Phylodynamic Inference of Bacterial Outbreak Parameters Using Nanopore Sequencing.

Area of Science:

• Genomics
• Epidemiology
• Bioinformatics

Background:

• Nanopore sequencing and phylodynamic modeling are vital for viral epidemic reconstruction but challenging for bacterial pathogens.
• Cost-effective bacterial genome sequencing and variant calling on nanopore platforms are needed for enhanced surveillance in resource-limited settings.

Purpose of the Study:

• To adapt random forest models for single nucleotide polymorphism (SNP) polishing to estimate divergence and effective reproduction numbers (Re) for bacterial outbreaks.
• To assess the utility of nanopore sequencing for reconstructing transmission dynamics of methicillin-resistant Staphylococcus aureus (MRSA) outbreaks in remote communities.

Main Methods:

• Adapted random forest models for SNP polishing to analyze low-coverage nanopore sequencing data (>5× to 10×) from MRSA isolates.
• Used birth-death skyline models for phylodynamic inference of epidemiological parameters.
• Developed Nextflow pipelines for SNP polisher training, evaluation, and outbreak alignment.

Main Results:

• Achieved high-resolution genotype inference (>90% accuracy and precision) for ST93 MRSA outbreaks compared to Illumina references.
• Successfully reconstructed phylogenetic topology, outbreak origins, and epidemic growth (Re > 1).
• Demonstrated nanopore sequencing's capability for bacterial outbreak reconstruction at competitive costs.

Conclusions:

• Nanopore sequencing combined with random forest models provides a cost-effective solution for bacterial outbreak reconstruction.
• This approach enhances infection control capabilities in remote areas and communities lacking advanced sequencing infrastructure.
• The developed methods enable the reconstruction of within-lineage transmission dynamics for bacterial disease outbreaks using portable nanopore platforms.