Skip to content


Genetics Selection Evolution

Open Access

Alternative implementations of Monte Carlo EM algorithms for likelihood inferences

Genetics Selection Evolution200133:443

Received: 14 September 2000

Accepted: 23 April 2001

Published: 15 July 2001


Two methods of computing Monte Carlo estimators of variance components using restricted maximum likelihood via the expectation-maximisation algorithm are reviewed. A third approach is suggested and the performance of the methods is compared using simulated data.


restricted maximum likelihoodMarkov chain Monte CarloEM algorithmMonte Carlo variancevariance components

(To access the full article, please see PDF)

Authors’ Affiliations

Departamento de Genética, Universidad de Zaragoza, Zaragoza, Spain
Section of Biometrical Genetics, Department of Animal Breeding and Genetics, Danish Institute of Agricultural Sciences, Tjele, Denmark


© INRA, EDP Sciences 2001