Genet. Sel. Evol. 40 (2008) 161-176
DOI: 10.1051/gse:2007042

A comparison of strategies for Markov chain Monte Carlo computation in quantitative genetics

Rasmus Waagepetersen¹, Noelia Ibánez-Escriche² and Daniel Sorensen<sup>3

1</sup>  Department of Mathematical Sciences, Aalborg University, 9220 Aalborg, Denmark
²  IRTA, Avda. Rovira Roure, 25198 Lleida, Spain
³  Department of Genetics and Biotechnology, Danish Institute of Agricultural Sciences, P.O. Box 50, 8830 Tjele, Denmark

(Received 14 February 2007; accepted 7 September 2007; published online 27 February 2008)

Abstract - In quantitative genetics, Markov chain Monte Carlo (MCMC) methods are indispensable for statistical inference in non-standard models like generalized linear models with genetic random effects or models with genetically structured variance heterogeneity. A particular challenge for MCMC applications in quantitative genetics is to obtain efficient updates of the high-dimensional vectors of genetic random effects and the associated covariance parameters. We discuss various strategies to approach this problem including reparameterization, Langevin-Hastings updates, and updates based on normal approximations. The methods are compared in applications to Bayesian inference for three data sets using a model with genetically structured variance heterogeneity

Key words: Langevin-Hastings / Markov chain Monte Carlo / normal approximation / proposal distributions / reparameterization

Correspondence and reprints: rw@math.aau.dk

© INRA, EDP Sciences 2008