Haploi

Haploi infers haplotypes from multi-population genotype data and estimates the number and configuration of haplotype ancestor pools for analyses of SNP variation.

Key Features:

• Nonparametric Bayesian Model: Utilizes a hierarchical Dirichlet process (HDP), a nonparametric Bayesian model that serves as a surrogate for the coalescent process, enabling flexible modeling of haplotype distributions.
• Multi-Population Capability: Handles genotype data from multiple ethnically distinct populations and leverages individual ethnic information to inform inference.
• Statistical Framework: Provides posterior inference of individual haplotypes and the size and configuration of haplotype ancestor pools for datasets with thousands of single nucleotide polymorphisms (SNPs).
• Exchangeable and Unbounded Model: Employs an exchangeable and unbounded model that couples demographic information across different populations.
• Performance: Often demonstrates superior speed and accuracy compared to existing state-of-the-art programs.

Scientific Applications:

• Biological and Medical Research: Inferring haplotypes of SNPs in studies where haplotype structure is relevant to biology or medicine.
• Analyses of Heterogeneous Populations: Processing large-scale genotype data from diverse human subpopulations to characterize haplotype variation.
• Complex SNP Datasets: Performing analyses involving thousands of SNPs that require estimation of haplotype ancestor pool size and configuration.

Methodology:

Implements a hierarchical Dirichlet process (HDP) nonparametric Bayesian model as a surrogate for the coalescent process to perform posterior inference of individual haplotypes and haplotype ancestor pool size and configuration across multiple populations.