Dual processes and an invariance result for exchangeable models in population genetics

Area of Science:

• Population genetics
• Stochastic processes
• Mathematical biology

Background:

• Exchangeable models are fundamental in population genetics.
• Understanding large population dynamics is crucial for evolutionary insights.
• Dual processes offer a powerful analytical framework.

Purpose of the Study:

• To construct dual processes for a broad range of exchangeable population genetics models.
• To investigate the behavior of these dual processes in large populations.
• To leverage duality for analyzing model properties and deriving approximations.

Main Methods:

• Representation of models as interactive particle systems.
• Construction of dual processes based on these systems.
• Analysis of limiting behavior as population size increases.
• Application of duality relationships for quantitative analysis.

Main Results:

• A tractable limiting dual process is identified for large populations.
• The duality relationship simplifies the analysis of various population genetics models.
• Approximate expressions for key quantities are derived.
• Diffusion approximations are readily obtained from invariance properties.

Conclusions:

• Dual processes provide an effective method for studying exchangeable population genetics models.
• The identified limiting dual process offers significant analytical advantages.
• The approach facilitates the derivation of approximations and diffusion models for large populations.