CloudAligner

CloudAligner maps sequencing reads to reference genomes using Hadoop MapReduce to provide distributed alignment of short and long reads, including bisulfite-treated and paired-end data.

Key Features:

• Cloud computing integration: Implements read alignment on Hadoop using the MapReduce framework to distribute computation across many nodes.
• Performance optimization: Employs a map-only workflow by omitting the reduce phase, yielding reported performance gains of 35% to 80% over CloudBurst.
• Comprehensive functionality: Supports bisulfite sequencing, paired-end mapping, and handling of long reads from second- and third-generation NGS instruments.
• Accuracy and versatility: Reported to produce more accurate results than local-based approaches such as RMAP and supports a variety of input and output formats.
• Scalability and parallel processing: Partitions and parallel-processes large reference genomes and sequencing reads to scale to extensive genomic datasets.

Scientific Applications:

• Genome assembly: Provides large-scale read alignment suitable for inputs to genome assembly workflows.
• Variant calling: Produces aligned reads required for downstream variant detection analyses.
• Epigenetic studies (bisulfite sequencing): Aligns bisulfite-treated reads for DNA methylation and epigenomic analyses.
• General NGS alignment: Handles both short and long reads from multiple sequencing technologies for broad genomics research needs.

Methodology:

Performs read-to-reference mapping on Hadoop MapReduce using a map-only workflow (reduce phase omitted) and partitions reference genomes and reads for parallel processing.