Algoritmo para minimização da endogamia e maximização de índice econômico em sistemas de acasalamento

Abstract

Animal breeding processes seek to increase the quality of the herds and the profitability of rural producers. Selection and mating systems form the core of breeding systems, aiming to target the offspring’s genetic and phenotypic traits through the combination of parental traits. The selection and mating processes must consider the coancestry between the animals in the herd, avoiding the crossing of related animals to minimize inbreeding. Inbreeding occurs from matings between related individuals, which can be useful for setting standards but generate health problems, less rusticity, and lower genetic value, consequently harming the producer. This work proposes an algorithmic solution for the animal breeding process to minimize coancestry while maximizing the economic index, seeking a balance between these values through a weighting parameter of the algorithm. The developed algorithm aims to guarantee an optimal mating scheme, considering the most important and desired aspects within the scope of animal breeding. For this, the problem was modeledas a binary linear programming problem. To treat it, a combination of simplex techniques is used to solve the relaxed linear programming problem and branch-and-bound to find the best integer solution. The algorithm called CoancestryMIN-Selection is finished and in the future it may be integrated with the already developed graph-based database. It was created in the R and C++ languages with the RStudio IDE. The developed algorithm is able to perform the computation of the pedigree matrix of the herd, always ensuring an optimal solution for the problem in question. Analyzes carried out based on the tests indicate that the minimization of inbreeding is a matter of greater complexity than the maximization of the economic index.