Bowtie 2

Bowtie 2 aligns sequencing reads to long reference sequences to enable downstream analyses such as polymorphism detection and transcriptome studies.

Key Features:

• Read length range: Supports reads from approximately 50 bases up to several hundred or even thousands of bases, including long reads >100 bp.
• Indexing: Uses an FM Index / full-text minute index to index reference genomes (e.g., human genome index typically requiring ~3.2 GB of memory).
• Alignment algorithms: Combines the full-text minute index with dynamic programming algorithms, including hardware-accelerated implementations noted in the documentation.
• Alignment modes: Supports gapped, local, and paired-end alignment modes to accommodate different sequencing protocols.
• Throughput and accuracy: Provides high throughput and competitive sensitivity and accuracy relative to other mappers, with particular effectiveness for longer reads.
• Benchmarking: Has been evaluated against other alignment tools including BWA, MAQ, RMAP, GSNAP, Novoalign, and mrsFAST in published benchmarks.
• NGS applicability: Designed to process large volumes of next-generation sequencing (NGS) data for large genomes such as mammalian genomes.

Scientific Applications:

• NGS read alignment: Maps short and long reads from next-generation sequencing experiments to reference genomes for downstream analysis.
• Polymorphism detection: Provides alignments used as input for variant calling and polymorphism discovery workflows.
• Transcriptome analysis: Supports alignments for RNA-seq analyses, including local alignment modes relevant to transcriptome studies.
• Epigenomic investigations: Produces mappings applicable to epigenomic assays that require genome-scale read alignment.
• Large-genome studies and medical diagnostics: Suited for alignment tasks involving large genomes (e.g., mammalian) relevant to research and diagnostic contexts.

Methodology:

Indexes reference genomes using an FM Index / full-text minute index and performs alignments with dynamic programming algorithms, supporting gapped, local, and paired-end modes.