gblup: prediction of genomic breeding values for use in GWA

In steibelj/gwaR: Functions for performing GWA from GBLUP

Description

A class for GBLUP prediction to implement GWA.

Usage

gblup(rsp, data, design, G, vdata = NULL, wt = NULL, ...)

## Default S3 method:
gblup(rsp, data, design, G, vdata = NULL, wt = NULL, ...)

Arguments

rsp	a string with the name of the response variable if a vector, first element is used.
data	either a gpData object from synbreed or a dataframe containing records for response variable and classification factors used as fixed or random effects
design	a list or a vector with formulas for fixed and random effects
G	a proportional covariance for genetics effects being predicted.
vdata	a list with proportional covariance matrices of additional random effects (not included in data)
wt	a vector of weigths (proportional to diagonals of residual covariance matrix), or a matrix of weights with anmimals in rows and response variables in columns
...	additional parameters for regress function
vdata	a list with proportional covariance matrices of additional random effects (not included in data)

Value

an object of the class gblup: a list the following components

• name trait name

• llik log-likelihood value at parameter estimates

• cycle number of iterations to convergence

• Q a matrix: see regress

• V phenotypic covariance matrix

• Vinv inverse of V

• sigma vector of estimated variance components

• coefm matrix of estimated fixed effects coefficients

• model the fit model

• ehat vector of residuals

• pos see regress

steibelj/gwaR documentation built on May 30, 2019, 10:45 a.m.