Reducing the bias of estimates of genotype by environment interactions in random regression sire models 英文参考文献.docVIP

Genetics Selection Evolution BioMedCentral Research Open Access Reducing the bias of estimates of genotype by environment interactions in random regression sire models Marie Lillehammer*1, J?rgen ?deg?rd1,2 and Theo HE Meuwissen1 Address: 1Department of Animal and Aquacultural Sciences, Norwegian University of Life Sciences, N-1432 ?s, Norway and 2NOFIMA, N-1432 ?s, Norway Email: Marie Lillehammer* - marie.lillehammer@ssb.no; J?rgen ?deg?rd - jorgen.odegard@umb.no; Theo HE Meuwissen - theo.meuwissen@umb.no * Corresponding author Published: 19 March 2009 Received: 10 March 2009 Accepted: 19 March 2009 Genetics Selection Evolution 2009, 41:30 doi:10.1186/1297-9686-41-30 This article is available from: /content/41/1/30 ? 2009 Lillehammer et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Abstract The combination of a sire model and a random regression term describing genotype by environment interactions may lead to biased estimates of genetic variance components because of heterogeneous residual variance. In order to test different models, simulated data with genotype by environment interactions, and dairy cattle data assumed to contain such interactions, were analyzed. Two animal models were compared to four sire models. Models differed in their ability to handle heterogeneous variance from different sources. Including an individual effect with a (co)variance matrix restricted to three times the sire (co)variance matrix permitted the modeling of the additive genetic variance not covered by the sire effect. This made the ability of sire models to handle heterogeneous genetic variance approximately equivalent to that of animal models. When residual variance was hete