Inference in two dimensions: allele frequencies versus lengths of shared sequence blocks

Area of Science:

• Population Genetics
• Evolutionary Biology
• Genomics

Background:

• Understanding population structure and gene flow is crucial for evolutionary studies.
• Estimating key parameters like neighborhood size (N) and dispersal rate (σ(2)) is challenging over intermediate timescales.

Purpose of the Study:

• To develop and present two novel approaches for inferring neighborhood size (N) and dispersal rate (σ(2)).
• To analyze population genetics over intermediate timescales (10-100 generations) using spatial correlations and sequence block lengths.

Main Methods:

• Inference based on allele frequencies and spatial correlations, assuming a Gaussian distribution for maximum likelihood estimates of N and κ.
• Inference based on the distribution of lengths of identical-by-descent sequence blocks, utilizing the Wright-Malécot formula for large N and a geometric distribution for small N.

Main Results:

• Spatial correlations of allele frequencies allow estimation of N and a local scale parameter (κ), but not dispersal rate (σ(2)).
• The distribution of long sequence blocks (>0.1 cM) directly relates to N and σ(2) under a quasi-equilibrium assumption.
• For small N, the distribution of identical-by-state sequence block lengths is geometric, dependent on N and σ(2).

Conclusions:

• Two distinct methods provide robust estimates of neighborhood size and dispersal rates in populations.
• These approaches are applicable to intermediate timescales, offering valuable tools for population genetic inference.
• The methods account for population structure and history, enhancing the accuracy of genetic parameter estimation.