Analyses of breeding values for simulated discrete visual score data with different distributions and different genetic parameters

The aim was to evaluate me effect of me form (Y) of assigning discrete visual scores (VS) based on a continuous underlying scale (US) on me estimares of me breeding values (BV) for two heritability values (H: 0.25 and 0.49) and two contemporary group variançevalues (GC: 0.04 and 0.16). Herds wim 40 bulls and 1,200 cows, mated at ranJam, were simulated for 20 years. Direçt and maternal BV; maternal permanent environmental, contemporary group and age of dam effects were generated and combined wim an independent error term to form me phenotype in me USo The VS data were assigned according to symmetric relative and fixed and asymmetric rdative distributions. The BV was estimated using a linear modeI wim me Gibbs SampIer. The procedure was repeated five times for each situation. CorreIation (R) between estimated and true BV was obtained for each animal cIass (sires, dams and olfspring).~ignificant elfects ofH on R for alI animal classes were found (0.49 presented greater R). For bulls, significant elfects ofH x GC interaction on R were found (greater H and smalIer GC presented greater R). Y had no elfect on R. Larger sampIes mar be needed for better evaluating me elfeçts of me factors on BV estimares for VS