Solving Mixed Model Equations in R
Description
Multivariate linear mixed model solver for estimation of heterogeneous variances and specification of variance covariance structures. Maximum and Restricted Maximum Likelihood (ML/REML) estimates can be obtained using the Direct-Inversion Newton-Raphson (NR), Direct-Inversion Average Information (AI), MME-based Expectation-Maximization (EM), and Efficient Mixed Model Association (EMMA) algorithms. Designed for genomic prediction and genome wide association studies (GWAS) to include additive, dominance and epistatic relationship structures or other covariance structures in R, but also functional as a regular multivariate mixed model software. Multivariate models (multiple responses) can be fitted currently with NR, AI and EMMA algorithms.