nxk PCA matrix from the predictor variables.

PCAs

vector of names for the population without the test set.

candidates

name of the individuals in the test or VP set.

Test

number of individuals to select in the training population.

ntoselect

number of solutions at each iteration.

npop

number of elite solutions for TP pops.

nelite

probability of mutation for each solution

mutprob

niterations

lambda

This function is a wrapper from the STPGA package to obtain the best subset of individuals to predict a group of individuals minimizing the predicted error variance (PEV). Is used internally in the TP.prep function.

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.

PEV: Selecting the best training population for genomic selection

Description

Usage

Arguments

Value

References

See Also