Why This Matters:
- Traditional transmission analysis often focuses on single pathogens isolated in culture, limiting interpretation of polymicrobial and microbiome-associated transmission events.
- Increasing use of metagenomic sequencing in healthcare and environmental surveillance is generating complex multi-kingdom datasets that require new analytical approaches.
- Strain-level resolution is critical for distinguishing true transmission events from shared background microbiota or environmental contamination.
- Improved reconstruction of microbial transmission networks may enhance outbreak detection, infection prevention, microbiome research, and understanding of microbial ecology within healthcare systems.
Key Findings: The authors developed TRACS, a computational framework for strain-level transmission inference from longitudinal metagenomic data across bacteria, viruses, fungi, and bacteriophages.1 The platform enables culture-independent strain resolution using SNP-level variation and integrates genomic similarity, time, and epidemiologic metadata to infer transmission events. The method is implemented in Python/C++ and using github.com/gtonkinhill/tracs. Comparisons were made against InStrain, StrainGE, and StrainPhlAn.
- Benchmarking: TRACS showed improved accuracy, avoiding inflated genetic distance estimates, particularly in low-abundance strains. In SARS-CoV-2 datasets it inferred zero SNP differences without manual filtering, and in S. pneumoniae it estimated intermediate host counts using sampling dates.
- Birth cohort transmission: In 1,288 infants (UK BabyBiome), TRACS found strain sharing was rare among unrelated children at the sme facility (0.73–0.78%) but showed high strain sharing among siblings. Caesarean delivery was associated with reduced strain sharing. Key taxa showing delivery-associated effects included B. bifidum, B. longum, P. vulgatus, P. distasonis, and E. coli.
- Within-host persistence: TRACS tracks longitudinal strain stability, revealing species-dependent persistence patterns.
- FMT engraftment: Across 3 clinical cohorts, TRACS showed high sensitivity and low false-positive rates, outperforming comparator tools in detecting multi-strain engraftment, including multiple B. longum strains per sample.
- Transmission vs colonization: TRACS reduces false-positive transmission inference by distinguishing true strain transmission from shared background species using fine-scale genomic relatedness.
Bigger Picture: As metagenomic sequencing becomes increasingly integrated into clinical microbiology and public health surveillance, analysis is shifting from data generation to interpretation of highly complex microbial community data. In this context, TRACS highlights how strain-resolved metagenomics may expand transmission analysis beyond single bacterial pathogens to include fungi, viruses, and bacteriophages within interconnected microbial ecosystems. This is particularly important in healthcare environments, where transmission may involve mixed microbial communities rather than isolated organisms.
Limitations: Performance depends on sequencing depth, genome coverage, longitudinal sampling density, and availability of epidemiologic metadata. Transmission directionality remains probabilistic rather than definitive.
With validation, TRACS could be applied to:
- enhanced outbreak reconstruction
- earlier detection of healthcare-associated transmission
- microbiome-informed infection prevention and treatment strategies
- improved environmental surveillance
(Image Credit: iStock/ sergunt)
