Hallander J, Waldmann P, Wang CK, Sillanpää MJ
Bayesian inferenceof genetic parameters based on conditional decompositions of multivariate normal distributions
Genetics: 2010 185:645-654

Abstract
It is widely recognized that the mixed linear model is an importanttool for parameter estimation in the analysis of complex pedigrees,which includes both pedigree and genomic information, and wheremutually dependent genetic factors are often assumed to followmultivariate normal distributions of high dimension. We havedeveloped a Bayesian statistical method based on the decompositionof the multivariate normal prior distribution into productsof conditional univariate distributions. This procedure permitscomputationally demanding genetic evaluations of complex pedigrees,within the user-friendly computer package WinBUGS. To demonstrateand evaluate the flexibility of the method, we analyzed twoexample pedigrees: a large noninbred pedigree of Scots pine(Pinus sylvestris L.) that includes additive and dominance polygenicrelationships and a simulated pedigree where genomic relationshipshave been calculated on the basis of a dense marker map. Theanalysis showed that our method was fast and provided accurateestimates and that it should therefore be a helpful tool forestimating genetic parameters of complex pedigrees quickly andreliably.

E-link to journal