SNP-Slice

SNP-Slice employs a slice sampling Markov Chain Monte Carlo algorithm to resolve mixed infections by learning SNP-haplotypes and linking them to hosts.

Key Features:

• SNP-haplotype reconstruction: Learns the SNP-haplotypes of all strains in populations without reference panels.
• Host-strain linkage: Determines which strains infect which hosts.
• Multiplicity of infection estimation: Outperforms state-of-the-art methods at estimating the multiplicity of infections.
• Allele frequency estimation: Accurately estimates allele frequencies from polygenomic data.
• Individual heterozygosity inference: Reconstructs individual heterozygosities accurately.
• Scalability: Provides a scalable approach to analyze polygenomic data.

Scientific Applications:

• Mixed infection resolution: Enables analysis of multi-strain infections that would otherwise be discarded due to requirements for monogenomic inputs.
• Molecular epidemiology: Supports understanding of genetic diversity, co-infection patterns, and genomic relatedness of pathogens.
• Molecular surveillance: Facilitates resolving complex infection patterns in pathogen surveillance programs.
• Malaria research: Illustrated on empirical malaria datasets with validated performance.
• HIV research: Demonstrated on empirical HIV datasets with recommendations for empirical use.

Methodology:

SNP-Slice uses a slice sampling Markov Chain Monte Carlo algorithm to learn SNP-haplotypes and link strains to hosts.