Supplemental data to publication "a fitted sparse-group lasso for genome-based evaluations"

Abstract: A simulation study has been conducted to analyse the association between genetic and phenotypic variation with grouped penalization approaches. The study design resembles a breeding population consisting of several half-sib families which is typical, for example, in livestock or crop. Genotypes and phenotypes have been simulated with publicly available software. Two strategies have been followed to generate genetic effects captured by markers: effects have been sampled independently (option a) or in groups (option b). The dataset includes a detailed description of the simulation design and comprises phased genotypes of parents and progeny, progeny phenotypes and marker effect sizes. The simulation has been repeated 100 times.