breedR-package: Frequentist and Bayesian methods for breeders, quantitative...

In famuvie/breedR: Statistical Methods for Forest Genetic Resources Analysts

Frequentist and Bayesian methods for breeders, quantitative geneticists and forest genetic resources analysts.

Description

This package provides statistical tools to build predictive models for the breeders, quantitative genetists and forest genetic resources analysts communities. It aims to assess the genetic value of individuals under a number of situations, including spatial autocorrelation, genetic/environment interaction and competition. It is under active development as part of the Trees4Future project, particularly developed having forest genetic trials in mind. But can be used for animals or other situations as well.

Details

The package functionality builds up on a wrapping up of Ignacy Misztal's progsf90 suite of Fortran programs. Particularly, the function reml performs classical Restricted-Maximum Likelihood inference by interfacing Misztal's programs with several high-level options such as spatial components, etc. The Fortran back-end allows for fast inference on rather large datasets (hundreds of thousands of individuals) with complex pedigrees.

Author(s)

Maintainer: Facundo Muñoz [e-mail]

Authors:

• Leopoldo Sanchez

Other contributors:

• Ignacy Misztal [contributor]

• Pablo Cappa [contributor]

• Timothée Flutre [contributor]

References

Most functionality in the package is based on Ignacy Misztal's suite of Fortran programs for mixed model computations in breeding. http://nce.ads.uga.edu/wiki/doku.php

M. Lynch \& B. Walsh (1998). Genetics and Analysis of Quantitative Traits. Sinauer Associates, Inc.

See Also

pedigreemm

Examples

  # Load, summarize and visualize data
  data(m4)
  summary(m4)
  plot(m4)

  # Fit Mixed Model using REML
  res.f90 <- remlf90(fixed = phe_X ~ gen, 
                     genetic = list(model = 'add_animal', 
                                    pedigree = get_pedigree(m4),
                                    id = 'self'),
                     data = as.data.frame(m4))
  
  # Summary of results
  summary(res.f90)
  
  # Observed phenotypes vs. Fitted values
  library(ggplot2)
  qplot(phe_X, fitted(res.f90), color=gen, data = as.data.frame(m4)) +
    geom_abline(intercept=0, slope=1)

famuvie/breedR documentation built on Aug. 6, 2024, 9:10 p.m.