Inference of Pairwise Interactions from Strain Frequency Data Across Settings and Context-Dependent Mutual...

Area of Science:

• Microbiology and Epidemiology
• Mathematical Modeling
• Population Genetics

Background:

• Understanding microbial population dynamics requires characterizing interactions between different strains.
• Previous methods often struggle to infer complex interaction networks from cross-sectional data.

Purpose of the Study:

• To develop and validate a novel computational framework for estimating pairwise strain interactions from population-level frequencies.
• To apply this method to Streptococcus pneumoniae serotype data from diverse global settings.

Main Methods:

• Utilized replicator dynamics derived from a multi-strain SIS model with co-colonization.
• Integrated epidemiological data on Streptococcus pneumoniae serotype frequencies from five countries.
• Employed basic reproduction number (R0), mean global susceptibility (k), and pairwise deviations (αij) to model interactions.

Main Results:

• Successfully inferred over 70% of the 92x92 Streptococcus pneumoniae serotype interaction matrix.
• Demonstrated that within- and between-serotype interaction coefficients exhibit unimodal distributions.
• Showcased the method's proof-of-concept for inferring multi-species interactions from cross-sectional data.

Conclusions:

• The proposed framework provides a high-resolution map of pneumococcal serotype interactions.
• This approach enables robust investigation of intervention effects in complex microbial ecosystems.
• The method is adaptable for both cross-sectional and longitudinal data analysis.