Likelihoods and simulation methods for a class of nonneutral population genetics models

Methods for simulating samples and sample statistics, under mutation-select ion-drift equilibrium for a class of nonneutral population genetics models, and for evaluating the likelihood surface, in selection and mutation param eters, are developed and applied for observed data. The methods apply to la rge populations in settings in which selection is weak, in the sense that s election intensities, like mutation rates, are of the order of the inverse of the population size. General diploid selection is allowed, but the appro ach is currently restricted to models, such as the infinite alleles model a nd certain K-models, in which the type of a mutant allele does not depend o n the type of its progenitor allele. The simulation methods have considerab le advantages over available alternatives. No other methods currently seem practicable for approximating likelihood surfaces.