Selective constraint in protein polymorphism: study of the effectively neutral mutation model by using an improved pseudosampling method.

To investigate the pattern of allelic distribution in enzyme polymorphism, with special reference to the relationship between the mean (H) and the variance (V(H)) of heterozygosity, we used the model of effectively neutral mutations involving multiple alleles in which selective disadvantage of mutant alleles follows a Gamma distribution. A simulation method was developed that enables us to study efficiently the process of random drift in a multiallelic genetic system and that saves a great deal of computer time. It is an improved version of the pseudosampling-variable (PSV) method [Kimura, M. (1980) Proc. Natl. Acad. Sci. USA 77, 522-526] previously used to simulate random drift in a diallelic system. This method will be useful for simulating many models of population genetics that involve behavior of multiple alleles in a finite population. By using this method, it was shown that, as compared with the model of strictly neutral mutations, the present model gives the reduction of both H and V(H) and an excess of rare variant alleles. The results were discussed in the light of recent observations on protein polymorphism with special reference to the functional constraint of proteins involved.