s-leaping

s-leaping implements an approximate downsampling algorithm for large high-throughput sequencing datasets to enable efficient, accurate estimation of sequencing coverage for omics analyses.

Key Features:

• Efficiency: Processes large datasets and is reported up to 39% faster than the second-fastest existing downsampling methods while maintaining comparable accuracy to exact downsampling.
• Accuracy: Uses an approximate approach that achieves accuracy closely aligned with traditional exact downsampling methods, suitable for practical omics-study applications.
• Lightweight implementation (fadso): Includes fadso, a C implementation for FASTQ downsampling that is up to 12% faster and uses 21% less memory than commonly used FASTQ tools with similar capabilities.
• High throughput: Reduced memory usage of the fadso implementation enables up to 40% higher throughput in parallel computing environments.

Scientific Applications:

• Omics study design: Enables evaluation of sequencing coverage requirements to inform cost-effective experimental design in omics research.
• Sequencing coverage estimation: Facilitates estimation of coverage needed for downstream analyses by producing representative downsampled subsets of reads.
• Large-scale genomic processing: Supports high-throughput downsampling workflows for large genomic datasets where memory and runtime are limiting factors.

Methodology:

s-leaping employs an approximate downsampling algorithm optimized for large datasets, and the accompanying fadso C implementation performs resource-efficient FASTQ downsampling with reduced memory usage.