near neutrality

Ohta T & Tachida H 1990 Theoretical study of near neutrality. I. Heterozygosity and rate of mutant substitution. Genetics 126:219-229.

• the definition of near neutrality is that the product of effective population size (N) and selection coefficient (s) is not larger than unity

• our model is also different from the landscape model of Gillespie (1983)
• we are concerned with the interaction effect of random drift and natural selection
• Gillespie treated the situation of strong selection and weak mutation

• our model differs from the previous models of Ohta (1976), Kimura (1979) and others
• the distribution of selection coefficients is fixed regardless of the allelic state occupying the population
• this is the same as Kingman's house-of-cards model, if random drift is ignored
• the fixed model corresponds to that adopted by Zeng, Tachida and Cockerham (1989)

• in several previous attempts to formulate the nearly neutral mutation theory, the selection coefficients of new mutations were assumed to follow a certain distribution such as exponential (Ohta 1976) or gamma (Kimura 1979)
• these distributions have no probability density for s > 0
• i.e., all mutations are slightly deleterious

• the present model incorporates very slightly advantageous mutations for the region s > 0
• in this regard, our study is an attempt to expand the concept of near neutrality to the realm of the selectionists