Hogwash: three methods for genome-wide association studies in bacteria

Area of Science:

• Microbiology
• Genomics
• Bioinformatics

Background:

• Bacterial genome-wide association studies (bGWAS) identify links between genetic variations and observable traits.
• Convergence-based bGWAS methods detect genomic mutations that arise independently multiple times, suggesting a strong selective pressure related to a phenotype.

Purpose of the Study:

• Introduce `hogwash`, an open-source R package for implementing convergence-based bGWAS.
• Provide tools for gene or pathway analysis to increase the power of detecting weakly penetrant or related genotypes.

Main Methods:

• Implemented three convergence-based bGWAS algorithms within the `hogwash` R package.
• Incorporated two burden testing approaches for gene and pathway-level analysis.
• Utilized simulated data with varying phylogenetic signals and convergence patterns to test `hogwash`.

Main Results:

• The `hogwash` package offers a robust framework for convergence-based bGWAS.
• Simulated data analysis helped identify optimal use cases for the implemented algorithms.
• Publicly available simulated datasets facilitate reproducible research and method validation.

Conclusions:

• `hogwash` provides a valuable, open-source resource for bacterial evolutionary genomics research.
• The package enhances the ability to detect genotype-phenotype associations, particularly for convergent mutations.
• Future applications can leverage `hogwash` for exploring bacterial adaptation and trait evolution.