SEAL

Seal: Distributed short-read alignment and duplicate removal framework

Seal performs distributed alignment of short DNA sequencing reads and duplicate removal using the Hadoop framework, generating mappings consistent with BWA outputs and applying duplicate marking according to Picard MarkDuplicates criteria.

Key Features:

• Scalability: Distributes alignment and duplicate removal tasks across a Hadoop cluster for large-scale genomic datasets.
• BWA-Compatible Alignment: Produces read mappings aligned with BWA outputs to ensure consistency and accuracy.
• Picard MarkDuplicates Integration: Removes duplicates following Picard MarkDuplicates criteria for standardized data cleaning.
• Performance: Processes approximately 13 GB/hour in map+rmdup mode and up to 19 GB/hour in mapping-only mode on a 16-node Hadoop cluster.

Scientific Applications:

• Short-Read Analysis: Supports genome sequencing, variant calling, and next-generation sequencing workflows requiring accurate alignment and duplicate removal.

Methodology:

Implements parallel processing of genomic data using the Hadoop distributed computing framework. Aligns short DNA reads to reference genomes with output consistent with BWA, followed by duplicate identification and removal based on Picard MarkDuplicates standards.