Comparison of repeatable and random regression models for genetic parameter estimation on Thoroughbreds
Introduction – The combination of multiple factors is effective in achieving the maximum level of race performance in thoroughbreds.
Aim – The aim of this study is to estimate and compare genetic parameters on the number of race success characteristics in Thoroughbreds with Random Regression Models (RRM) using L(3,3) Legendre polynomials and Repeatable animal model with Residual Maximum Likelihood (REML). It was also aimed to investigate which number of observation points would be sufficient for genetic parameter estimation for Thoroughbred.
Materials and methods – As data, 111312 test day race completion time (sec) records of 13625 thoroughbreds raced taken from the Jockey Club of Turkey between 2005 and 2016 were used. Competition performances were compared with different measurements using the same repeatability model. BV estimations for Thoroughbreds were obtained by using random regression models (RRM) and repeatable animal models with REML methods using DFREML and WOMBAT package, respectively.
Results and discussion – When AIC and BIC values were examined, it was observed that the values in RRM method were lower than REML method.
Conclusion – For this reason, RRM method can be preferred compared to REML method. Our results showed that number of race of five were sufficient to estimate genetic parameters for Thoroughbred horse.