-to-end alignment example

The following is an “end-to-end” alignment because it involves all the characters in the read (default).

Local alignment example

The following is a “local” alignment because some of the characters at the ends of the read do not participate. In this case, 4 characters are omitted (or “soft trimmed” or “soft clipped”) from the beginning and 3 characters are omitted from the end.

Bowtie2 Documentation

Loading and Running the Module

$ module spider bowtie2
$ module load bowtie2   # which version loaded?
$ module list # did you purge?
$ module spider bowtie2/2.5.4 # sometimes more detail about specific version
$ bowtie2 # to run

$ bowtie2 -help outputs a summary of command-line options.

Command Structure:

bowtie2 [options]* -x {-1 -2 | -U } [-S < sam >]

-x <bt2-idx>  Index filename prefix (minus trailing .X.bt2).

• NOTE: Bowtie 1 and Bowtie 2 indexes are not compatible.

-1 <m1>       Files with #1 mates, paired with files in <m2>.

• Could be gzip‘ed (extension: .gz) or bzip2‘ed (extension: .bz2).

-2 <m2>       Files with #2 mates, paired with files in <m1>.

• Could be gzip‘ed (extension: .gz) or bzip2‘ed (extension: .bz2).

-U <r>        Files with unpaired reads.

• Could be gzip‘ed (extension: .gz) or bzip2‘ed (extension: .bz2).

-S <sam>      File for SAM output (default: stdout)

<m1>, <m2>, <r> can be comma-separated lists (no whitespace) and can be specified many times. 

E.g. -U file1.fq,file2.fq -U file3.fq.

Using Bowtie2

First, build an index for your genome using sample data provided with the bowtie2 download.

$ ls example

Output:

index/
reads/
example/

This command creates the index:

$ __ bowtie2-build __ lambda_virus.fa lambda_virus *.bt2 index files

These index files will now be available in your directory:

lambda_virus.4.bt2
lambda_virus.3.bt2
lambda_virus.1.bt2
lambda_virus.2.bt2
lambda_virus.rev.1.bt2
lambda_virus.rev.2.bt2

Next, align the reads using this command:

$ __bowtie2__ -x lamda_virus -1 lamda-reads_1.fastq -2 lamda-reads_2.fastq -S lamda.sam

Bowtie2 Output

Building a SMALL index

10000 reads; of these:

<span style="color:#ff0000"> __Concordant alignment__ </span>

10000 (100.00%) were paired; of these:

834 (8.34%) aligned concordantly 0 times

__9166 (91.66%) aligned concordantly exactly 1 time__

0 (0.00%) aligned concordantly >1 times

----

<span style="color:#ff0000"> __Discordant alignment__ </span>

__834 pairs aligned concordantly 0 times; of these:__

42 (5.04%) aligned discordantly 1 time

----

<span style="color:#ff0000"> __The rest of the reads either align as singles __ </span>

__792 pairs aligned 0 times concordantly or discordantly__ ; of these:

1584 mates make up the pairs; of these:

1005 (63.45%) aligned 0 times

579 (36.55%) aligned exactly 1 time

0 (0.00%) aligned >1 times

__94.97% overall alignment rate__

The result summary is divided into 3 sections:

1. Concordant alignment

This shows the reads that were aligned properly as pairs. In this example, 91.66% of reads were aligned concordantly.

2. Discordant alignment

This shows the reads that did not align in the expected configuration. In this example, of the remaining 8.34% of reads, 792 reads align discordantly. That is to say, of the non-concordant fraction, 5.04% of reads (42 reads) align discordantly.

3. Single or unpaired alignments

Alignments, whether concordant or discordant, are aligned in paired-end mode. The rest of the reads either align as singles (i.e. Read1 in one locus and Read2 in a different locus, or one mate aligned and the other unaligned) or may not align at all. The reads in this section are

$$Total - (Concordant + Discordant)$$

In this example, that is

$$1000 - (9166 + 42) = 792$$

To reach the overall alignment, count the mates in total (i.e. mates aligned in paired and mates aligned in single fashion). In this example, that would be:

$$(9166 * 2) + (42 * 2) + 579 = 18995$$

That is 18885 mates aligned of total (10000 x2) mates, which is 94.97%.