poolfstat

poolfstat computes f-statistics and constructs admixture graphs from allele count and pool-seq read count data to infer demographic history across populations.

Key Features:

• Data types supported: Computes statistics from allele count data and pooled sequencing (pool-seq) read count data.
• Implementation: Provided as an R package for performing population genetics analyses.
• Unified F-statistics framework: Reinterprets F and D parameters in terms of the probability of gene identity to provide a unified framework for allele count and pool-seq data.
• Unbiased estimators for pooled data: Implements unbiased estimators applicable to pooled sequencing without requiring haplotype information.
• Three-Population Test: Performs the three-population test to assess admixture among three populations.
• Four-Population Test: Performs the four-population test to evaluate treeness among four populations.
• F4-ratio estimation: Estimates admixture rates between populations using F4-ratios.
• Admixture graph fitting and construction: Fits admixture graphs, supports visualization, and offers semi-automatic graph construction.
• Evaluation on simulated data: Includes evaluation on simulated pool-seq data across varying sequencing coverages and error rates to assess accuracy, including low-coverage designs.

Scientific Applications:

• Demographic inference: Infers admixture, migration, and population splits from genetic variation using both individual genotype and pooled sequencing data.
• Drosophila suzukii case study: Analyzed pool-seq data from 14 populations of Drosophila suzukii to refine demographic history and invasion routes.

Methodology:

Computes f-statistics from allele count and pool-seq read count data using a probability-of-gene-identity framework with unbiased estimators, applies three- and four-population tests and F4-ratio estimation, fits and visualizes admixture graphs with semi-automatic construction, and evaluates performance on simulated pool-seq data across varying coverage and error rates.