Seedability

Seedability optimizes seed-based sequence alignment parameters to improve alignment sensitivity and accuracy by estimating optimal k-mer length and the minimal number of shared seeds for specified alignment identity thresholds.

Key Features:

• Optimization of k-mer length: Dynamically estimates the most effective k-mer length to increase sensitivity and accuracy of seed-based alignments across datasets.
• Minimal shared seeds estimation: Calculates the minimal number of shared seeds required to achieve reliable alignment outcomes, aiding alignments of short or divergent sequences.
• Alignment identity thresholding: Uses user-specified alignment identity thresholds to tailor sensitivity and specificity of parameter selection.

Scientific Applications:

• Improving seed-based aligners: Provides optimized parameters that improve alignment results compared with default settings in Minimap2.
• Short and divergent sequence alignment: Enhances alignment sensitivity for short or highly divergent sequences that are challenging for fixed-parameter methods.
• Comparative genomics and variant analysis: Supports detection and analysis of genomic relationships relevant to evolutionary studies, variant calling, and comparative genomics.

Methodology:

Systematic evaluation of different k-mer lengths and shared seed counts across specified alignment identity thresholds to identify parameter configurations that maximize alignment sensitivity while maintaining computational efficiency, with experimental comparisons to Minimap2 default settings.