A zero-inflated Poisson (ZIP) regression model, applicable to count data with an excess of zeros, was adopted for genetic analysis of number of clinical mastitis cases in 36 178 first-lactation daughters of 245 sires. The ZIP model mixes a point mass at zero, with probability p, with a Poisson lambda distribution. Here, p is the probability of the 'perfect' state, an animal is resistant to mastitis, and lambda is the mean number of mastitis cases in the 'imperfect' case. A Bayesian approach via a Gibbs-Metropolis algorithm was used for inferring the unknown parameters. The most important source of variation in lambda was residual, followed by herd and sire of cow. The posterior mean (SD) of p was 0.21 (0.021), suggesting that there is a 21% of chance that a cow will belong to the perfect state and, hence, be resistant to mastitis.

M. R Motta, Daniel Gianola, Y. M Chang, B. Heringstad

Proceedings of the World Congress on Genetics Applied to Livestock Production, Volume , , 26.05, 2006
Download Full PDF BibTEX Citation Endnote Citation Search the Proceedings

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.