Large deviations for the Fleming–Viot process with neutral mutation and selection, II

Large deviation principles are established for the Fleming–Viot process with neutral mutation and with selection, and the associated equilibrium measures as the sampling rate approaches zero and when the state space is equipped with the weak topology. The path-level large deviation results improve the results of Dawson and Feng (1998, Stochastic Process. Appl. 77, 207–232) in three aspects: the state space is more natural, the initial condition is relaxed, and a large deviation principle is established for the Fleming–Viot process with selection. These improvements are achieved through a detailed study of the behaviour near the boundary of the Fleming–Viot process with finite types.