Author Information

Brent D. Burch
Ian R. Harris

Abstract

For many researchers the restricted maximum likelihood (REML) method of estimation is the procedure of choice for estimating heritability. In most applications the REML estimate can only be obtained via an iterative method. In some cases the algorithm used to compute the REML estimate may be slow or fail to converge. These predicaments have provided the motivation to develop closed-form approximations to the REML estimator of heritability in mixed linear models having two variance components. These estimators are compared to the REML estimator by considering their large and small sample properties. We provide guidance on how to select the closed-form estimator that provides the best approximation to the REML estimator. A simple one-way random effects model and an animal breeding model with correlated genetic effects are presented.

Creative Commons License


This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Share

COinS
 
Apr 29th, 11:00 AM

ASSESSING THE PERFORMANCE OF CLOSED-FORM APPROXIMATIONS TO THE REML ESTIMATOR OF HERITABILITY

For many researchers the restricted maximum likelihood (REML) method of estimation is the procedure of choice for estimating heritability. In most applications the REML estimate can only be obtained via an iterative method. In some cases the algorithm used to compute the REML estimate may be slow or fail to converge. These predicaments have provided the motivation to develop closed-form approximations to the REML estimator of heritability in mixed linear models having two variance components. These estimators are compared to the REML estimator by considering their large and small sample properties. We provide guidance on how to select the closed-form estimator that provides the best approximation to the REML estimator. A simple one-way random effects model and an animal breeding model with correlated genetic effects are presented.