Comparing subsampling strategies for efficient pairwise analysis of large pathogen genomic and spatial datasets: An application to Mycobacterium tuberculosis transmission.

Pairwise case-control approach with five controls per case enables efficient, unbiased analysis of large pathogen datasets, matching divide-and-conquer methods in performance.

• Why it matters: Analyzing genomic and spatial data for pathogen transmission is computationally demanding, limiting the feasibility of pairwise analysis in large datasets, which hampers understanding of disease spread.
• What they did: The study compares divide-and-conquer Bayesian modeling and pairwise case-control methods using a large Mycobacterium tuberculosis dataset (n=4,154) and simulation studies across different clustering scenarios.
• The result: The case-control approach showed negligible bias, reliable credible intervals, and fewer outliers, offering a practical, resource-efficient alternative for large-scale pairwise analysis without sacrificing accuracy.