Read-Consistent Minimum Unique Substrings: A Parameter-Free, Linear-Time Framework for Genomic Sequence...

Area of Science:

• Genomics and Bioinformatics
• Computational Biology
• Sequence Analysis

Background:

• Fixed-length k-mers are standard in sequence analysis but lack resolution for heterogeneous genomes.
• K-mers can cause redundancy and lose contextual sensitivity due to uniform length.
• Existing methods struggle with accurate repeat boundary definition and genomic complexity.

Purpose of the Study:

• Introduce Minimum Unique Substrings (MUSs) as variable-length sequence units.
• Develop a context-aware method that adapts to local genomic complexity.
• Provide a biologically meaningful alternative to fixed-length k-mers for sequence analysis.

Main Methods:

• Developed a linear-time algorithm (O(n)) using a generalized suffix tree.
• Introduced 'outposts' for precise localization of MUS boundaries.
• Defined a read-consistent measure of uniqueness for sequencing reads.

Main Results:

• MUS framework achieves 100% unique coverage with an average length of 36.08 bp.
• Demonstrated distinct MUS length distributions reflecting bacterial (30.44 bp) and human (36.08 bp) genomes.
• Achieved >99% reduction in tokens compared to k-mers, offering higher resolution and data compression.

Conclusions:

• MUSs provide a biologically meaningful, context-sensitive alternative to k-mers.
• The MUS framework enhances genome assembly, repeat characterization, and comparative genomics.
• Variable-length MUSs adapt to genomic architecture, overcoming limitations of fixed-length k-mers.