SNP-Slice Resolves Mixed Infections: Simultaneously Unveiling Strain Haplotypes and Linking Them to Hosts

Area of Science:

• Molecular epidemiology
• Computational biology
• Infectious disease dynamics

Background:

• Multi-strain pathogen infections are common but challenging to analyze.
• Existing methods require discarding mixed infection samples, hindering research on genetic diversity and co-infection.
• A tool to resolve single-nucleotide polymorphism (SNP) haplotypes from polygenomic data is crucial.

Approach:

• Developed SNP-Slice, a slice sampling Markov Chain Monte Carlo algorithm.
• SNP-Slice learns SNP haplotypes and identifies infecting strains from mixed samples.
• The method reconstructs haplotypes and heterozygosities without reference panels.

Key Points:

• SNP-Slice accurately estimates infection multiplicity and allele frequencies.
• Outperforms existing state-of-the-art methods in resolving polygenomic SNP data.
• Demonstrated effectiveness on malaria and HIV empirical datasets.

Conclusions:

• SNP-Slice offers a novel approach for analyzing polygenomic pathogen data.
• Enables deeper insights into complex infection patterns and pathogen genomic relatedness.
• Opens new avenues for molecular surveillance and understanding infectious disease evolution.