Identifying mixed Mycobacterium tuberculosis infections from whole genome sequence data.

Area of Science:

• Genomics
• Microbiology
• Epidemiology

Background:

• Mixed, polyclonal Mycobacterium tuberculosis infections are prevalent in natural populations.
• Accurate detection methods are vital for treatment monitoring and transmission reconstruction.
• Whole genome sequencing (WGS) data can be utilized for identifying mixed infections.

Purpose of the Study:

• To assess two WGS-based methods for detecting mixed Mycobacterium tuberculosis infections.
• To evaluate the effectiveness of these methods in distinguishing pure strains from mixed infections.
• To determine the frequency and characteristics of mixed infections in a large clinical isolate dataset.

Main Methods:

• Assessed two WGS data analysis methods: (i) heterozygous site proportions and (ii) Bayesian clustering of allele frequencies.
• Validated methods using in silico and in vitro artificially mixed and pure M. tuberculosis samples.
• Analyzed 1963 clinical isolates from the Karonga Prevention Study in Malawi.

Main Results:

• Both WGS-based methods effectively distinguished pure strains from mixed infections with >10% minor strain proportion.
• The frequency of mixed M. tuberculosis infection in the study population was approximately 10%.
• Mixed infection showed an association with the year of diagnosis but not with age, sex, HIV status, or prior tuberculosis.

Conclusions:

• Whole genome sequencing enables robust identification of mixed Mycobacterium tuberculosis infections.
• The presented methods are applicable for population-wide tuberculosis analyses and individual case identification.
• Understanding mixed infections is critical for disease evolution, transmission dynamics, and patient management.