Alpaca: a kmer-based approach for investigating mosaic structures in microbial genomes

Abstract

Microbial genomes are often mosaic: different regions can possess different evolutionary origins due to genetic recombination. The recent feasibility to assemble microbial genomes completely and the availability of sequencing data for complete microbial populations, means that researchers can now investigate the potentially rich evolutionary history of a microbe at a much higher resolution. Here, we present Alpaca: a method to investigate mosaicism in microbial genomes based on kmer similarity of large sequencing datasets. Alpaca partitions a given assembly into various sub-regions and compares their similarity across a population of genomes. The result is a high-resolution map of an entire genome and the most similar scoring clades across the given population.Availability: https://github.com/AbeelLab/Alpaca Contact: t.abeel@tudelft.nl

Nanopore sequencing and comparative genome analysis confirm lager-brewing yeasts originated from a single hybridization

“…Therefore, we developed Alpaca: a simple and computationally inexpensive method to investigate complex non-linear ancestry via comparison of sequencing datasets (61). Alpaca is based on short-read alignment of a collection of strains to a partitioned reference genome, in which the similarity of each partition to the collection of strains is independently computed using k-mer sets (61). Reducing the alignments in each partition to k-mer sets prior to similarity analysis is computationally inexpensive.…”

“…Reducing the alignments in each partition to k-mer sets prior to similarity analysis is computationally inexpensive. Phylogenetic relationships are also not recalculated, but simply inferred from previously available information on the population structure of the collection of strains (61). The partitioning of the reference genome enables the identification of strains with high similarity to different regions of the genome, enabling the identification of ancestry resulting from non-linear evolution.…”

“…We developed Alpaca (61) to investigate non-linear ancestry of a reference genome based on large sequencing datasets. Briefly, Alpaca partitions a reference genome into multiple sub-regions, each reduced to a k-mer set representation.…”

“…Non-linear ancestry can therefore be inferred by tracing the population origin of the most similar genome(s) in each sub-region. Detailed explanation Alpaca can be found in our method description (61).…”

“…MASH (version 2.1) ( 64 ) was then used to hierarchically cluster each genome based on their MASH distance using k-mer size of 21, sketch size of 1,000,000, and minimum k-mer frequency of 2. The resulting trees were used as population reference trees for Alpaca (61).…”