Censored Bayesian models for genetic evaluation of age at first calving in Brazilian Brahman cattle

Abstract

We compared different Bayesian models to handle censored data for genetic parameters estimation of age at first calving (AFC) in Brazilian Brahman cattle. Data from females with AFC above 1825 days of age were assumed to have failed to calve and were considered as censored records. Data including information of 53,703 cows were analyzed through the following methods: conventional linear model method (LM), which consider only uncensored records; simulation method (SM), in which the data were augmented by drawing random samples from positive truncated normal distributions; penalty method (PM), in which a constant of 21 days was added to censored records; and the bivariate threshold-linear method (TLcens). The LM was the most suited for genetic evaluation of AFC in Brazilian Brahman cattle based on the predictive ability evaluation through cross-validation analysis. The similar results for LM and PM regarding Spearman correlations, and the higher percentages of selected animals in common, indicated that there was not relevant reranking of animals when censored records were used. In summary, the heritability estimates for AFC ranged from 0.09 (TLcens) to 0.20 (LM). Given its poor predictive performance, the SM is not recommended for handling censored records for genetic evaluation of AFC.

Impact of Censored or Penalized Data in the Genetic Evaluation of Two Longevity Indicator Traits Using Random Regression Models in North American Angus Cattle

“…Setiaji et al [ 39 ], analyzing different penalty methods to access interval from the first to the last successful insemination in Japanese Black heifers, concluded that EBV correlations decrease at higher levels of censored (missing) records. Using only uncensored data and the conventional linear model was also the most recommended method for the genetic evaluation of age at first calving in Brahman cattle [ 11 ]. However, the authors pointed out that similar EBV correlations were observed using either uncensored or penalized data, which suggests that there was not relevant re-ranking of animals when censored records were used [ 11 ].…”

“…Using only uncensored data and the conventional linear model was also the most recommended method for the genetic evaluation of age at first calving in Brahman cattle [ 11 ]. However, the authors pointed out that similar EBV correlations were observed using either uncensored or penalized data, which suggests that there was not relevant re-ranking of animals when censored records were used [ 11 ]. Similarly, Costa et al [ 40 ] also found that the linear-threshold model without censored data showed the best predictive ability (computed based on the EBV correlation estimated between the complete and reduced datasets) for the genetic evaluation of both age at first calving and stayability in Nellore cattle.…”

“…The penalty method [ 10 ] has been considered suitable to treat censored records in the genetic evaluations of fertility related traits, which consists on applying a penalization criterion for animals with censored records. For age at first calving, for example, censored records are replaced by a set of augmented data (i.e., penalized data), which is obtained by adding a constant value of 21 days to the highest value of age at the first calving within each contemporary group [ 9 , 10 , 11 ]. Specifically for longevity-related traits, penalized data can be obtained by including a culling date for cows without culling information recorded in the dataset.…”

“…Therefore, the greatest challenge is to defined the optimal threshold to create the penalized data for longevity. Using the augmented data instead the censured records has been shown to increase the prediction accuracy of EBVs for fertility in some studies [ 11 , 12 ]. However, to our best knowledge, the use of the penalty method has not yet been investigated for the genetic evaluation of longevity indicator traits, especially when using random regression models (RRM).…”