Comparison of linear-linear vs. linear- threshold models for estimation of genetic and environmental parameters using simulated data

In order to compare two linear-linear (LLM) and linear-threshold (LTM) models for estimation of genetic and environmental parameters, a population consisting of 6,700 animals was simulated using the Visual Basic programming language. The heritabilities for the first and second trait were assumed to be 0.30 and 0.10, respectively, and the genetic correlation between two traits was 0.50. Selection of parents was based on a two-trait selection index, and mating of selected parents conducted randomly for 15 years. The second trait was transformed into a categorical trait, assigning 50% ofobserved data in the first category, 34% in the second, and 16% in the third, using the probit function. In the LLM, genetic and environmental parameters were estimated under animal model specifications and derivative free restricted maximum likelihood methodology. In the LTM, predicted breeding values were obtained using a Bayesian method, implemented via Gibbs sampling procedure. The two models showed high ability in estimating genetic and environmental parameters for the continuous trait, and results were similar. Estimated genetic parameters for the threshold trait in LTM were higher than LLM, as the estimate of additive genetic variance and heritability was higher in LTM (0.084 and 0.10) than in the LLM (0.035 and 0.07) for the threshold trait. Therefore, the LTM offered advantages over the LLM in estimation of genetic and environment parameters, specifically in categorical trait) but due to a wide range of traits’ heritabilities and correlation between them under natural situation, further researchs usingfield data and other parameters are required.