動物育種におけるシミュレーテッド・アニーリングを利用した選抜と交配の最適化

Abstract

Artificial selection results in an inflation of inbreeding through reduced effective population size. In recent years, various selection and mating procedures have been proposed to reduce rates of inbreeding in breeding populations. Since these procedures are generally based on optimal solutions of mathematical programming with a huge number of variables, they could not be implemented in most of the practical situations. To overcome this problem, the use of quasi-optimal solution via simulated annealing was considered in this study. The simulated annealing algorithm is an analogy to the fact that if a metal is cooled slowly, the atoms find their optimal positions to achieve a state of minimum energy. The algorithm was illustrated with the problem for finding mating allocation to give minimum inbreeding (minimum coancestry mating). Effect of parameters to control the imaginary temperature was examined in simulated populations, and a reasonable set of parameters in the practical applications was suggested. Using the simulated annealing with the suggested parameters, the effectiveness of the minimum coancestry mating and selection for maximizing genetic diversity was evaluated in simulated pig and broiler populations.