Burrows-Wheeler Alignment Tool: A Comprehensive Guide for High-Throughput Sequence Mapping (with examples)

The Burrows-Wheeler Alignment tool (BWA) is a powerful software package widely used in bioinformatics for mapping low-divergent DNA sequences against large reference genomes. By utilizing advanced algorithms, it enables efficient alignment of short reads to reference sequences, making it a favored choice in genomic research, especially in studies involving human genomes. BWA’s accuracy and speed, especially with large datasets, makes it indispensable in analyzing high-throughput sequencing data.

Use case 1: Index the Reference Genome

Code:

bwa index path/to/reference.fa

Motivation:

Before performing any alignment, it is crucial to prepare or ‘index’ the reference genome. Indexing helps to organize the sequences in such a way that allows for faster and more efficient read mapping. Think of it as creating a roadmap before embarking on a journey; without it, the alignment process would be chaotic and time-consuming.

Explanation:

• bwa: This invokes the BWA program.
• index: This specifies the command that tells BWA to prepare an index of the reference genome.
• path/to/reference.fa: This is the file path to the reference genome sequence in FASTA format. It is the file that will be indexed by BWA.

Example output:

The command doesn’t produce an output directly onto the terminal but generates several index files (.bwt, .pac, .ann, etc.) in the same directory as the reference file. These files are essential for the efficient mapping of reads.

Use case 2: Map Single-End Reads (with Compression)

Code:

bwa mem -t 32 path/to/reference.fa path/to/read_single_end.fq.gz | gzip > path/to/alignment_single_end.sam.gz

Motivation:

Mapping single-end reads to a reference genome is a common initial step in genome analysis. This example emphasizes both speed and storage efficiency by using 32 threads for parallel processing and compressing the output. By utilizing multiple threads, researchers can significantly decrease the runtime, making it feasible to process large datasets in a shorter period.

Explanation:

• bwa: The command to invoke the BWA tool.
• mem: Specifies the BWA-MEM algorithm, optimized for high accuracy with longer reads.
• -t 32: Uses 32 threads to increase processing speed through parallel execution.
• path/to/reference.fa: The path to the indexed reference genome.
• path/to/read_single_end.fq.gz: The path to the compressed file containing single-end reads.
• | gzip > path/to/alignment_single_end.sam.gz: Pipes the output through gzip to compress the alignment results, saving disk space and making storage more manageable.

Example output:

The resulting file, alignment_single_end.sam.gz, is a compressed version of the SAM file containing the alignment of each read to the reference genome, preserving essential information needed for subsequent analyses.

Use case 3: Map Paired-End Reads (with Compression)

Code:

bwa mem -t 32 path/to/reference.fa path/to/read_pair_end_1.fq.gz path/to/read_pair_end_2.fq.gz | gzip > path/to/alignment_pair_end.sam.gz

Motivation:

Paired-end reads provide more context than single-end reads, as they consist of sequences from both ends of a DNA fragment. This additional data can significantly improve the accuracy of alignment, helping to resolve repetitive sequences and structural variations. Just like the previous use case, this example ensures fast processing and efficient data storage through multi-threading and compression.

Explanation:

• bwa: Invokes the BWA tool.
• mem: Deploys the BWA-MEM algorithm for better accuracy.
• -t 32: Utilizes 32 threads for faster processing.
• path/to/reference.fa: Path to the indexed reference genome file.
• path/to/read_pair_end_1.fq.gz and path/to/read_pair_end_2.fq.gz: Paths to the files containing paired-end reads.
• | gzip > path/to/alignment_pair_end.sam.gz: Compresses the resulting alignments using gzip.

Example output:

A compressed SAM file of alignments is saved as alignment_pair_end.sam.gz. Each line represents the alignment of a read pair, indicating how each sequence aligns with the reference genome.

Use case 4: Map Paired-End Reads with Marking Shorter Split Hits

Code:

bwa mem -M -t 32 path/to/reference.fa path/to/read_pair_end_1.fq.gz path/to/read_pair_end_2.fq.gz | gzip > path/to/alignment_pair_end.sam.gz

Motivation:

When working with certain downstream tools, such as Picard, it’s necessary to mark shorter split hits as secondary alignments. This ensures compatibility and allows for correct interpretation of the data for further analyses, such as variant calling or read group processing.

Explanation:

• bwa: Starts the alignment tool.
• mem: Uses the BWA-MEM algorithm.
• -M: Marks shorter split hits as secondary, improving compatibility with Picard.
• -t 32: Employs 32 threads for rapid computation.
• path/to/reference.fa: Location of the indexed reference genome.
• path/to/read_pair_end_1.fq.gz and path/to/read_pair_end_2.fq.gz: Paths to paired-end read files.
• | gzip > path/to/alignment_pair_end.sam.gz: Compresses and redirects the output.

Example output:

The output, alignment_pair_end.sam.gz, contains alignments where secondary alignments are explicitly marked, making it suitable for use with post-processing tools that require this distinction.

Use case 5: Map Paired-End Reads with FASTA/Q Comments

Code:

bwa mem -C -t 32 path/to/reference.fa path/to/read_pair_end_1.fq.gz path/to/read_pair_end_2.fq.gz | gzip > path/to/alignment_pair_end.sam.gz

Motivation:

Incorporating FASTA/Q comments into the alignment file can offer valuable metadata that might be required for complex analyses. This feature allows researchers to tie additional contextual data to reads, facilitating detailed insights especially important in advanced genomic studies.

Explanation:

• bwa: Command used to run the BWA tool.
• mem: Specifies the selection of the BWA-MEM algorithm.
• -C: Appends FASTA/Q comments to the SAM output, such as barcodes or indices, which could be critical for certain analytical insights.
• -t 32: Enlists 32 threads for efficient processing.
• path/to/reference.fa: The reference genome file path.
• path/to/read_pair_end_1.fq.gz and path/to/read_pair_end_2.fq.gz: Paths to paired-end read files.
• | gzip > path/to/alignment_pair_end.sam.gz: Ensures the output is compressed.

Example output:

A compressed SAM file, alignment_pair_end.sam.gz, is generated with included comments from the FASTA/Q files. This additional data can be leveraged for specialized analysis or additional downstream applications.

Conclusion:

The Burrows-Wheeler Alignment tool provides immense flexibility and power for researchers working with genomic data. From indexing reference genomes to detailed alignment processes with various features, BWA stands out as a critically important tool in bioinformatics pipelines. These use cases demonstrate its capacity to handle complex tasks efficiently, allowing for precise analysis of high-throughput sequencing data.