The paper aims to analyze the structure of additive genetic groups as an alternative to include animals with unknown parentage in genetic evaluation. Several traits were studied: weaning weight; yearling weight; post-weaning weight gain in 345 days; scrotal circumference at 18 months of age; and visual muscling scores at 18 months of age. As a control group it was used a database where all animals had known paternity. Thus, three scenarios were defined where 30%; 50% or 70% of the observations were randomly taken as animals with unknown parentage. For these three simulated databases, the strategies considered to define additive genetic groups were: effect of year at birth; effect of farm at birth; effect of year and farm at birth; and a negative control group, which considered animals with unknown paternity. The most appropriate strategy for additive genetic group was chosen as the one that resulted in a higher regression coefficient and additive genetic value closest to the predicted for those animals in the control group. The results considering the strategy of additive genetic show large agreement with the control group. Among the strategies proposed, farm of birth and year of birth of the animal with unknown parentage achieved the best results.
correlation; livestock; linear regression; multiple sire; natural mating