Seedability: optimizing alignment parameters for sensitive sequence comparison

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Sequence alignment is crucial in bioinformatics.
• Current tools often use fixed k-mer lengths, potentially limiting sensitivity.
• Optimizing k-mer length is vital for accurate alignment, especially for short sequences.

Purpose of the Study:

• To develop a framework for estimating optimal seed k-mer length and minimum shared seeds.
• To improve the sensitivity of pairwise alignment for short and divergent sequences.
• To provide a method for selecting appropriate parameters in seed-based alignment tools.

Main Methods:

• Developed Seedability, a seed-based alignment framework.
• Implemented a method to estimate optimal k-mer length based on alignment identity.
• Evaluated performance using short and divergent sequence alignments.

Main Results:

• Seedability determined parameter values that improved alignments of short and divergent sequences compared to default values.
• Demonstrated cases where default parameters failed to produce alignments, but Seedability-derived parameters yielded plausible results.
• Showcased enhanced sensitivity in pairwise alignment of challenging sequence pairs.

Conclusions:

• The Seedability framework effectively identifies optimal k-mer lengths for enhanced sequence alignment sensitivity.
• Parameter optimization using Seedability leads to more robust and accurate alignments, particularly for difficult sequence types.
• This approach offers a valuable tool for improving the performance of existing bioinformatics alignment software.