An alignment- and reference-free strategy using k-mer present pattern for population genomic analyses

Area of Science:

• Genomics
• Population Genetics
• Bioinformatics

Background:

• Pangenome analysis traditionally relies on graph-based methods, requiring multiple high-quality genomes and intensive computational alignment.
• Existing methods for pangenome analysis are computationally demanding and necessitate complete genome sequences.
• The direct application of k-mer decomposition for population genetic analyses remained unexplored.

Purpose of the Study:

• To develop and validate a novel strategy for population genetic analysis using k-mer count variants.
• To establish an alignment- and reference-free (ARF) framework for population genetics.
• To explore the utility of k-mers in detecting various genetic variations beyond single nucleotide polymorphisms (SNPs).

Main Methods:

• Developed a novel strategy utilizing k-mer count variants for population genetic analyses.
• Compared the k-mer based method against SNP-based methods using simulated and real datasets.
• Analyzed population structure, genetic diversity, and identified genetic variants in Saccharomyces cerevisiae.

Main Results:

• The k-mer based approach successfully recapitulated population structure identified by SNPs.
• The k-mer method revealed higher genetic diversity in the real dataset, indicating a greater capture of genetic variants.
• Identified SNPs, insertion/deletion, and horizontal gene transfer (HGT) fragments associated with adaptive evolution using k-mer frequencies.

Conclusions:

• The developed k-mer based strategy is effective for population genetic analyses, providing an alignment- and reference-free alternative.
• This ARF method is particularly advantageous for species with incomplete or highly divergent genomes.
• The approach broadens the scope of pangenome analysis by incorporating k-mer based insights into population genetics.