Beyond the SNP Threshold: Identifying Outbreak Clusters Using Inferred Transmissions.

Area of Science:

• Epidemiology
• Genomics
• Bioinformatics

Background:

• Whole-genome sequencing (WGS) is crucial for understanding pathogen transmission dynamics.
• Defining transmission clusters based on single-nucleotide polymorphism (SNP) thresholds is common but lacks standardization.
• Existing methods struggle with variable mutation rates and diverse sampling times.

Purpose of the Study:

• To develop a probabilistic framework for defining transmission clusters.
• To improve the accuracy of identifying direct pathogen transmissions.
• To offer a more robust alternative to SNP-threshold methods.

Main Methods:

• Developed a probabilistic model combining SNP differences with temporal data (case timing, molecular clock).
• Incorporated variable mutation rates across the genome and epidemiological data.
• Applied the framework to tuberculosis (TB) datasets from British Columbia and the Republic of Moldova.

Main Results:

• The transmission-based method demonstrated superior performance in identifying direct transmissions compared to SNP-threshold methods.
• Simulation results showed lower dissimilarity (0.27 bits) for the proposed method versus SNP-threshold (0.37 bits) and random data (0.84 bits).
• The approach is particularly effective with variable molecular clock rates and spread-out sample collection times.

Conclusions:

• The proposed probabilistic, transmission-focused approach offers a more accurate and reliable method for pathogen transmission cluster analysis.
• This framework enhances the utility of WGS data in public health surveillance and outbreak investigations.
• The method is implemented in the R package 'transcluster' for broader application.