Pangenome-spanning epistasis and coselection analysis via de Bruijn graphs

Area of Science:

• Microbiology
• Genomics
• Evolutionary Biology

Background:

• Studying bacterial adaptation and evolution is challenging due to difficulties in measuring traits like virulence and drug resistance in large populations.
• Advancements in long-read sequencing now enable high-quality, complete bacterial genome assemblies, presenting an opportunity for new analytical methods.

Purpose of the Study:

• To develop a novel, alignment-free method for identifying co-selected and interacting genomic variations from bacterial genome assemblies.
• To leverage advances in genome sequencing to study bacterial adaptation and evolution more effectively.

Main Methods:

• A phenotype- and alignment-free approach using a compact colored de Bruijn graph.
• Approximation of intragenome distances between loci to account for linkage disequilibrium (LD).
• Application to genome assemblies covering both core and accessory genomic regions.

Main Results:

• The method efficiently identifies associations between genomic loci and key bacterial traits.
• Successfully identified loci linked to drug resistance in *Streptococcus pneumoniae* and *Enterococcus faecalis*.
• Demonstrated association between loci and adaptation to the hospital environment in these pathogens.

Conclusions:

• The developed method is versatile and efficient for discovering functionally relevant genomic variation.
• This approach facilitates a deeper understanding of bacterial adaptation, evolution, and the genetic basis of traits like drug resistance.
• Enables the exploitation of large-scale genomic data for microbial research.