Genetic improvement of feed efficiency (FE) is an important issue since feed is the major cost in dairy farms. However, individual feed intake is difficult or costly to measure, which limits selection accuracy. Genomic selection is expected to be a promising tool to start selection for FE. The objective of this study is to investigate accuracy of genomic prediction for dry matter intake (DMI) in the case of a small reference data in Nordic Holsteins. Two data sets were used in this study. The first data set included 53,914 week-records of DMI records from 1,623 lactations of 827 cows. Among these cows, 330 individuals were genotyped with 54k chip. The other data set included 5,409 progeny-tested bulls with marker data of 54k chip and de-regressed proof of milk yield. Breeding value were predicted using a univariate and a bivariate (with milk yield as assistant trait) conventional BLUP model, as well as a univariate single-step model (ssGBLUP) and a bivariate ssGBLUP model. Using single-trait analysis, correlation between EBV and corrected 305d DMI for genotyped animals increased from 0.287 to 0.393 in validation scenario without sibs in training data, and from 0.366 to 0.432 in scenario allowing sibs in training data, when moving conventional BLUP model to ssGBLUP model. The ssGBLUP also led to a slight improvement of EBV accuracy for non-genotyped animals. The bivariate models using milk yield information resulted in a considerable increase of model-based reliability for both genotyped and non-genotyped animals, but an increase of correlation between EBV and corrected 305d DMI only for genotyped animals. Although we used a small reference population, the results indicated that genomic prediction leads to better prediction accuracy than conventional BLUP method. To obtain a higher accuracy of EBV for DMI, more data are required. Keywords: dairy cattle, dry matter intake, feed efficiency, genomic prediction, single-step GBLUP
Proceedings of the World Congress on Genetics Applied to Livestock Production, Volume Electronic Poster Session - Methods and Tools - Prediction 2, , 474, 2018
|Download Full PDF||BibTEX Citation||Endnote Citation||Search the Proceedings|
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.